Successful Trouble Shooting for Process Engineers: A Complete Course in Case Studies

Successful Trouble Shooting for Process Engineers D. R. Woods Successful Trouble Shooting for Process Engineers. Don Wo...

Author: Donald R. Woods

137 downloads 1622 Views 4MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Successful Trouble Shooting for Process Engineers D. R. Woods

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Further Titles of Interest: Bchel, K. H., Moretto, H.-H., Woditsch, P.

Hattwig, M., Steen, H. (Eds.)

Industrial Inorganic Chemistry

Handbook of Explosion Prevention

Second, Completely Revised Edition 2000 ISBN 3-527-29849-5

Weissermel, K., Arpe, H.-J.

Industrial Organic Chemistry

2004 ISBN 3-527-30718-4

Oetjen, G.-W., Haseley, P.

Freeze-Drying

Fourth, Completely Revised Edition

Second, Completely Revised and Extended Edition

2003 ISBN 3-527-305789-5

2004 ISBN 3-537-30620-X

Mollet, H., Grubenmann, A.

Hagen, J.

Formulation Technology

Industrial Catalysis

Emulsions, Suspensions, Solid Forms

A Practical Approach

2001 ISBN 3-527-30201-8

1999 ISBN 3-527-29528-3

Sundmacher, K., Kienle, A. (Eds.)

Jakobi, R.

Reactive Distillation

Marketing and Sales in the Chemical Industry

Status and Future Directions 2003 ISBN 3-527-30579-3

Rauch, J. (Ed.)

Multipurpose Plants 2003 ISBN 3-527-29570-4

Elias, H. G.

An Introduction to Plastics Second, Completely Revised Edition 2003 ISBN 3-527-29602-6

Second, Completely Revised Edition 2002 ISBN 3-527-30625-0

Bamfield, P.

Research and Development Management In the Chemical and Pharmaceutical Industry Second, Completely Revised and Extended Edition 2003 ISBN 3-527-30667-6

Donald R. Woods

Successful Touble Shooting for Process Engineers A Complete Course in Case Studies

Author Prof. Donald R. Woods Chemical Engineering Department McMaster University Hamilton Ontario Canada, L8S 4L7

&

All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate. Library of Congress Card No.: applied for British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. Bibliographic information published by Die Deutsche Bibliothek Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data is available in the Internet at . 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – nor transmitted or translated into machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law. Printed in the Federal Republic of Germany. Printed on acid-free paper. Typesetting Khn & Weyh, Satz und Medien, Freiburg Printing betz-druck GmbH, Darmstadt Bookbinding J. Schffer GmbH, Grnstadt ISBN-13: ISBN-10:

978-3-527-31163-7 3-527-31163-7

V

Contents Preface 1 1.1 1.1.1 1.1.2 1.2 1.2.1 1.2.2 1.3 1.4 1.5 1.6

XIII What is Trouble Shooting? 1 Characteristics of a Trouble-Shooting Problem 2 Similarities among TS Problems 2 Differences between TS Problems 3 Characteristics of the Process Used to Solve Trouble-Shooting Problems 3 How the Type of Problem Guides the TS Process or Strategy 3 Five Key Elements Common to the TS Process 4 Self-Test and Reflections 5 Overview of the Book 9 Summary 9 Cases to Consider 9

2.2.2 2.2.3 2.3 2.3.1 2.3.2 2.4 2.5 2.6

The Mental Problem-Solving Process used in Trouble Shooting 17 Problem Solving 19 Trouble Shooting 23 Considerations when Applying the Strategy to Solve Trouble-Shooting Problems 23 Problem-Solving Processes Used by Skilled Trouble Shooters 24 Data Collection and Analysis: Approaches Used to Test Hypotheses 25 Overall Summary of Major Skills and a Worksheet 25 Getting Organized: the Use of a Trouble-Shooter’s Worksheet 25 Feedback about your Trouble Shooting 29 Example Use of the Trouble-Shooter’s Worksheet 35 Summary 40 Cases to Consider 40

3 3.1 3.1.1 3.1.2

Rules of Thumb for Trouble Shooting 43 Overall 43 General Rules of Thumb and Typical Causes Corrosion as a Cause 45

2 2.1 2.2 2.2.1

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

43

VI

Contents

3.1.3 3.1.4 3.1.5 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.3 3.3.1 3.3.2 3.3.3 3.3.4 3.3.5 3.3.6 3.4 3.4.1 3.4.2 3.4.3 3.4.4 3.4.5 3.4.6 3.4.7 3.4.8 3.4.9 3.4.10 3.4.11 3.4.12 3.5 3.5.1 3.5.2 3.5.3 3.5.4 3.5.5 3.5.6 3.5.7 3.5.8 3.5.9 3.5.10 3.5.11 3.5.12 3.5.13 3.6 3.6.1

Instruments, Valves and Controllers 46 Rules of Thumb for People 47 Trouble-Shooting Teams 48 Transportation Problems 51 Gas Moving: Pressure Service 52 Gas Moving: Vacuum Service 53 Liquid 54 Solids 56 Steam 58 Energy Exchange 58 Drives 58 Thermal Energy: Furnaces 60 Thermal Energy: Fluid Heat Exchangers, Condensers and Boilers Thermal Energy: Refrigeration 65 Thermal Energy: Steam Generation 66 High-Temperature Heat-Transfer Fluids 66 Homogeneous Separation 67 Evaporation 67 Distillation 69 Solution Crystallization 72 Gas Absorption 73 Gas Desorption/Stripping 75 Solvent Extraction, SX 76 Adsorption: Gas 77 Adsorption: Liquid 77 Ion Exchange 77 Membranes: Reverse Osmosis, RO 79 Membranes: Nanofiltration 79 Membranes: Ultrafiltration, UF, and Microfiltration 79 Heterogeneous Separations 79 Gas–Liquid 80 Gas–Solid 81 Liquid–Liquid 82 Gas–Liquid–Liquid Separators 84 Dryer for GS Separation 85 Screens for Liquid Solid Separation or Dewatering 85 Settlers for LS Separation 86 Hydrocyclones for LS Separation 86 Thickener for LS Separation 86 Sedimentation Centrifuges 87 Filtering Centrifuge 87 Filter for LS Separation 88 Screens for Solid–Solid Separation 88 Reactor Problems 88 PFTR: Multitube Fixed-Bed Catalyst, Nonadiabatic 89

61

Contents

3.6.2 3.6.3 3.6.4 3.6.5 3.6.6 3.6.7 3.6.8 3.6.9 3.6.10 3.6.11 3.6.12 3.7 3.7.1 3.7.2 3.7.3 3.8 3.8.1 3.8.2 3.8.3 3.9 3.9.1 3.9.2 3.9.3 3.9.4 3.9.5 3.9.6 3.9.7 3.10 3.11 3.12 3.12.1 3.12.2

PFTR: Fixed-Bed Catalyst in Vessel: Adiabatic 91 PFTR: Bubble Reactors, Tray Column Reactors 93 PFTR: Packed Reactors 94 PFTR: Trickle Bed 94 PFTR: Thin Film 96 STR: Batch (Backmix) 96 STR: Semibatch 98 CSTR: Mechanical Mixer (Backmix) 99 STR: Fluidized Bed (Backmix) 101 Mix of CSTR, PFTR with Recycle 106 Reactive Extrusion 106 Mixing Problems 107 Mechanical Agitation of Liquid 107 Mechanical Mixing of Liquid–Solid 108 Solids Blending 108 Size-Decrease Problems 109 Gas Breakup in Liquid: Bubble Columns 109 Gas Breakup in Liquid: Packed Columns 109 Gas Breakup in Liquid: Agitated Tanks: 110 Size Enlargement 110 Size Enlargement: Liquid–Gas: Demisters 110 Size Enlargement: Liquid–Liquid: Coalescers 110 Size Enlargement: Solid in Liquid: Coagulation/Flocculation 111 Size Enlargement: Solids: Tabletting 111 Size Enlargement: Solids: Pelleting 111 Solids: Modify Size and Shape: Injection Molding and Extruders 112 Coating 126 Vessels, Bins, Hoppers and Storage Tanks 126 “Systems” Thinking 127 Health, Fire and Stability 130 Individual Species 130 Combinations 131

4 4.1 4.2 4.3 4.4 4.5 4.6

Trouble Shooting in Action: Examples 133 Case ’3: The Case of the Cycling Column 133 Case ’4: Platformer Fires 138 Case ’5: The Sulfuric Acid Pump 141 Case ’6: The Case of the Utility Dryer 144 Case ’7: The Case of the Reluctant Crystallizer 157 Reflections about these Examples 162

5 5.1 5.1.1 5.1.2

Polishing Your Skills: Problem-Solving Process 165 Developing Awareness of the Problem-Solving Process 165 Some Target Skills 166 The TAPPS Roles: Talker and Listener 166

VII

VIII

Contents

5.1.3 5.1.4 5.2 5.2.1 5.2.2 5.2.3 5.2.4 5.3 5.3.1 5.3.2 5.4 5.4.1 5.4.2 5.4.3 5.4.4 5.5 5.5.1 5.5.2 5.5.3 5.6 6 6.1 6.1.1 6.1.2 6.1.3 6.1.4 6.2 6.2.1 6.2.2 6.2.3 6.2.4

6.2.5 6.2.6 6.3 6.3.1 6.3.2 6.4 6.4.1

Activity 5.1: (35 minutes) 168 Feedback, Self-Assessment 172 Strategies 173 Some Target Skills 174 The Extended TAPPS Roles: Talker+ and Listener+ 175 Activity 5.2: (time 35 minutes) 176 Feedback, Self-assessment 179 Exploring the “Context”: what is the Real Problem? 180 Example 180 Activity 5-3 181 Creativity 183 Some Target Skills 183 Example: Case ’10: To dry or not to dry! (based on Krishnaswamy and Parker, 1984) 186 Activity 5-4 190 Feedback, Self-Assessment 191 Self-Assessment 191 Some Target Skills 192 Activity for Growth in Self-Assessment 192 Feedback About Assessment 193 Summary and Self-Rating 194 Polishing Your Skills: Gathering Data and the Critical-Thinking Process 195 Thinking Skills: How to Select Valid Diagnostic Actions 196 How to Select a Diagnostic Action 196 Select from among a Range of Diagnostic Actions 196 More on Gathering and Interpreting Data 200 Summary 209 Thinking Skill: Consistency: Definitions, Cause–Effect and Fundamentals 209 Consistent Use of Definitions 210 Consistent with How Equipment Works: Cause ﬁ Effects: Root Cause-Symptoms 212 Consistent with Fundamental Rules of Mathematics and English 217 Consistent with Fundamental Principles Of Science: Conservation of Mass, Energy, High to Low Pressure, Properties of Materials 218 Consistent with Experience 218 Summary 219 Thinking Skills: Classification 219 Classify the Starting Information 219 Classifying Ideas from Brainstorming 220 Thinking Skills: Recognizing Patterns 221 Patterns in the Symptoms 221

Contents

6.4.2 6.5 6.5.1 6.5.2 6.5.3 6.5.4 6.5.5 6.5.5a 6.5.5b 6.5.5c 6.5.5d 6.5.6 6.5.7 6.5.8 6.5.9 6.5.10 6.6 6.7 6.8

Patterns in the Evidence 223 Thinking Skill: Reasoning 223 Step 1: Classify the Information 224 Step 2: Write the Conclusion 225 Step 3: Identify the Context 225 Step 4: Clarify the Meaning of the Terminology 226 Step 5: Consider the Evidence 227 Identify the Evidence 227 Check for Consistency 227 Which Evidence is Pertinent? 228 Diagram the Argument 229 Step 6: Formulate the Assumptions 230 Step 7: Assess the Quality of the Reasoning 230 Step 8: Assess the Strengths of the Counterarguments 232 Step 9: Evaluate the Consequences and Implications 232 Activity 6-14 232 Feedback and Self-Assessment 233 Summary 233 Exercises 234

7 7.1 7.1.1 7.1.2 7.1.3 7.1.4 7.1.5 7.2 7.2.1 7.2.2 7.2.3 7.2.4 7.2.5 7.3 7.4 7.5

Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance 237 Interpersonal Skills 237 Communication 237 Listening 238 Fundamentals of Interaction 239 Trust 240 Building on Another’s Personal Uniqueness 243 Factors that Affect Personal Performance 244 Pride and Unwillingness to Admit Error 244 Stress: Low and High Stress Errors 245 Alienation and Lack of Motivation 249 “I Know Best!” Attitude 249 Tendency to Interpret 249 The Environment 253 Summary 253 Exercises and Activities 253

8 8.1 8.1.1 8.1.2 8.2 8.2.1

Prescription for Improvement: Put it all Together Approaches to Polish Your Skill 259 Triad Activity 259 Individual Activity 262 Cases to Help you Polish Your Skill 263 Guidelines for Selecting a Case 263

259

IX

X

Contents

8.2.2 8.3 9 9.1 9.2 9.3 9.3.1 9.3.2 9.4 9.5

The Cases and Understanding the Choice of Diagnostic Actions for each Case 263 Summary 396 What Next? 397 Summary of Highlights 397 Reflection and Self-Assessment are Vital for the Development of Confidence 401 Going Beyond this Book: Setting Goals for Improvement 402 Prepare Yourself for Success 402 Use Reflection and Self-Assessment Effectively 403 Going Beyond this Book: Updating your Rules of Thumb and Symptom ‹ Cause Data for Process Equipment 403 Beyond this Book: Sources of Other Cases 403

Literature References Index

405

I 1

CD Contents Appendix A Feedback about Experience with Process Equipment Appendix B Improving “Systems Thinking”

411

415

Appendix C Feedback on the Cases in Chapters 1, 2 and 7 423 Appendix D Coded Answers for the Questions Posed to Solve the Cases Appendix E Debrief for the Trouble-Shooting Cases

435

537

Appendix F Other Tasks for the Skill-Development Activities in Chapter 5 Appendix G Selected Responses to the Activities in Chapters 6 and 7 Appendix H Data about “Causes” for Selected Process Equipment

569

573

565

Contents

Appendix I Feedback about Symptoms for Selected Causes

579

Appendix J Guide for Students: How You Can Get the Most from this Book 581 J-1 Getting Started: Get the Big Picture 581 J-2 Try a Trouble-Shooting Case where the Problem is Reasonably Well Defined 582 J-3 See How Others Handle a Case 591 J-4 Pause, Reflect on the Pretest, and Invest Time Polishing Specific Skills 591 J-5 Work your First Cases Starting with Case ’19 591 J-6 Trouble Shooting on the Job 591 J-7 Summary 592

XI

XIII

Preface My McMaster University colleague Tom Marlin describes trouble shooting as “the bread and butter” activity of engineering. Indeed, the financial health of a process unit depends so much on the skill of the engineers to trouble shoot problems promptly, safely and effectively. Training in trouble shooting should be part of every undergraduate engineer’s education. Yet, it rarely is, even though the introduction of trouble-shooting examples receives a warm welcome by the students. As Scott Lynn of UC Berkeley reports, “Our experience was that most students really got into the spirit of the thing and trouble shooting was one of the most popular parts of the course.” Perhaps some of the reasons why the development of trouble-shooting skill is not introduced are the need for excellent problem-solving skills, the lack of a variety of industrial problems and, perhaps most significantly, the student’s lack of a rich set of practical experience and understanding of equipment. There may also be a lack of the faculty’s confidence in using such open-ended experiences. Whatever the reason, I have designed this book to overcome these shortcomings. I hope that trouble-shooting skill development becomes part of every undergraduate experience. Training in trouble shooting in industry tends to occur from the school of hard knocks, by trial and error and gradually from the experience of solving problems as they occur, with no well-designed program of instruction. This is relatively inefficient and it does little to develop confidence. This book is designed to improve skill and confidence of process engineers and engineering students. This book is based on my experience developing trouble-shooting skills in undergraduate engineering programs, in short courses in industry and in courses presented at conferences. This book is designed to help develop your skill and confidence. This book is tailored to help you improve your skill no matter where you are in your journey to become an outstanding trouble shooter. A number of excellent books have been published about trouble shooting. Liberman (“Trouble-shooting process Operations”) describes a wide range of problems that he encountered, the fault that he discovered and the corrective action. His personal approach to trouble shooting is illustrated. Saletan (“Creative Trouble Shooting in the Chemical Process Industries”) provides interesting examples to illustrate different components in the trouble-shooting process. However, no specific educaSuccessful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

XIV

Preface

tional plan is apparent. No activities, with feedback, are provided for skill development. Branan’s “Rules of Thumb for Chemical Engineers” is mainly an excellent collection of rules of thumb. In addition he has chapters on trouble shooting and plant startup. He includes material from a range of topics and resources but he does not presented a synthesis of this material. The focus in these books tends to be a personalized approach of how the author solved trouble-shooting problems. Not everyone will or should follow Lieberman’s (1985), Saletan’s (1994), Gans’s (1983), Kister’s (1979) or my style in trouble shooting. The key is to identify your style and develop confidence in using it. Developing your style and skill requires that we draw on the extensive research about the trouble-shooting process. A skill development program should give you a chance to solve a wide range of trouble-shooting problems, to think about how you solved them and to set goals for improvement. The central core of this book is 52 trouble-shooting cases that are presented in a unique format that allows you to select the process you will use to solve the problem. Feedback is given to help you assess your approach. Target skills used by successful trouble shooters are given; structured activities provided, and feedback is supplied. This book is unique in its coverage, ease in use, focus on skill development using proven methods, self-selection and inclusion of activities that are challenging but fun to do. Included are a range of selfassessment tools. Here are the details: Chapter 1 outlines four types of trouble-shooting problems and summarizes the five key skill areas needed in trouble shooting: skill in problem solving, practical knowledge about a range of process equipment, specific knowledge about safety, hazards, systems thinking and people skills. A self-test is included to help identify which of the five key skill areas might be of most interest to you. Five trouble-shooting cases are posed from a variety of industries and unit operations: distillation, heat exchange, pumps, adsorption and crystallization and that pose a range of difficulty. The results of the self-tests can be used to guide you as to how best to use the remaining chapters and appendix material. The focus of this book is on developing skill in the mental process used to solve trouble-shooting problems. Chapter 2 summarizes the research evidence of what skilled trouble shooters do, provides a Trouble-Shooter’s Worksheet (and illustrates its application) and a feedback form to help focus attention on the problem-solving, synthesis, data-handling and decision-making activities used. This gives you a chance to compare the processes used with those used by skilled trouble shooters, and hence improve your skill and confidence in trouble shooting. To illustrate the application of these skills five scripts are provided in Chapter 4 of trouble shooters tackling the trouble-shooting problems Cases ’3–7. These are real problems taken from industrial experience; only the names of the trouble shooters have been changed. Each of the five scripts consists of about three parts with each part concluding with a few questions for you to consider. This reflective break was introduced to give you a chance to reflect on how you would have handled the case, and to decide what you should do next. I recommend that, as you read each script, you play the game. An assessment is given of the problem-solving processes used by

Preface

each of the trouble shooters. Other examples of the process are given in Appendix C and scattered as activities throughout most of the Chapters. Case ’3 in Chapter 6; Case ’6, in Chapter 6; Case ’8, in Chapters 2 and 6; Cases ’9 and 10, in Chapters 5 and 6; Case ’11 in Chapter 6; Cases ’12–18 in Chapter 7. The central activity of the book is in Chapter 8. Here, trouble-shooting problems are posed so as to help you develop your skill. The activity asks you to select an action or question to take for each selected case (from about 30 possible actions). A coded answer to each action is given in Appendix D. By posing a series of actions you will gather evidence until you have “solved the problem”. Feedback about the process is given in Appendix E where an answer is given and key elements of the process used by an experienced trouble shooter are listed. These problems are sequenced and classified so that you can start with easy and familiar Cases and build up your confidence gradually. The classification notes the degree of difficulty, the type of equipment involved, and the chemicals/process technology involved. Some of the Cases relate to similar processes. For example, six cases relate to the depropanizer-debutanizer system. Two, to the ethylene process; five, to the ammonia-reformer. Since the trouble-shooting cases require the use of the five key skills, the rest of the book provides skill-development activities for each of these five skills. Skill ’1: problem solving. The development of problem-solving skills is the theme of Chapters 5 and 6. In Chapter 5 the focus is on awareness, strategies, exploring the problem, creativity and self-assessment. Target skills are given, activities are introduced, a range of tasks are given (in Chapter 5 and Appendix F) and feedback is provided. Chapter 6 provides activities to develop skill in gathering data, checking hypotheses and critical thinking. For the various skills being developed, the process is illustrated (in the context of a trouble-shooting case), and tasks are given. This activity-based, workshop-style approach has been proven to be extremely effective; the proof is given in the award-winning paper” Developing problem-solving skills: the McMaster Problem Solving program, “Journal of Engineering Education”, April, vol 86, no 2, pp. 75–91, 1997. A wide range of tasks are provided from which you can select those most pertinent to your experience with feedback available in Appendix G. Skill ’2: knowledge of process equipment.

Chapter 3 provides a convenient summary of the practical aspects about equipment needed for trouble shooters of over 50 different types of process equipment. For most types of process equipment the following information is given: overall fundamentals, guidelines for good operation, and trouble shooting. For trouble shooting, typical symptoms are given together with a prioritized list of typical causes. Some will want to keep this text handy for just this summary of practical know-how. More details are given in Appendices A, H and I.

Skill ’3: process safety and properties of materials.

Chapter 3 also gives some succinct rules of thumb related to safety and hazard identification in Section 3.12.

Skill ’4: systems thinking.

Guides to and activities to help develop “Asystems thinking” are given in Chapter 3 in Sections 3.1 and 3.11 and Appendix B.

XV

XVI

Preface

Skill ’5: people skills.

Chapter 7 addresses interpersonal skills and looks at the factors influencing personal performance. More is given in Appendices C, F and G. Chapter 9 offers ideas of what to do next. This book would not have been possible without the help of many. In the seven companies for whom I worked before coming to McMaster University, I was fortunate to have worked with a variety of excellent trouble shooters who patiently helped me polish my skill, Don Ormston and Ted Tyler of Distiller’s Company Ltd, Saltend, UK; Stan Chodkiewicz, Polysar, Sarnia, J. Mike F. Drake, British Geon Ltd, Barry, South Wales. I thank Tom Marlin, Adam Warren, Iryna Bilovous, the late R.B. Anderson, Archie Hamielec, Terry Hoffman, Cam Crowe, John Vlachopoulos, Raja Ghosh, Douglas Dick, Dave Cowden and Lisa Crossley, McMaster University; Peter Silveston, University of Waterloo; Jud King and Scott Lynn, University of California, Berkeley; Ian Doig, University of New South Wales; Frank Bajc; Pierre Cote, Zenon Environmental, Douglas R. Winter and Robert French, Universal Gravo-plast, inc, Toronto, my students, my alumni who sent back problems (and answers), participants in the industrial workshops and in the conference workshops and Esso Chemicals, Nova Corporation, Prices of Bromborough, Unilever who generously provided me with problems and gave me permission to use them. McMaster Alumni who sent me Cases (and answers) include Bill Taylor (B Eng ’66), Ian Shaw (B. Eng. ’67), John Gates (B. Eng. ’68), Don Fox (B. Eng. ’73), R.J. Farrell (B. Eng. ’74), Jim Sweetman (B. Eng. ’77), Mike Dudzic (B Eng ’80), Mark Argentino (B. Eng. ’81), Vic Stanilawczik (B. Eng. ’83), Gary Mitchell (B. Eng. ’83), David Goad (B Eng and Mgt, ’91), Kyle Bouchard (B Eng ’93), Doug Coene (B. Eng. ’97) and Jonathan Yip (B. Eng. ’97). I have learned much from the cases solved and the approaches taken by Norman Lieberman, David Saletan and Henry Kister that they published in their books and articles. I thank Tom Marlin, and Brendan J. Hyland (B Eng and Society, ’97). With financial support from McMaster University Instructional Development program, they produced detailed versions of over 40 cases, some of which were used in this book. I am especially indebted to Luis J. Rodriguez, Downstream Oil Company, Waterdown; Douglas C. Pearson, Technical Support Consultant, Parry Sound and Tom Marlin, McMaster University, who gave me feedback and detailed suggestions on the case problems. Many colleagues supplied me with interesting trouble-shooting cases and information about the cause and perhaps some details about the TS process used to solve the problem. However, in writing up the interactive cases, I had to provide additional information to flesh out the case, provide some red herrings and address a broad range of possible hypotheses so that the fault is not immediately obvious. I have done my best, and any errors in this elaboration are mine. Waterdown, September 2005

Don Woods

1

1

What is Trouble Shooting? Process plants operate about 28 days of the month to cover costs. The remaining days in the month they operate to make a profit. If the process is down for five days, then the company cannot cover costs and no profit has been made. Engineers must quickly and successfully solve any troublesome problems that occur. Sometimes the problems occur during startup; sometimes, just after a maintenance turn-around; and sometimes unexpectedly during usual operation. A trouble-shooting (TS) problem is one where something occurs that is unexpected to such an extent that it is perceived that some corrective action may be needed. The trouble occurs somewhere in a system that consists of various pieces of interacting equipment run by people. The TS “corrective” action required may be: . . .

.

to initiate emergency shut-down procedures, to forget the situation; it will eventually correct itself, to return the situation to “safe-park” and identify and correct the cause and try to prevent a reoccurrence, to identify and correct the cause while the process continues to operate under current conditions.

Here are two example TS problems. Example Case ’1: “During the startup of the ammonia synthesis reactors, the inlet and outlet valves to the startup heater were opened. The pressure in the synthesis loop was equalized. The valves to the high-pressure stage of the synthesis gas compressor were opened and the firing on the start-up heater was increased. However, we experienced difficulty getting the fuel-gas pressure greater than 75 kPa; indeed a rumbling noise is heard if we try to increase the pressure. The process gas temperature is only 65 C. What do you do?” Example Case ’2: “The pipe on the exit line from our ammonia storage tank burst between the vessel and the valve. An uncontrolled jet of –33 C ammonia is streaming out onto the ground. What do you do?”

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

2

1 What is Trouble Shooting?

Trouble shooting is the process used to diagnose the fault safely and efficiently, decide on corrective action and prevent the fault from reoccurring. In this chapter we summarize the characteristics of a trouble-shooting problem, give an overview of the trouble-shooting process and “systems” thinking used to correct the fault and present an overview of this book.

1.1

Characteristics of a Trouble-Shooting Problem

TS problems share four common characteristics; TS problems differ in their seriousness and when they occur. Here are the details for each. 1.1.1

Similarities among TS Problems

TS problems share the following four characteristics: a) exhibit symptoms of deviations from the expected, b) have tight time constraints, c) are constrained by the physical plant layout and d) involve people. a)

Trouble-shooting situations present symptoms. The symptoms may suggest faults on the plant or they might be caused by trouble upstream or downstream. The symptoms may be false and misleading because they result from faulty instruments or incorrect sampling. The symptoms might not reflect the real problem. For example, in Example Case ’1 the cause is not that the fuel-gas pressure is too low. Instead, the suction pressure of the synthesis gas compressor was lower than normal, the alarms on the cold bypass “low flow” meter had been disarmed and the real problem was that there was insufficient process gas flow through the heater. b) The time constraints relate to safety and to economics. Is the symptom indicative of a potential explosion or leak of toxic gas? Should we initiate immediate shutdown and emergency procedures? The release of ammonia, in Example Case ’2, causes an immediate safety hazard. Time is also an economic constraint. Profit is lost for every minute when off-specification or no product is made. c) The process configuration constrains a trouble shooter. The process is fabricated in a given way. The valves, lines and instruments are in fixed locations. We may want to measure or sample, but no easy way is available. We have to work within the existing process system. d) Sometimes the cause of the problem is people. Someone may not have followed the expected procedure and was unwilling to admit error. Someone may have opened the bypass valve in the belief that “the process operates better that way.” As in Case ’1, the alarm may have been turned off. The orifice plate may have been put in backwards. Someone may have left his lunch in the line during the construction. Instructions may have been misinterpreted.

1.2 Characteristics of the Process Used to Solve Trouble-Shooting Problems

1.1.2

Differences between TS Problems

Here are four ways that TS problems differ. Some TS problems pose a) safety and health hazards. TS problems can arise b) during startup, c) after a shutdown for maintenance or after a change has been made and d) during usual operations.

1.2

Characteristics of the Process Used to Solve Trouble-Shooting Problems

The TS process or strategy used differs depending on the type of TS problem. Yet, the TS process has five common key elements. 1.2.1

How the Type of Problem Guides the TS Process or Strategy

The four different types of TS problems (described in Section 1.1.2) call for different TS strategies. .

Handling trouble that poses a hazard

At the design stage engineers should anticipate causes of potentially unsafe and dangerous operation (through such analyses as HAZOP and fault tree) and prevent hazardous conditions from ever occurring. They should include the four elements of control: the usual control, alarms, system interlock shutdown, SIS, and shutdown/relief. However, despite best efforts trouble can occur – such as in Example Case ’2. The TS strategy is to recognize unsafe conditions and initiate emergency measures or, where possible, to return the operation to “safe-park” conditions where operation is safe until the trouble is solved. .

Handling trouble during the startup of a new process

When we start up a process or new approach for the first time, we may encounter trouble-shooting problems. However, because these are “first-day” problems they have characteristics that differ from the usual trouble that can occur on an existing process. Hence, a different set of information or experience, and sometimes approach, can be useful. In particular, four events could cause trouble: 1. 2. 3. 4.

garbage or stuff left in the lines or equipment, incorrect installation, for example, a pipe hooked up to the wrong vessel, during startup, there are often many people around to get things going correctly – this can interfere with the lines of communication, residual water or air left in process vessels and lines.

3

4

1 What is Trouble Shooting?

Furthermore, although we have theory and often computer simulations to provide ideas about how the plant or process should be operating; we have no actual data. Example Case ’1 is a startup problem. The TS strategy is to focus on the basic underlying principles and create hypotheses about how the process and operations should function. The financial penalty is usually higher for delays during startup. The penalties include penalties written into the contract for delays, insurance costs and government regulation costs. .

Handling trouble that occurs after a maintenance turnaround or a change.

Changes that can cause faulty operation include 1. 2. 3.

equipment is taken apart for maintenance, processing conditions change because, for example, the feedstock is changed, there is a change in operating personnel.

In these examples, we have information about performance before and after the change. The TS strategy is to identify the change that seems to have triggered the fault. .

Handling trouble that occurs during usual operation or when conditions change gradually.

Sometimes we encounter trouble when the process is operating “normally” or when we gradually increase the production rate. The TS strategy is to focus on the basic underlying fundamentals of how the process works, create hypotheses that are consistent with the evidence and use tests to confirm the hypothesis. 1.2.2

Five Key Elements Common to the TS Process

Skill in trouble shooting depends on five key elements: 1) skill in problem solving, 2) knowledge about a range of process equipment, 3) knowledge about the properties, safety and unique characteristics of the specific chemicals and process conditions where the trouble occurs, 4) “system” thinking and 5) people skills. Here are some details about each. For general problem solving, one of the most important skills is in identifying which evidence is significant and how the evidence relates to appropriate hypotheses and conclusions. Concerning the importance of knowledge about process equipment, the differences between skilled and unskilled trouble shooters are more in their repertory of their experiences than in differences in general problem-solving skills. In other words, it is the knowledge about process equipment, common faults, typical symptoms and their frequency that is of vital importance. A trouble-shooter’s effectiveness depends primarily on the quality of the knowledge that relates i) symptom to cause; and ii) the relative frequencies of the symptoms and the likelihood of causes.

1.3 Self-Test and Reflections

Specific knowledge about the chemicals and equipment configuration must be known to handle safety and emergencies. For example, if knowledge of the hazards of ammonia is not known, then Example Case ’2 is not treated with the urgency required. Trouble occurs in a process “system” even though it might initially appear as though it is in an isolated piece of equipment. Equipment interacts; people interact with the equipment. Viewing the trouble-shooting problem in the context of a “system” is vital. Interpersonal skills are needed. The interpersonal skills needed between the trouble shooter and the people with whom he/she must interact include good communication and listening skills, building and maintaining trust and understanding how biases, prejudice, and preferences lead to interpersonal differences in style.

1.3

Self-Test and Reflections

Reflect on your trouble-shooting skills based on the five common key elements described in Section 1.2.2. Rate yourself-on the five or six elements in each category and then set goals to improve. A rating of 0 means that nothing is known. The maximum scale is 10. Descriptions are given for ratings of 1, 5 and 10. (1) Problem-solving skill as applied to trouble shooting –

Monitoring, being organized and focusing on accuracy:

rate: _____

1 = aware that it’s important when problem solving. 5 = monitor about once per 5 minutes, use a personal “strategy”, tend to let time pressures dominate. 10 = monitor about once per minute, use an evidence-based strategy flexibly and effectively, focus on accuracy, check and double check frequently. –

Data handling, collecting, evaluating and drawing conclusions:

rate: _____

1 = think of a variety of data to be collected. 5 = systematically collect data that seem to test the hypotheses, unclear of accuracy of data, unaware of common faults in reasoning, emphasis on opinions. 10 = systematically decide on data to collect and correctly identifies its usefulness; aware of the errors in measurements; use valid reasoning, focus on facts, aware of own biases in collecting data. –

Synthesis: creating and working with hypotheses as to the cause:

rate: _____

1 = aware that should have a hypothesis. 5 = can identify several working hypotheses that seem technically reasonable. 10 = can generate 5 to 7 technically reasonable hypotheses for any situation; willing to change hypotheses in the light of new data.

5

6

1 What is Trouble Shooting?

–

Decision making:

rate: _____

1 = use intuitive criteria. 5 = systematic, consider many options, unaware of any biases. 10 = use measurable must and want criteria explicitly, prioritize decisions and aware of personal biases and try to overcome these. (2) Experience with process equipment –

Centrifugal pumps:

rate: _____

1 = flow capacity and head, location of inlet and exit, principle of operation. 5 = NPSH and problems related to this, impact of reverse leads on the motor, correct location of the pressure gauge on the exit and the implications of shutting the exit valve, pumps operate on the head-capacity curve and the implications. 10 = implications of worn volute tongue and worn wear rings, lubrication, seals and glands. –

Shell and tube heat exchangers:

rate: _____

1 = size area. 5 = size on area and Dp, baffle-window orientation, correction to MTD for multipass system, some options for control. 10 = tube vibrations, steam traps, nucleate versus film boiling and conditions, different causes of fouling, maldistribution issues and can use a variety of control options. –

Distillation columns:

rate: _____

1 = estimate the number of trays, know impact of feed conditions, reflux ratio and bottoms and overhead composition. 5 = familiar with a variety of internals and can size/select, size downcomers, issues related to sealing downcomers, familiar with some control options, can describe the interaction between condenser and reboiler. 10 = jet versus downcomer flooding, surface tension positive vs negative, pump arounds, vapor recompression, wide variety of control options. (3) Knowledge about safety and properties of material on the processes with which I work I can list the conditions and species that pose: –

Flammable risk

rate: _____

1 = can identify individual species and conditions for five chemicals that might produce “flammable risk”. 5 = can identify individual and combinations of species and conditions for over 30 chemicals and the process faults or failures that might produce “flammable risk”. 10 = can identify individual and combinations of species and conditions for over 100 chemicals and the process faults or failures that might produce “flammable risk”. –

Health risk

rate: _____

1 = can identify individual species and conditions for five chemicals that might produce “health risk”. 5 = can identify individual and combinations of

1.3 Self-Test and Reflections

species and conditions for over 30 chemicals and the process faults or failures that might produce “health risk”. 10 = can identify individual and combinations of species and conditions for over 100 chemicals and the process faults or failures that might produce “health risk”. –

Explosive risk

rate: _____

1 = can identify individual species and conditions that might produce “explosive risk” for five chemicals. 5 = for over 30 chemicals and can identify individual and combinations of species and conditions and the process faults or failures that might produce “explosive risk”. 10 = for over 100 chemicals and can identify individual and combinations of species and conditions and the process faults or failures that might produce “explosive risk”. –

Mechanical risk

rate: _____

1 = can identify pressure and moving equipment risk for about five types of equipment. 5 = can identify overpressure, thermal and moving equipment risk for a P&ID with 20 pieces of Main Plant Items, MPI. 10 = can identify overpressure, thermal, corrosive and moving equipment risk for a P&ID with 50 MPI. –

Unique physical and thermal properties:

rate: _____

1 = can identify chemicals and conditions that have “unique properties” for one chemical. 5 = for 10 chemicals. 10 = for 30 chemicals. (4) “Systems” thinking. –

Faulty operation of and carryover from/to upstream/downstream equipment: rate: _____

Can estimate/predict the effects of pulses, cycling, contamination on downstream equipment. Can predict potential sources of pulses, cycling and contamination from upstream equipment. 1 = for one piece of equipment. 5 = for a P&ID with 10 MPI. 10 = for a P&ID with 40 MPI. –

Impact of environmental conditions

rate: _____

1 = can estimate the environmental impact for the atmosphere from about 10 main plant items. 5 = for about 20 MPI and atmospheric, aqueous and solid impact. 10 = for about 50 MPI and atmospheric, aqueous and solid impact. –

Pressure profile:

rate: _____

1 = can calculate a pressure profile for one pipe from detailed calculations. 5 = can use rules of thumb to estimate the pressure profile for about five piping configurations. 10 = can estimate pressure profiles for a P&ID with interconnecting piping with 50 MPI.

7

8

1 What is Trouble Shooting?

–

Process control:

rate: _____

0 = Unable to identify and rationalize a process control system. 5 = For a P&ID with 10 MPI, can identify good and bad process control; can identify the presence and absence of four levels of process control (control, alarm, SIS, relief and shutdown). 10 = For a P&ID with 40 pieces of equipment, can identify good and bad process control; can identify on the P&ID the presence of and absence of four levels of process control (control, alarm, SIS, relief and shutdown). (5) People skills –

Communication skills:

rate: _____

1 = write or speak to tell them what you know, use acceptable grammar and follow expected format. 5 = correctly identifies single audience, answers needs and questions; includes some evidence related to conclusions, reasonably well organized with summary, coherent and interesting, defines jargon or unfamiliar words, grammatically correct and follows the expected format and style. Some misunderstanding occurs in some verbal or written instructions. 10 = correctly identify multiple audiences, answer their needs and questions; include evidence to support conclusions, well organized with summary and advanced organizers, coherent and interesting, defines jargon or unfamiliar words, grammatically correct and follows the expected format and style. Verbal and written instructions are carried out correctly. –

Listening skills:

rate: _____

1 = listen intuitively. 5 = aware of some elements of listening and usually can demonstrate attending. 10 = aware of the characteristics and foibles of listening, skilled at opening conversations, attending, following and reflecting. –

Fundamentals of relationships:

rate: _____

1 = handles relationships intuitively. 5 = aware of most of the fundamentals and unacceptable behavior. 10 = claims and respects fundamental rights and avoids using contempt, criticism, withdrawal and defensiveness. –

Developing and building trust:

rate: _____

1 = knows a few principles for developing trust; 5 = understands how to develop trust. 10 = can develop mutual trust naturally. –

Building on another’s personal preferences:

rate: _____

1 = intuitively aware of own preferences and that others are different. 5 = explicitly aware of own preferred style and aware of uniqueness of others but not very effective in exploiting the differences positively. 10 = familiar with my uniqueness and those of my colleagues and use the differences to improve our work instead of promoting conflict.

1.6 Cases to Consider

Total your scores. Identify the areas with the lowest scores and set goals for yourself. For problem solving, see Chapters 2, 5 and 6. For experience with process equipment, see Chapter 3 and Appendix A. For knowledge about safety, see Chapter 3. For “systems thinking”, see Chapter 3 and Appendix B. For people skills, see Chapter 7. If you have high scores in all areas, Congratulations. Go directly to Chapter 8 and enjoy!

1.4

Overview of the Book

This book is about improving your approach to trouble shooting. This book has basically five parts. Chapters 2 and 3 provide details about the mental process and practical knowledge of common symptoms and causes for a variety of process equipment. Chapter 4 gives some examples of trouble shooters in action as they work through a variety of problems. This is included to give you a chance to reflect on your approach. Chapters 5, 6 and 7 provide example training opportunities to polish your skill in trouble shooting in the areas of problem solving, critical thinking and testing hypotheses and interpersonal skills, respectively. Chapter 8 gives cases that you, the reader, can use to polish your skill. The final chapter suggests the next level of considerations to polish your skill further.

1.5

Summary

Trouble-shooting situations present symptoms, symptoms that may not reflect the real problem. Trouble shooters are constrained by time and the existing equipment layout. Trouble-shooting situations inevitably include people. Solving a trouble-shooting problem uses the five elements: skill in problem solving, knowledge about equipment and about hazards, skill in systems thinking and people skills. Problems occur that pose a hazard, when the process is started up for the first time, when the process is started up after change or maintenance or during usual operations or when we are trying to increase the capacity of the process. Slightly different TS strategies are used for the different types of TS problem.

1.6

Cases to Consider

Here are five cases. Consider each and write out the approach you would take to start each. For example, you might ask What is the problem? What questions might I ask? What are the possible causes? What tests might I do? What samples might be taken for analysis?

9

10

1 What is Trouble Shooting?

Case ’3: The Case of the cycling column The shut-down and annual maintenance on the iC4 column has just been completed. When the operators begin to bring the column back on-stream the level in the bottom of the column cycles madly, that is, the level rises slowly about 0.6 m above the normal operating level and then quickly drops to about 0.6 m below normal. The process then repeats. You have been called in as chief trouble shooter to correct this fault. It costs our company about $500/h when this plant is off-stream. Get this column working satisfactorily. The system is given in Figure 1-1.

Condensate

Steam Temp 1

Column

Trap

Figure 1-1

A distillation column for Case ’3.

The case of the platformer fires Heavy naphtha is converted into high octane gasoline in “Platforming”. Byproducts of the reaction include low-pressure gas and hydrogen-rich gas containing 60–80% hydrogen. The products from the platformer reactor (at 4.8 MPa g and 500 C) are heat exchanged with the feed naphtha to preheat the reactor feed. Figure 1-2 illustrates the layout. In the past three weeks since startup we have had four flash fires along the flanges of the stainless steel, shell and tube heat exchanger. The plant manager claims that because of the differential thermal expansion within the heat exchanger, because of the diameter of the exchanger (1 m), and because it’s hydrogen, we’re bound to have these flash fires. The board of directors and the factory manager, however, refuse to risk losing the $90 million plant. Although the loss in downtime is $10,000/h, they will not let the plant run under this flash-fire hazard condition. “Fix it!” says the technical manager. Maintenance have already broken six bolts trying to get the flange tighter, but they just can’t get the flanges tight enough.

Case ’4:

1.6 Cases to Consider

Platformer

Naphtha feed

Figure 1-2

The platformer for Case ’4.

The sulfuric acid pump problem Dilute sulfuric acid is stored in a horizontal, cylindrical tank in the basement, as is shown in Figure 1-3. The tank diameter is 1.8 m; the length, 3.6 m. An exit line goes from the bottom of the tank and rises 3.6 m up through the ground floor to a centrifugal transfer pump that pumps the acid to a reservoir 7.5 m above the ground level.

Case ’5:

to reservoir at elevation +7.5 m ground level

vent 3.6 m

acid return lines

site guage 1.8 m acid storage tank

basement

Figure 1-3

The configuration for Case ’5.

11

12

1 What is Trouble Shooting?

Acid is recycled to the tank from different parts of the process at such a rate that about every two hours the pump is activated to transfer the acid to the elevated reservoir. However, each time the transfer pump operates, the level gauge at the side of the tank shows that there is still about 0.7 m of acid in the bottom of the tank when the transfer pump makes a “crackling” noise that the operator says “sounds like cavitation”. At this time the operator stops the pump. This means that the transfer pump has to operate more frequently than need be and that cavitation may be eroding the impeller. What do you suggest that I do to fix the problem? The case of the utility dryer (courtesy of C.J. King, University of California, Berkeley) Our plant has a utility air drying unit, which dries all of the utility air used for the pneumatic instrument lines and other purposes. The air is compressed to about 550 kPa g and then passes through the drying unit, the flow diagram of which is attached. For this unit, two beds are hooked in series with the first bed being regenerated and the second bed drying. The first bed experiences two hours of regeneration with hot air followed by a one hour flow of cold incoming air to cool down the regenerated bed. The second bed drys the air for 3 hours. After 3 hours, the flows switch so that the regenerated bed becomes the drying bed and vice versa. The plant operators are following the vendor’s instructions in setting the timer dials on the various valves: all the valves (the four way valves, V2 and V3, and the three-way valve, V1) are thrown every three hours. The 3-way valve, V1, is also thrown two hours after a cycle change to send fresh air to cool the regenerated bed. The hot air used to regenerate the bed is heated in a steam heater with the TRC-1 set at 175 C. The present utility air flowrate through the dryer is 4000 Ndm3/s or about 12 design flowrate. The proportionating valve is governed by pressure P3. At present, full pressure is kept on the valve; the valve is shut so that no air goes directly to the dryer bed. The diagram shows the valve settings for Bed A being regenerated and Bed B, drying. All the air flow goes, via the 3-way valve V1 to the steam heater for the regeneration phase of Bed A. The adsorbent is activated alumina with typically 0.14–0.22 kg water adsorbed/kg dry solid. Each bed contains 5000 kg of activated alumina. The available sample valves are labeled “S”. Now that it is winter we have been experiencing much colder nights, and we have encountered several instances where the instrument air lines have been freezing. This has been traced to the air coming out of the drying unit being too wet, on average. We estimate that this problem will cost us about $8,000 per day until we get it fixed. The job is yours – fix it. Case ’6:

1.6 Cases to Consider

Figure 1-4

The utility dryer for Case ’6.

The case of the reluctant crystallizer (the case is supplied by W.K. Taylor, B Eng. McMaster, 1966 and used with permission) Process solution, at 55 C, enters the vacuum crystallizer (VC) where it is concentrated and cooled to cause precipitation of the product. Normally, the first and second stage ejectors are used to start syphoning feed solution into the VC until it is two-thirds to three-quarters full. The first hour of operation is done at 6.5 kPa absolute supplied by the first- and second-stage ejectors with city water to the interstage condenser. When the batch cools to 40 C the booster ejector and the barometric condenser are turned on to give an absolute pressure of 2.5 kPa abs. The batch time is 8 hours during this time the liquid level in the VC slowly drops about 40 to 50 cm. The city water is much colder than the bay water and so to ensure that the temperatures in the barometric leg is less than 26 C, city water can be used to supplement the bay water. If the booster ejector is turned on too soon, it will not hold but rather kicks out. This happens when the steam goes directly into the VC instead of through the ejector nozzle. This phenomena makes a recognizable sound. Today, the plant operator phones, “The booster does not hold! After about half to one hour of operation it kicks out.”

Case ’7:

13

14

1 What is Trouble Shooting?

While the booster was holding, the liquid level dropped at such a “fantastic rate” that you could actually watch the level drop, whereas it would normally drop 40 to 50 cm over an 8-hour period. Pressure gauge P7 indicated a “wildly fluctuating pressure”. The needle jumped back and forth from 140 to 550 kPa g while the booster was “holding”. All the other pressures and temperatures were normal. Here is a summary: Pressure, kPa g Steam

Normal readings Today

Water

P1

P2

P3

P4

P8

P5

P6

P7

550 550

550 550

550 550

725 725

550 550

0–35 0–35

310 310

205 140–550

Temperature, C Barometric legs

Normal readings Today

T1

T2

< 27 < 27

< 27 < 27

Figure 1-5 illustrates the system.

1.6 Cases to Consider

Figure 1-5

The vacuum crystallizer for Case ’7.

Feedback for these cases is provided in Chapter 4.

15

17

2

The Mental Problem-Solving Process used in Trouble Shooting A trouble-shooting problem occurs. As the trouble shooter processes the evidence, he/she mentally scans past experience to see if he/she has successfully solved anything like it before. If that past experience is limited; if there are no past examples that bear any relationship to the current troubling problem, then this is a “problem to solve”. The process used will be called “problem solving.” Problem solving is illustrated in Figure 2-1. Data are gathered and an internal, mental representation is created of the problem situation. That mental representation is compared with past experience to see if a problem similar to this has been solved successfully in the past. If not, then “it’s a problem!”. We systematically draw on our problem-solving skills, scan our bank of pertinent knowledge and combine our skills and knowledge to “solve the problem”. We then elaborate and take time to encode and store that successful solution to the problem in our mental experience bank. In the future, when we encounter a similar problem we don’t have to agonize through all of the problem-solving process. We simply recall a solution from our experience in a process we call “exercise solving”. “Exercise solving” is illustrated in Figure 2-2. Data are gathered and an internal, mental representation is created of the problem situation. That mental representation is compared with past experience to see if a problem similar to this has been solved successfully in the past. If yes, then “it’s an exercise!”. The past solution is recalled from experience and modified to solve the current problem. Beginning trouble shooters with limited experience start their journey as problem solvers and gradually build up experience. Experienced trouble shooters are primarily exercise solvers who draw on their knowledge and experience. Research evidence suggests that for experienced trouble shooters 95% of the situations they encounter will be “exercises”. They still need problem-solving skill for 5% of the situations. In Section 2.1 we summarize research about problem solving, in general. These characteristics of skilled problem solvers are used in solving any type of problem such as setting goals, making decisions, making a purchase and trouble shooting. In Section 2.2 additional research evidence into the process of solving troubleshooting problems is given. In Section 2.3 is given a worksheet or template for solving trouble-shooting problems. Also given is an assessment form to provide feedback about one’s performance as a trouble shooter. An example use of the Worksheet is given in Section 2.4 for Case ’8. Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

18

2 The Mental Problem-Solving Process used in Trouble Shooting

Figure 2-1

Problem solving.

2.1 Problem Solving

Figure 2-2

Exercise solving.

2.1

Problem Solving

Here are eighteen characteristics of skilled problem solvers. The first eight could be called “the problem-solving process or how”; the second set of characteristics are called “synthesis”. The third class is called “decision making”. The other important characteristics, “data and analysis”, are given in Section 2.2.3. Research has shown that attitudes and other related skills are also important.

19

20

2 The Mental Problem-Solving Process used in Trouble Shooting

The “problem-solving process, how”: 1. 2.

3. 4. 5.

6.

7. 8.

Be able to describe your thought processes as you solve problems. Be organized and systematic. An evidence-based strategy for solving problems, in general, is given in Figure 2-3. The six stages are as follows. 1. Engage with the problem or dilemma, listen, read carefully and manage your distress well. Say “I want to and I can!” 2. Analyze the data available and classify it: the “goal, the givens, the system, the constraints and the criteria”. 3. Explore: build up a rich visual/mental picture of the problem and its environment; through simplifying assumptions explore the problem to see what is really important; identify the real problem. 4. Plan your approach to solving the problem. 5. Carry out the plan and 6. Check the accuracy and pertinence of your answer. Did it answer the problem? satisfy the criteria? Reflect on the problem-solving process used to discover new insights about problem solving. Elaborate on the answer and the problem situation to discover answers to other problems, to extend the solution to other situations and relate this problem experience to other technical problems you have solved in the past. Cue this experience into memory. This systematic approach is not sequential. Skilled problem solvers bounce back and forth between the stages. A typical approach would be engage, analyze, engage, explore, engage, explore, analyze, engage, explore, plan, engage and so on. Focus on accuracy instead of speed. Actively write things down. Make charts, draw diagrams, write down goals, list measurable criteria and record ideas from brainstorming. Monitor and reflect. Mentally keep track of the problem-solving process and monitor about once per minute. Typical monitoring thoughts are “Have I finished this stage? What have I discovered so far? Why am I doing this: if I calculate this, what will this tell me? What do I do next? What seems to be the problem? Is this the real problem? Should I recheck the criteria?” Typical reflections that look back on the process and attitudes used are: “This didn’t work, so what have I learned? Am I focusing on accuracy or am I letting the time pressures push me to make mistakes? Am I managing my stress? I can do this! Am I monitoring the process? “ Explore the “real” problem by creating a rich perspective of the problem. During the explore stage, see it from many different points of view. Be willing to spend at least half the total available time defining the problem. Ask many what if questions. Try to bound the problem space. “Swim with the data” to see how it responds. Identify the real problem, by asking a series of Why? questions to generalize the situation and to see the problem in the context of a “system”. This activity of identifying the real problem was called the Explore stage and is the heart of the problem-solving process. Identify the subcomponents of the problem, yet keep the problem in perspective. Are skilled at creative and critical thinking.

2.1 Problem Solving

Figure 2-3

The MPS Strategy for problem solving.

These first eight items we could call “Problem-solving process, how”. Table 2-1 lists detracting and enriching behaviors. Activities to help develop these skills are given in Chapter 5. The next two items are related to “Synthesis” with detracting and enriching behaviors listed in Table 2-1.

21

22

2 The Mental Problem-Solving Process used in Trouble Shooting

9. Be flexible. 10. Keep at least five hypotheses active. Do not quickly close on one hypothesis. Issues related to “decision making” are next. 11. Spend time where it benefits you the most. Use Pareto’s principle (80% of the results can be found from 20% of the effort). Find the key 20%. 12. Be an effective decision maker. Express the goal as results to be achieved rather than as actions to be taken. Make decisions based on criteria that are explicit and measurable. Distinguish between must criteria (the process must have an internal rate of return of 35%) and want criteria (the process might have the potential to be licensed). Reject options that do not meet the must criteria. Use a rating system to score the want criteria. The remaining research evidence relates to “attitude toward problem solving” and some related skills. 13. See challenges and failure as opportunities for new perspectives. 14. Be willing to risk. 15. Manage stress well. Solving problems is stressful. When we initially encounter a problem we experience distress because of the uncertainty. Such stress tends to immobilize us. When we successfully solve a problem we experience the joy and exhilaration of stress (that distracts us from checking and double checking that our answer is the best). A certain level of stress motivates us. Excessive stress makes us make mistakes. Data suggest that operators with confidence and training working under high stress make 1 mistake in 10 actions. Operators with confidence and training who receive feedback about their actions and are under low stress make 1 mistake in 1000 actions. Although these data refer to plant operators, the same trends can be extended to suggest how stress, lack of reflection and feedback might interfere with engineering practice. High stress would be a rating of over 450 on the Holmes–Rahe scale (Holmes and Rahe, 1967). Ten suggested approaches to managing stress include: worry only about things over which you have control, include physical exercise as part of your routine, have hobbies and destimulating activities in which you can lose yourself, plan ahead, avoid negative self-talk, rename the events that are stressful to you, build a support system, be decisive, put the situation into perspective and use role models of others who have succeeded. 16. Manage your time well. Covey (1990) offers excellent suggestions on time management. Identify problems and decisions according to their importance and urgency. Shift the important situations to being non-urgent. Learn to say “No”. 17. Understand your strengths, limitations and preferred style. See Section 6.1.3, part c. 18. For problems involving people, use the 85/15 rule. 85% of the problems occur because of rules and regulations; 15% of the problems are because of people.

2.2 Trouble Shooting

2.2

Trouble Shooting

The general problem-solving characteristics listed in Section 2.1 apply to troubleshooting problems. However, unique to this type of problem are other research findings. Here we summarize the use of the strategy, in Section 2.2.1, general considerations, in Section 2.2.2 and hypothesis testing in Section 2.2.3. 2.2.1

Considerations when Applying the Strategy to Solve Trouble-Shooting Problems

Two different types of ideas help us focus on our use of the general strategy: problems where there is an apparent change and where there is no apparent change. a. Using this Strategy for “Change” Problems The overall strategy, described in Section 2.1.1, is applied to identify the change that occurred to cause to trouble. The hypothesis is that the symptoms arise because of some change made to the system. Therefore the plan is to identify the change. The basis of the approach is to learn to ask the right questions. Kepner–Tregoe (1985) illustrate the application of this approach. The questions that usually are most helpful are those that help identify an obvious change. .

. . .

what is happening and what should be happening but is not, and “is this difference significant?” where is and where is not, who is and who is not, when is and when is not.

This approach is usually most helpful for “people problems”, for problems that occur just after maintenance and for processes that have worked well in the past and now seem to be malfunctioning after a change has been made – in raw material, in operating procedure, in operators, in weather or in season. b. Using this Strategy for “Basics” Problems The same overall strategy, described in Section 2.1.1, is applied when a change is not obvious. The emphasis is different in that we focus on the basics instead of on a change. The conditions when this apply could be because 1) we are starting up a process for the first time, and we have no practical data of what should be happening or 2) something internal to the process changes and we have no simple way to identify that change. No one ordered the raw materials from a new supplier. No one repaired a pump. No one changed the temperature setting on the heater. Instead, inside the equipment a hunk of corroded metal fell into the liquid; or a truss weakened and gave way inside the vessel. The catalyst bed collapsed. We cannot easily identify the change because there is nothing to “see” from the outside. For this situation, we rely on our fundamental principles and knowledge of how the process and equipment should operate, we create hypotheses, check for consistency between the

23

24

2 The Mental Problem-Solving Process used in Trouble Shooting

hypotheses and the evidence and test the hypothesis. The questions that tend to be successful are as follows. . . .

.

What basically is going on in this process? What fundamentals are important? What are the key operating principles that guide the operation of the equipment? How are the fundamentals reflected in the observed data?

2.2.2

Problem-Solving Processes Used by Skilled Trouble Shooters

Here are the characteristics of skilled trouble shooters especially when we are trying to diagnose the cause when using a “basics” strategy. 1. 2. 3. 4. 5.

6. 7. 8. 9.

Generate hypotheses early based on limited cues. Consider the most common hypothesis first. Be systematic and organized: each piece of information requested should relate to an organized plan of attack. Make hypotheses consistent with the evidence: no hypothesis should be more specific or more general than the evidence justifies. Keep two to five competing hypotheses under consideration at any one time. Explore the option of multiple causes especially when evidence suggests a single rare cause. Do not neglect the possibility of two common causes. If two or more common causes would produce disastrous results, and you cannot confirm or refute these causes; act as if both are the cause. Whenever a new or revised hypothesis is generated, check the implications of previous cues. Prioritize test procedures; use the simple, inexpensive ones first before exploring the high-cost option. Do not guess. Use a systematic TS process. Find the root cause; do not correct the symptoms.

Some example data (from the Medical literature, Elstein et al., 1978) are: . . . . . .

total different bits of information sought/given: 200 per case, total bits accumulated before the first hypothesis was generated: 20, number of active hypotheses: 3, total number of different hypotheses throughout whole case: 6, number of cues acquired: 50 number of critical cues acquired: 50

Table 2-1 describes detracting and enriching behaviors for these activities under the topic synthesis.

2.3 Overall Summary of Major Skills and a Worksheet

2.2.3

Data Collection and Analysis: Approaches Used to Test Hypotheses

Successful trouble shooters .

. .

.

.

.

.

.

assign a weighting of +++, ++, +, 0, –, – –, – – – to the cues/evidence and then select the hypothesis that maximizes the positive cues or that has the maximum difference between positive and negative cues. use Bayes’ theorem if the probabilities of various causes are known. are sensitive to and try to overcome personal biases (related to premature closure and anchoring). consider the evidence with respect to all hypotheses (to overcome the most commonly encountered bias of pseudodiagnosticity or overinterpretation). gather data to disconfirm a hypothesis and are willing to discard a “favored” hypothesis (to overcome confirmation bias). consistently use fundamentals when analyzing the evidence-hypothesis link (to overcome representativeness bias). use diagrams, trees and tables to systematically chart hypotheses, cause and evidence (to overcome omitting cues and overcome the limitations of ShortTerm Memory). restrain from creating a new hypothesis for each new clue and thereby generate excessive data and have trouble with closure.

Table 2-1 lists detracting and enriching behaviors for these activities under the two topics of data analysis and decision making. Ideas on how to improve this skill are given in Chapters 5 and 6.

2.3

Overall Summary of Major Skills and a Worksheet

The research summarized in Sections 2.1 and 2.2 can be converted into a TroubleShooter’s Worksheet to guide our approach and an assessment form, to give feedback for growth. These are discussed in turn. 2.3.1

Getting Organized: the Use of a Trouble-Shooter’s Worksheet

To help us to be systematic, we use the McMaster 6-step strategy in Figure 2-3 and convert this into a worksheet. A succinct version highlighting the key features is given in Trouble shooters Worksheet 2-1.

25

26

2 The Mental Problem-Solving Process used in Trouble Shooting

Trouble-Shooter’s Worksheet 2-1: Succinct summary ( copyright 2003 D. R. Woods and T. E. Marlin)

1. Engage: Gather initial information. .

. . .

Establish if emergency priority: safety? damage? shut down or safe-park or continue? Describe what’s going on. Manage panic: “I want to and I can.” Monitor: Have you finished this stage? Can you check? What next?

2. Define the stated problem: based on given information. If the information is not known at the stage, gather it later.

WHAT WHEN WHERE WHO

IS

IS NOT

_____________________ _____________________ ?& _________________ ?& _________________ ?& _________________

(should be happening but it is not) _________________________________ ?& ______________________________ ?& ______________________________ ?& ______________________________

Identify situation as 1) startup new process; 2) startup after maintenance or change, 3) usual operation. Monitor: Have you finished this stage? Can you check? What next? 3. Explore: Exercise? or problem? Strategy for change or basics? Useful to broaden with Why? Why? Why? Gather information. Perspectives: customers? suppliers? weather? changed economics? politics? environment? . . . . .

. .

Prioritize: product quality, production rate or profit? Goal: safe-park? short term? long term? SMARTS$ Data consistency? Pertinent fundamentals? Likelihood of problem type. Explore with What if? List changes made and/or list trouble-shooting experience: root causes based on symptom. (Chapter 3) Brainstorm hypotheses Hypotheses and evidence of symptoms:

Evidence of symptoms: a. ________________ b. ________________________ c. ___________________ d. ___________________ e. ___________________

2.3 Overall Summary of Major Skills and a Worksheet

Working Hypotheses

Initial Evidence a

b

c

Diagnostic Actions d

e

A

B

C

D

1 2 3 4 5

S = supports; D = disproves, N = neutral

Diagnostic actions: A. ______________________ B. _____________________ C. __________________________ D. _________________________________ 4. Plan 5. Do it 6. Look back

1. Engage: Take in the evidence. Listen carefully to the phone call. Sense the evidence. Establish priority: Safety? Hazard containment? Equipment damage? If yes, then invoke emergency measures or alter conditions to safe-park the process in the interim. For example, a distillation column might be “parked” by isolating it and continuing to operate under full reflux. If emergency or safe-park options are not needed, continue. Take time to really understand the physical process. We find it useful to write down a word description of what happens in the process. If a diagram is available, trace around the lines and describe what is flowing in each line, how it is controlled and what should be happening. Once you feel that you have some understanding of the process, manage your distress by saying “I want to and I can. I have a strategy that works. Let’s systematically follow it.” 2. Define the stated problem Understand the given information. Classify and chart the information using IS and IS NOT for WHAT, WHERE, WHEN and WHO. List the symptoms as given. Don’t guess! Don’t overinterpret! Don’t infer! Just systematically classify the given information. Some may not be known, so identify this as information to be gathered. Note whether this is startup of a new plant, startup after maintenance or change or usual operations. If this is not known, then identify this as information to be gathered. 3. Explore: Build a rich description of the situation. Gather information to be gathered noted in Define. See the situation from many different perspectives. Decide if this is an exercise or a problem. Decide if a more effective strategy might be to focus on change or basics.

27

28

2 The Mental Problem-Solving Process used in Trouble Shooting

Gain a rich perspective about the situation by considering the problem situation from the following points of view: a) viewpoint of fundamentals and practical operability, b) viewpoint of trouble-shooting rules of thumb: likelihood specific to type of situation or conditions, example startup (from Section 3.1), likelihood specific to equipment (Chapter 3 plus experience) c) viewpoint of controlling factors via What if? plus order-of-magnitude estimates to bound the behavior and identify key assumptions and variables, d) viewpoint of isolated equipment, equipment in the context of a subsystem, in the context of a mini-system, in the context of the plant site including utilities and in the context of the corporation; e) viewpoint in context of the weather, political, environmental, legal and economic environment; f) viewpoint of trends and time changes, g) perhaps broaden the context of the situation by asking Why? Why? Why? h) viewpoint of stakeholders: customers, plant manager, operators, vendors, process control, auditors, statisticians, instrument and control specialists, or unit operation specialist. Then select the “real” goal. Decide priority based on product quality, production rate and profit. Describe the goal and criteria with the aid of the acronym SMARTS$, introduced by colleague Tom Marlin. The Goal should describe results – not actions – and be expressed in Specific and Measurable terms. The Goal should be Attainable. The Goal should produce a Reliable and stable result. Use as Criteria: Timely: the problem should be solved quickly. Promptness is critical for emergency priorities in safety, containment of hazards and damage to equipment. This was considered in the Engage stage. The Goal should produce a Safe operation, safe product, safe startup and shutdown. Use as Criteria: $ The downtime costs, the testing costs and the corrective and preventative costs should be minimized. Chart the given evidence and symptoms. Brainstorm hypotheses as to the root cause. Defer judgement. Don’t be afraid to list crazy ideas and to spend an extra five minutes dreaming up and building on wild ideas. Remember, often the unique ideas are found in the midst of completely useless ideas. List five to seven working hypotheses as to the cause. These should be expressed as potential root cause rather than symptom cause. Systematically chart and check, based on experience and fundamentals, as to whether the evidence Supports, Disproves or is Neutral toward each hypothesis. Look for obvious flaws and inconsistencies. For example, the measured data on a refrigeration unit are not consistent with refrigerant data on a pressure–enthalpy diagram. or On a pipe, the pressure gauge downstream reads higher than the upstream gauge. 4. Plan Use the criteria to select a sequence of diagnostic actions. Sometimes several actions can be combined. However, usually it is best to wait for the results from the first action before we recycle back to the Explore stage and relook at our hypotheses. The criteria in selecting an action include: will the action provide background information to ensure the problem is understood in context or is the action to test a hypothesis? will the action produce results that give the accuracy needed? is the

2.3 Overall Summary of Major Skills and a Worksheet

action simple? inexpensive? safe? Stopping production to “inspect” or “change equipment” is usually very costly. 5. Do it Carry out the first action in the plan. Check. Monitor. 6. Look back Compare the results obtained with the hypotheses. Look back at the process used. Self-assess. Return to previous stages of Engage, Explore, Plan and continue. An example of the use of this Trouble-Shooter’s Worksheet is given in Section 2.4. 2.3.2

Feedback about your Trouble Shooting

Based on the evidence presented in Sections 2.1 and 2.2, the four things to look for in an effective trouble shooter are: the overall approach to problem solving, data handling, synthesis, and decision making. Table 2-1 summarizes the detracting and enriching behaviors for each. Worksheet 2-2 provides a summary worksheet that can be used by you to self-assess your TS process or can be completed by an observer to give you feedback about your TS process.

Table 2-1

Summary of detracting and enriching behaviors for trouble shooters.

Problem solving in general: How you did it Theme

Detracting behaviors

Monitors the thought No assessment of potential process gain from a question or action.

Enriching behaviors Asks “What will this get me?”

Unclear of type and purpose of question asked; just asks what pops into mind.

Knows clearly the purpose: ask fishing or shooting questions; whether creating hypotheses or checking for change or gathering information for clarification.

Does not monitor or ask questions as to Why? or implications.

Asks “Am I through?”, “Am I finished with this task?”, “Where is this leading me?”, “This should tell me ...” Checks and double checks instruments; checks if the equipment and lines are as on diagrams. Calibrates and recalibrates instruments.

Checks and double checks

Assumes everything is OK. Does not check instruments, diagrams, hardware, procedures.

Is systematic

Jumps all around, confused, and no apparent plan.

Identifies plan and follows it systematically yet flexibly. Uses tables or charts to keep track of idea flow.

29

30

2 The Mental Problem-Solving Process used in Trouble Shooting Subproblems and perspective

Confuses issues, factors, fault detection, solutions.

Breaks overall task into ones of situation clarification, or hypothesis testing and/or identify the change; into emergency action, cause identification, fault correction and future problem prevention. Identifies phases clearly and works through systematically.

Solves a minor fault while the process explodes.

Keeps situation in perspective, does not get lost in a subproblem.

Keeps whole problem and does not identify subproblems. No identification of a strategy.

Data handling: What you did Theme

Detracting behaviors

Enriching behaviors

Data resolution

Gathers data but does not know what it tells him/her.

Correctly identifies the usefulness of the data collected.

Asks any old question.

Matches hypotheses with the observed evidence to see if the hypothesis is consistent with the evidence.

Believes all he/she sees and hears; unclear of errors in information.

Explicitly states limitations of the instruments, measurements and checks these systematically.

No data gathered explicitly. Jumps in making corrective action without stating possible hypothesis or cause.

Gathers data for problem clarification and hypothesis testing/or change rather than jumping in with corrective action without any data.

Gathers data expensively, Takes process apart for everything. Overlooks simple ways of gathering information.

Gathers data easily through simple changes in operating procedure, puts controllers on manual.

Asks for samples, but assumes that sample locations and procedures are as usual.

Is present when samples are taken, bottles labelled.

Gives imprecise instructions: “Check out the instrument”; “Open up the exchanger”.

Gives precise instructions.

Based on intuition.

Based on fundamentals; estimates behavior based on fundamentals. Does mass and energy balances with at least two independent measurements.

Actions based on fundamentals

Does pressure profiles through units.

2.3 Overall Summary of Major Skills and a Worksheet Reasoning

Completeness

Jumps to invalid conclusions.

Draws valid conclusions; tests both positive and negative: what is; what is not; if it does happen; if it does not.

Error in inference: Confirmational bias. Uses only part of the information. Uses all resources. Doesn’t check the design calculations, or data from startup or data from initial, clean fluid; didn’t think of human error.

Synthesis: How you put it all together Theme

Detracting behaviors

Enriching behaviors

Hypotheses

Becomes fixed, thinks of only or selects one hypotheses; selects one at the start and cannot become unfixed.

Keeps at least four working hypotheses; keeps options open as data are gathered.

Makes everything complex.

Keeps it simple, especially if there is a “big failure”.

One view.

Many viewpoints: operators, design, human error, instruments, corrosion.

Critical of ideas; limited brainstorming.

Defers judgement when appropriate.

Considers only a “basics” strategy or a “change” strategy and sticks with it regardless of the evidence.

Selects either a “basics” strategy or a “change” strategy. Shifts strategy when its warranted.

Considers steady state only; considers only the facts

Considers unsteady state as well; considers the people too (the stress they might be under; the environment that allows open discussion; turf fights).

Cannot put all the ideas together into a reasonable story. Becomes fixed on one cause even when evidence points otherwise.

Can put the ideas together into a plausible explanation.

Flexibility

Overall synthesis

31

32

2 The Mental Problem-Solving Process used in Trouble Shooting

Decision making: How you put it all together Theme

Detracting behaviors

Priorities

No priorities for hypotheses; Sets and uses priorities. Keeps track and moves from top just start anywhere; indeed, may not even create a hypothesis! priority to second. At least four working one hypothesis; keeps options open as data are gathered. No priorities for gathering evidence; just collects something No priorities about the urgency of the situation; diddles around while the plant explodes; keeps customers waiting until problem completely solved.

Bias

Overall process

Enriching behaviors

Prioritizes; gathers the easy and cheap tests first. Visits the site. Prioritizes urgency; willing to use a contingency plan to get things going safely and later corrects the real fault.

Unbiased. Selects either a “basics” Biased, stacks the deck so the favorite fault will be selected strategy or a “change” strategy. Shifts even when the evidence refutes it. strategy when it is warranted. Biased: tests for only positive elements.

Tests for both positive and negative. Proves that hypothesis is correct and the other options are not correct.

No criteria used, or if they are, they are not measurable.

Uses measurable criteria to make decisions.

2.3 Overall Summary of Major Skills and a Worksheet

Worksheet 2-2: Summary observation form for feedback about Trouble Shooting.

TS name _____________________ Case _____ Initials ES _____ Obs ___ Rough work area: Process: how Data/analysis: what Monitoring _____________________ Data resolution _______________ Checking _______________________ Fundamentals? _______________ Systematic ______________________ Reasoning ___________________ Subs and perspective _____________ Completeness ________________ Decision making: how Priorities ______________________ Bias ___________________________

Synthesis: what Hypotheses __________________ Flexibility ____________________

Rating and Feedback Clarity of Communication None

Some

Most

All

Some

Most

All

Some

Most

All

Some

Most

All

Some

Most

All

Process used: None

Data collections and analysis: None

Synthesis: None

Decision-making: None

Five Strengths: ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ Two areas for improvement ________________________________________________________________ ________________________________________________________________

33

P&ID for the depropanizer and debutanizer.

2 The Mental Problem-Solving Process used in Trouble Shooting

Figure 2-4

34

2.4 Example Use of the Trouble-Shooter’s Worksheet

2.4

Example Use of the Trouble-Shooter’s Worksheet

Case ’8 Depropanizer: The temperatures go crazy (used courtesy of T.E. Marlin, McMaster University, Hamilton, Canada) The process shown in Figure 2-4 is being started up for the first time. It has been running well for several shifts. Just an hour ago, a new operator came on duty and the operator changed the pressure at which the Depropanizer, C-8, is operated, raising the pressure by 0.1 MPa. About 10 minutes after the pressure was increased, the tray temperatures began to go crazy and the bottoms level started to decrease. Figure 2.4 shows the P and ID for this process.

Trouble-Shooter’s Worksheet 2-3: Case ’8: The depropanizer: the temperatures go crazy ( copyright 2003 Donald R. Woods and Thomas E. Marlin)

1. Engage: Write down what is said; what you sense, smell, hear. If someone is telling you, then use skilled reflective statements to ensure you accurately obtain the information. .

.

Emergency priority: Safety? Hazard? Equipment damage? shut down &; safe park &. If not &, then: Draw a sketch of the process and mark on values. Provide a description in words of what is going on. This is a simple distillation column but all the piping and instrumentation details make it look complex. On the LHS, feed from upstream processing enters drum V29. This feed is pumped (via either a steam driven or motor driven centrifugal pump, F25, 26) through a preheater, E24, and into the depropanizer at tray 18. As the name suggests, the purpose of this column is to take overhead “propane and all lighter species”. Let’s follow the overhead. The overhead is condensed in two condensers in series, E25, collected in overhead drum V-30 with the non-condensibles (such as methane and hydrogen) removed from the drum and vented to the fuel-gas system. The pressure on the column, C8, is controlled by the valve on the vent system, PV 10. Condensed propane is pumped, F-27, from the drum V-30 forward as product, through product cooler E-26, and returned to the column as reflux. The reflux is flow controlled. Following the bottoms: a thermosyphon reboiler is steam heated. The bottoms flows forward to the next column, the debutanizer. No pump is needed because of the pressure difference between the depropanizer, 1.7 MPa, and the debutanizer, 0.48 MPa. I’m not sure at this stage if this is a control “problem” so I won’t elaborate further on the system at this time. I also will focus on the depropanizer, and not explore the debutanizer at this time.

35

36

2 The Mental Problem-Solving Process used in Trouble Shooting .

.

Manage any panic you might feel by saying “I want to and I can. I have a strategy that works. Let’s systematically follow it.” Monitor: Have you finished this stage? Can you check? What next? I’ve systematically followed my way around the flow diagram. I think I understand enough for now.

2. Define the stated problem: Systematically classify the given information using IS and IS NOT. If the information is not known at the stage check ?& to remind you to gather this information IS

IS NOT

(should be happening but it’s not) tray temperatures steady; bottoms level steady. ? & before the pressure increase; running well for several shifts. First plant startup.

WHERE

tray temperatures “go crazy”; bottoms level decreases. ? & 10 minutes after the pressure in column C8 increased by 0.1 MPa. ? & depropanizer, C8.

WHO

? & new operator.

WHAT WHEN

?& maybe upstream; no information about downstream debutanizer, yet! ? & not with previous operators.

– startup new process & suggest use Basics – startup after maintenance or change & suggest use Change – usual operation but changes made in operation but not in equipment & suggest use Basics – usual operation & suggest use Basics. .

Monitor: Have you finished this stage? Can you check? What next? Yes. I think I’ve finished.

3. Explore: Gather information to be gathered ? & in Define stage &. Perhaps questions about downstream effects. The decrease in liquid level could be because the flow has increased to the debutanizer (because the Dp) has increased. Exercise? & or a problem? & . I haven’t seen anything like this before Strategy: change & or basics & . Perspectives. Why? Why? Why? no information at this time that suggests this might be useful. ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________

2.4 Example Use of the Trouble-Shooter’s Worksheet

› ____________________________________________________________ Why? › ____________________________________________________________ Why? › Why?

Start ﬁ _________________________________________________________ . .

Prioritize: product quality & ; production rate &; profit & Goal: safe-park? & : short term now with long term later & ; long term now &

Action to be achieved: Specific terms and Measurable: level out the temperatures and the bottoms level Attainable? total reflux is a start but I hope it’s attainable Reliable? depends on my short-term solution will work on solving it quickly Timely? Safe? cannot think of major hazard now $ .

.

.

Check consistency of data/symptoms: inter-data consistency? OK & no & data consistent with fundamentals? OK & no & Type of problem: startup new process & maybe mechanical electrical failure usual operation & : ambient temp? & maybe fluids problems; high temperature? & then maybe materials problems System? failure of heat exchanger & > rotating equipment & > vessels & > towers & Identify key and What if?

What if? then What if? then What if? then What if? then

temperatures “going crazy” = temperature cycling focus on cycling symptoms / causes only temperature “cycling” and no decrease in level bottoms and tops temperature and pressures should be cycling too only bottoms level drop and no temperature “going crazy” root cause related to bottoms level drop column pressure increases condensation temperature at top increases; DT condenser increases and condensation should be easier; boiling temperature at bottoms increases; DT reboiler decreases so might shift from film to nucleate boiling giving higher heat flux, causing increased boilup or if nucleate to start with then insufficient area and boilup decreases.

37

38

2 The Mental Problem-Solving Process used in Trouble Shooting

List changes made & and/or trouble-shooting causes based on symptom & . “Crazy temperatures” and decreasing bottoms level sounds like a control problem. From Chapter 3, no symptoms listed for bottoms level dropping, but symptoms related to “oversized condenser” and “undersized reboiler” are: “Insufficient boilup”: [ fouling on process side]*/ condensate flooding, see steam trap malfunction, Section 3.5 including higher pressure in the condensate header/ inadequate heat supply, steam valve closed, superheated steam/ boiling point elevation of the bottoms/ inert blanketing/ film boiling/ increase in pressure for the process side/ feed richer in the higher boiling components/ undersized reboiler/ control system fault/ for distillation, overdesigned condenser. But if there was “insufficient boilup”, then the bottoms level should be increasing and not decreasing. This doesn’t make sense?? For: “Cycling of column temperatures:” controller fault/ [ jet flooding]*/ [downcomer flooding]*/ [ foaming]*/ [dry trays]* with each of the []* items listed as separate symptoms with their own root causes. [ jet flooding]*: excess loading/ fouled trays/ plugged holes in tray/ restricted transfer area/ poor vapor distribution/ wrong introduction of feed fluid/ [ foaming]*/ feed temperature too low/ high boilup/ entrainment of liquid because of excessive vapor velocity through the trays/water in a hydrocarbon column. [downcomer flooding]*: excessive liquid load/ restrictions/ inward leaking of vapor into downcomer/ wrong feed introduction/ poor design of downcomers on bottom trays/ unsealed downcomers/ [ foaming]* [ foaming]*: surfactants present/ surface tension positive system/ operating too close to the critical temperature and pressure of the species/ dirt and corrosion solids. [Dry trays]*: flooded above/ insufficient reflux/ low feedrate/ high boilup / feed temperature too high. .

.

Brainstorm root causes: summary of major ideas generated:

change in feed, too much overheads in feed, not enough feed, tray collapsed in stripping section, too much vaporized feed, pump F26 failure, pump F26 cavitates, increased pressure and DT increases, dry tray, flooded in rectification, insufficient reflux, low feedrate, feed temperature too high, boilup too high, too much feed to debutanizer, leak in bottoms, vaporizer flashes 90% (instead of 67%), failure of check-valve on idle pump outlet, boilup controller fault. Some of these are symptoms and not root cause, e.g. “not enough feed” “dry trays” .

Hypotheses: list in Chart; Symptoms: code and list in chart; Analyze with S supports; D disproves and N neutral or can’t tell.

2.4 Example Use of the Trouble-Shooter’s Worksheet

Symptom a. 10 min after column pressure increased, column temperatures go crazy b. 10 min after column pressure increased, bottoms level decreases c. d. e. Working Hypotheses

1. tray collapsed stripping section 2. too much bottoms fed to debutanizer 3. too much overheads in feed 4. feed valve FV1 stuck 5. pump F-26 not working 6. check valve on idle pump allows backflow 7.

Initial Evidence a S N N S S S

b S S S S S S

c

d

Diagnostic Actions e

A 4 4 4

B

C

D

4 4 4

Diagnostic actions: A. B. C. D.

readings of instruments on column visit site and listen to pump for cavitation visit site and see location of valve stem on FV-1 shut isolation valves on idle pump

4. Plan Select “read instruments” as the first task because it is inexpensive and should help test many of the hypotheses. Many of the key variables are displayed in the control room. 5. Do it Go to the control room, notebook in hand. 6. Look back

Comment: This example illustrates the approach one might take in being systematic.

39

40

2 The Mental Problem-Solving Process used in Trouble Shooting

2.5

Summary

Research on problem solving and on trouble shooting provide key information about the trouble-shooting process. Based on this research the four key features of the process are: the problem-solving process: monitoring, checking and double checking, being organized and systematic and keeping the problem in perspective. data handling and critical thinking: data gathering and resolution, based on fundamentals, with valid reasoning and being complete. synthesis: having five to seven working hypotheses, being flexible and putting it all together well. decision-making: based on criteria, priorities and avoiding bias.

.

.

.

.

A feedback form is given in Figure 2-3. The Trouble-Shooter’s Worksheet was created to aid the process. Its use was illustrated.

2.6

Cases to Consider

For each of the five cases given in Section 1.6 complete the Trouble-Shooter’s Worksheet 2-4. Trouble-Shooter’s Worksheet 2-4: ( copyright 2003 Donald R. Woods and Thomas E. Marlin)

1. Engage: Write down what is said; what you sense, smell, hear. If someone is telling you, then use skilled reflective statements to ensure you accurately obtain the information. .

.

.

.

Emergency priority: Safety? Hazard? Equipment damage? shut down &; safe park &. If not, & then: Draw a sketch of the process and mark on values. Provide a description in words of what is going on. Manage any panic you might feel by saying “I want to and I can. I have a strategy that works. Let’s systematically follow it.” Monitor: Have you finished this stage? Can you check? What next?

2. Define the stated problem: Systematically classify the given information using IS and IS NOT. If the information is not known at the stage check ?& to remind you to gather this information

2.6 Cases to Consider

WHAT WHEN WHERE WHO

IS

IS NOT

_____________________ _____________________ ?& _________________ ?& _________________ ?& _________________

(should be happening but it is not) _________________________________ ?& ______________________________ ?& ______________________________ ?& ______________________________

– startup new process & suggest use Basics – startup after maintenance or change & suggest use Change – usual operation but changes made in operation but not in equipment & suggest use Basics – usual operation & suggest use Basics. .

Monitor: Have you finished this stage? Can you check? What next?

3. Explore: Gather information to be gathered ? & in Define stage & Exercise? & or a problem? &. Strategy: change & or basics &. Perspectives. Why? Why? Why? ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________ Why? › Start ﬁ _________________________________________________________ . .

Prioritize: product quality & ; production rate &; profit & Goal: safe-park? &: short term now with long term later &; long term now &

Action to be achieved: Specific terms and Measurable: __________________ Attainable? ______________________________________________________ Reliable? ________________________________________________________ Timely? _________________________________________________________ Safe? ___________________________________________________________ $ _______________________________________________________________

41

42

2 The Mental Problem-Solving Process used in Trouble Shooting .

.

.

Check consistency of data/symptoms: inter-data consistency? OK & no & data consistent with fundamentals? OK & no & Likelihood of problem: startup new process & maybe mechanical electrical failure usual operation &: ambient temp? & maybe fluids problems; high temperature? & then maybe materials problems System? failure of heat exchanger &> rotating equipment & > vessels & > towers & Identify key and What if?

What if? _______________________ then _____________________________ What if? _______________________ then _____________________________ What if? _______________________ then _____________________________ .

. .

List changes made & and/or trouble-shooting causes based on symptom &. Brainstorm root causes: Hypotheses: list in Chart; Symptoms: code and list in chart; Analyze with S supports; D disproves and N neutral or can’t tell.

Symptom a. b. c. d. e. Working Hypotheses

Initial Evidence a

1 2 3 4 5 6 7

Diagnostic actions: A. B. C. D.

b

c

Diagnostic Actions d

e

A

B

C

D

43

3

Rules of Thumb for Trouble Shooting In Section 3.1 we consider the general rules of thumb for processes and different types of problems, instruments and people and the environment. In Sections 3.2 to 3.10 we consider different types of equipment: transportation, energy exchange, homogeneous phase separations, heterogeneous phase separations, reactions, mixing, size reduction, size enlargement and bins. Sections 3.11 and 3.12 consider “systems” and hazards. Guidelines and trouble-shooting rules of thumb are not available for many pieces of equipment. Often, guidelines for good practice are given since trouble can often results from “poor practice”. The style used for presenting information about trouble shooting is as follows. The “symptom” is shown in italics in quotes. This is followed by the root causes, separated by a slash, / root cause/ root cause. The causes are listed with the most-likely cause first, next-likely cause second and so on. Some causes are not root causes. Such causes are shown as [cause]*. Those causes are listed in square brackets with an *, for example [corrosion]* might be listed as a cause. But what is the root cause of the corrosion? Such “cause–symptoms” are listed separately with their root causes. For example, [Corrosion]*: inadequate stress relief for metals/ wrong metals chosen/ liquid flows at velocities > critical value.

3.1

Overall

Consider general rules of thumb and typical causes, rules of thumb about corrosion, for instrumentation and for people, respectively. 3.1.1

General Rules of Thumb and Typical Causes

Gans et al. (1983) suggests that big failures usually have simple causes, such as a compressor that will not start. On the other hand, small failures (or deviations from the norm) often are caused by complex causes, such as the product does not quite meet specifications because of a buildup of contaminants.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

44

3 Rules of Thumb for Trouble Shooting

The general most likely causes for failure differ depending upon whether this is the . . .

startup of a new process or startup after a shutdown and maintenance or fault that develops for an on-going, operating process.

a) Common Faults for First Time Startup. The faults encountered are: 75% Mechanical/electrical failures 20% Faulty design or poor fabrication

5% Faulty/inadequate initial data

leaks, broken agitators, plugged lines, frozen lines, air leaks in seals. unexpected corrosion, overloaded motors, excessive pressure drop in heat exchangers, flooded towers. often chosen to be the scapegoat by inexperienced trouble shooters.

b) Startup after Maintenance Ask questions about what specifically was changed, repaired, or modified. c)

Trouble for On-going Processes

For ambient temperature operations, about 80% of the problems experienced are fluid dynamical. For high-temperature operations, about 70% of the problems experienced are materials failure. Frequency of failures based on type of equipment: 17% 16% 14% 12% 10% 8% 8% 7%

heat exchangers. rotating equipment: pumps, compressors, mixers. vessels. towers. piping. tanks. reactors. furnaces.

Another approach is to consult data for Mean Time Between Failures, MTBF. Some example MTBF, in years, are: reciprocating compressors (nonlubricated), 2.8 years; gas turbines, 3.9; centrifugal pumps, 4; screw compressors, 4; reciprocating compressors (lubricated), 4.1; motors, 11.4; large induction motors, 16. See, for example, H.P. Bloch, “Looking for RCTA databases. Consider Failure statistics”, Hydrocarbon Processing, Jan 2002, p 36.

3.1 Overall

3.1.2

Corrosion as a Cause

Corrosion could cause trouble. Here are some general ideas about corrosion and some details about trouble shooting. General ideas: 1. 2. 3.

4.

5. 6.

The strength of materials depends totally on the environment in which the materials function and not on the handbook values. All engineering solids are reactive chemicals – they corrode. The usual eight forms of material failure are: 1) uniform corrosion: uniform deterioration of the material (32%); 2) stress corrosion: simultaneous presence of stress and corrosive media (24%); 3) pitting: stagnant areas with high halide concentration (16%); 4) intergranular corrosion: most often found in stainless steels in heated areas (14%); 5) erosion: sensitive to high flowrates, local turbulence with particles or entrained gas bubbles. For flowing gas-solids systems the rate of erosion increases linearly with velocity and depends on abrasiveness of particles (9%); 6) crevice corrosion: concentration cells occur in stagnant areas (2%); 7) selective leaching or dealloying: removal of one species from a metallic alloy (1%) and 8) galvanic corrosion: dissimilar metals coupled in the presence of a solution with electrolyte (negligible). Stress corrosion (the second most significant form of corrosion) can start from perfectly smooth surfaces, in dilute environments in material with stresses well below the yield stress. >70% of stress corrosion cracking is related to residual – not applied – stresses. The penetration of stress corrosion cracking as a function of time depends on the alloy composition, structure, pH, environmental species present, stress, electrochemical potential and temperature.

Trouble shooting:

“High concentration of metals (Fe, Cr, Ni, Cu) in solution”: [corrosion]*/ contaminants from upstream processing. “Ultrasonic monitoring shows thin walls for pipes, internals or vessels”: faulty ultrasonic instrument/ [corrosion]*/ faulty design. “Failure of supports, internals, vessels”: [corrosion]*/ faulty design/ unexpected stress or load. “Leaks”: [corrosion]*/ faulty installation/ faulty gasket/ faulty alignment. [Cavitation in pumps]*: pump rpm too high/ suction resistance too high/ clogged suction line/ suction pressure too low/ liquid flowrate higher than design/ entrained gas. [Corrosion]*: [corrosive environment]*/ inadequate stress relief for metals/ wrong metals chosen/ liquid flows at velocities > critical velocity for the system; for amine circuits: > 1 m/s for carbon steel and > 2.5 m/s for stainless steel/ large step changes in diameter of pipes/ short radii of curvature/ flange or gasket material projects into the pipe/ [cavitation in pumps]*/ improper location of control valves.

45

46

3 Rules of Thumb for Trouble Shooting

[Corrosive environment]*: temperature too high; for amine solution: > 125 C/ high dissolved oxygen content in liquid/ the liquid concentration differs from design; for steam: trace amounts of condensate or condensate level in condensers > expected; for 316 stainless steel: trace amount of sodium chloride; for sulfuric acid: trace amounts of water diluting concentrated acid; for amine absorption: total acid gas loadings > 0.35 mols acid gas/mol MEA, > 0.40 mols acid gas/mol DEA, > 0.45 mols acid gas/mol MDEA; makeup water exceeds specifications; for amine absorption units exceeds: 100 ppm TDS, 50 ppm total hardness as calcium ion, 2 ppm chloride, 3 ppm sodium, 3 ppm potassium and 10 ppm dissolved iron; for sour-water scrubbers: cyanides present/ pH change/acid carryover from upstream units/ high concentration of halide or electrolyte/ presence of heat stable salts/ bubbles present/ particulates present/ invert soluble precipitates with resulting underlying corrosion/ sequence of alternating oxidation-reduction conditions. 3.1.3

Instruments, Valves and Controllers

Trouble-shooting sensors: Most sensor faults are because of improper selection, incorrect installation or adverse environmental conditions. “Wrong signal”: fouled or abraded sensors/ bubbles or solid in fluid/ sensing lines plugged or dry/ electrical interference or grounding/ sensor deformed/ process fluid flow < design or laminar instead of turbulent flow/ contamination via leaky gaskets or O-rings/ wrong materials of construction/ unwanted moisture interferes with measurement or signal/ high connection or wiring resistance/ nozzle flappers plugged or fouled/ incorrect calibration/ orifice plate in backwards/ orifice plate designed incorrectly/ sensor broken/ sensor location faulty/ sensor corroded/ plugged instrument taps: for sourwater strippers: water or steam purge of taps malfunctioning or local temperatures < 82 C at which ammonium polysulfides form. “Wrong input”: sensor at wrong location/ insufficient upstream straight pipe for velocity measurement/ feedback linkages shift or have excessive play/ variations in pressure, temperature or composition of the process fluid. “Fluctuating signal”: bubbles in the liquid/ flashing because Dp across an orifice plate > design or fluid too close to the boiling point causing cavitation. Trouble-shooting control valves: The signal to the valve should be midrange; otherwise the signal depends whether the valve is fail open or fail close. “Leaks”: erosion/ corrosion/ gaskets, packing or bolts at temperatures, pressures and fluids that differ from design. “Can’t control low flowrate”: miscalibration/ buildup of rust, scale, dirt/[ faulty design]*. “Can’t stop flow”: miscalibration/ damaged seat or plug. “Excessive flow”: excessive Dp. “Slow response”: restricted air to actuator/ dirty air filters. “Noise”: cavitation/ compressible flow. “Poor valve action:” dirt in instrument air/ sticky valve stem/ packing gland too tight/ faulty valve positioner. “Cycling”: stiction. [Faulty design]*: valve stem at design flowrate is not at midrange. [Stiction]* is the sticking and friction related to valve movement and measured as the difference between the driving values needed to overcome static friction upscale and downscale. Likely cause of small amplitude, continuous cycling: gland too tight/ insufficient

3.1 Overall

driving force/antistiction coating damaged/ faulty valve positioner/ poorly tuned control system/ incorrect valve/ incorrect actuator. Trouble-shooting transmitters: “Erratic or fluctuating output”: vibrations/ improper orientation/ loose connections. Trouble-shooting block valves: Check that the arrow on the valve is in the same direction as the flow through the valve. Test via “turn and seal” to check movement and closure. “Reduced flow”: valve not fully open/ plugged with dirt. “Poor control by the control system”: block valve on bypass partially open/ block valves upstream or downstream of the control valve partially closed. Trouble-shooting check valves: “Noisy”: backpressure too high. “Reduced flow”: backpressure too high. Trouble-shooting control systems: “Oscillation”: feedforward/ poorly tuned/ valve sticks or has excessive hysteresis. “Returns with offset”: proportional control only. “Two related variables start to deviate”: lack of ratio controllers/ failure to relate analyses to flows. 3.1.4

Rules of Thumb for People

Nine suggestions are given. 1. 2.

3. 4. 5.

6.

Become aware of your own uniqueness and personal style, and how you might differ from the style of others. Honor the seven fundamental rights of individuals, RIGHTS. R, to be Respected; I, Inform or to have an opinion and express it; G, have Goals and needs; H, have feelings and express them; T, trouble and make mistakes and be forgiven; S, select your response to others expectations and claim these rights and honor these in others. Avoid the four behaviors that destroy relationships: Contempt, Criticism, Defensiveness and Withdrawal/ stonewalling. Trust is the glue that holds relationships together. Three elements of trust are benevolence, integrity and competence. Build trust by benevolence through loyalty to others, especially when they are not present and by not doing anything that would embarrass or hurt them. Build trust by integrity by keeping commitments to yourself-and others; clarifying expectations that you have of yourself-and of others; showing personal integrity, and honesty; apologizing promptly and sincerely when you know you are wrong; honoring the fundamental RIGHTS listed above and avoiding the killers; listening and understanding another’s perspective; being truthful; and accepting others “warts and all”. Build trust by being competent in your area of expertise. Destroy trust by the reverse of the Builders of trust listed above, and by selectively listening, reading and using material out of context; not accepting the experience of others as being valid; making changes that affect others without consultation; blind-siding by playing the broken record until you’ve even-

47

48

3 Rules of Thumb for Trouble Shooting

7. 8.

9.

tually worn them out or subtly make changes in the context/issues/wording gradually so that they are unaware of what is happening until it is too late. The 12:1 rule applies to rebuild relationships. 12 positive experiences are needed to overcome 1 negative experience. To improve and grow we need feedback about performance. Give feedback to others to encourage and help them; not for you to get your kicks and put them down. Focus on five strengths for every two areas to improve on. Be skilled at responding assertively. “When you... I feel .. adjust by...” .

3.1.5

Trouble-Shooting Teams

Use Worksheet 3-1 after each meeting, set goals and celebrate achievement. Use the framework developed by Francis and Young (1979) for growth; consult Fisher et al. (1995) for more short-term ideas. Trouble-shooting team meetings

These are organized by symptom with possible corrective responses suggested for chair {C} and member {M}. Problems with Purpose and Chairperson

“No apparent purpose for the meeting”: {C} don’t have a meeting. {M} question the purpose of the meeting. See also Agenda and timing problems. Agenda and Timing Problems

“No agenda”: {M}: phone {C} and ask for agenda. Invoke “no agenda, not attendance.”/at meeting: “Perhaps the first thing we should do is to create an agenda.” / After 5 minutes, “We seem to be lost. Could we draw up an agenda and follow that?” “Meeting drags on and on”: {C} should have circulated an agenda with times for each item and used the 20 minute rule/ {M} “Perhaps we can follow the agenda.”/ {M} indicate to {C} ahead of time the amount of time you have available for the meeting and then leave at that time. “Get off the track”: {M} seek direction, purpose, summary of progress. see also Behavior problems: “Subgroups interrupting and talking”. “Group gets bogged down”: state problem/ summarize/seek agenda clarification/ invoke 20 min rule. “Decisions made just at the end of the meeting”: state frustration/ suggest tabling/ suggest future corrective way to handle in future. See also Agenda and chairperson problems. Behavior and Participation Problems

“People come into a meeting cold:” {C or M} suggest reconvene meeting when all are prepared.

3.1 Overall

“Late arrivals”: {C} start meeting on time and continue with the agenda through the disruption of the retardee/ {M} “I realize that not everyone is here but I suggest that we start. It looks like a long agenda to get through.” “Some people do all the talking and some remain silent”: wrong membership/ encourage quiet ones to contribute/ ask each, in turn, to summarize his/her point of view/ ask a “safe” question of the silent ones/ privately check with the silent ones and reevaluate whether they need to attend/ ask open ended questions/ use nominal group. “Sub groups interrupting and talking”: identify problem/ suggest discussing one issue at a time and add subgroup’s issues to agenda/ be silent until the side conversation stops. “Thank you.” / Interrupt the side conversation. “Indecisive members, continual question asker”: ask for their ideas early/ redirect questions he/she asks back to him/her. Conflict or Apparent Conflict

“Conflict because of differing views:” restate the importance and value of everyone’s opinions, restate the RIGHTS/ attempt to bring conflict into the open/ summarize different views/ focus on different performance or opinions and not personalities/ remind of fundamental RIGHTS. “Conflict over facts:” stop the argument, identify problem as you see it and check that that is a problem/ identify facts we need clarified and probable expert. “Conflict over values, goals, criteria, process or norms:” stop discussion, identify problem as you see it and check that that is a problem/ use problem solving. “Resistance to new ideas, we tried that before, it won’t work, over my dead body, we don’t have the resources”: surface the resistance/ honor the resistance/ invoke consequence of no decision or of repeating what we’ve always done before/ use consensus building techniques/ reflect on the home turf of the objector and the impact the decision might have on them; explore if this might be brought to the group as an issue to address/ root cause of most resistance is fear of change, apathy, vested interests, not invented here, negativism, overwhelmed by enormity of proposal.

49

50

3 Rules of Thumb for Trouble Shooting

Worksheet 3-1:

Rating form for teams

Assessment of your team and team meeting

Name: _________________ Date: __________________ Purpose of team: ______________________________________ unclear & Purpose of this meeting ________________________________ unclear & Agenda for this meeting: detailed, clear and circulated ahead of time bare minimum circulated ahead of time none

& & &

Three-minute team task to seek consensus about the rating of the Task and Morale: .

Teamwork: Task all members clear about and committed to goals; all assume roles willingly; all influence the decisions; know when to disband for individual activity; all provide their unique skills; share information openly; the team is open in seeking input; frank; reflection and building on each other’s information; team believe they can do the impossible; all are seen as pulling their fair share of the load.

The degree to which these descriptors describe your team’s performance (as substantiated by evidence: meetings, engineering journal, interim report, presentations). None of Few of these these behaviors but behaviors major omissions & 1

& 2 .

& 3

& 4

Most features demonstrated

All of these behaviors

& 5

& 7

& 6

Teamwork: Morale: Trust high, written communication about any individual difficulties in meeting commitments; cohesive group; pride in membership; high esprit de corps; team welcomes conflict and uses methodology to resolve conflicts and disagreements; able to flexibly relieve tension; sense of pride; we attitude; mutual respect for the seven fundamental rights of all team members; Absence of contempt, criticism, defensiveness and withdrawal.

The degree to which these descriptors describe your team’s performance (as substantiated by evidence: meetings, engineering journal, interim report, presentations).

3.2 Transportation Problems

None of Few of these these behaviors but behaviors major omissions & 1

& 2

& 3

& 4

Most features demonstrated

All of these behaviors

& 5

& 7

& 6

Each, in turn, gives a 30-second summary of his/her perception of his/her contribution. This is presented without discussion. Individual, 30 second reporting of his/her contribution to this meeting: ___________________________________________________________________ ___________________________________________________________________ ___________________________________________________________________ ___________________________________________________________________ Four-minute team task to reach consensus about the five strengths and the two areas for growth. Strengths of your team ___________________________ ___________________________ ___________________________ ___________________________ ___________________________

Areas to work on for growth ___________________________________ ___________________________________

D.R. Woods (2005) This form should be completed after each meeting and copies used as evidence of growth.

3.2

Transportation Problems

Fundamentals of why fluids move: Fluids move from high pressure to low pressure, vertically because of gravity force, dragged along by a moving boundary or belt or because of density differences; won’t flow out of a sealed vessel or vacuum unless there is a vent break. These are expressed, on the macroscopic level, as Bernoulli’s equation. Most trouble-shooting problems encountered are fluid-dynamical problems. Centrifugal pumps are often selected to pump liquids; such pumps operate on their head-capacity curve showing decreasing head with increasing capacity. For pumping liquids, cavitation usually occurs when pumping hot liquids near their boiling temperature or when sucking liquids out of a sump. Whenever plants startup for the first time or after a shutdown, wood, sandwiches, bolts and other

51

52

3 Rules of Thumb for Trouble Shooting

crud could have been left unintentionally in the lines. Symptoms and possible causes for specific pieces of equipment are presented as follows: gas-moving equipment for pressure, Section 3.2.1 and for vacuum service, Section 3.2.2. Pumping liquids are considered in Section 3.2.3; pumping solids, Section 3.2.4. Considerations for steam are given in Section 3.2.5. 3.2.1

Gas Moving: Pressure Service

Fans, blowers and centrifugal and reciprocating compressors. Fans: Trouble shooting: “Noise”: vortex, flow separation/ loose bearings. “Discharge pressure low”: instrument error/ fans in series rotating in the same direction/ operating below the stall point/ density increase. “Low flowrate”: instrument error/ flow separation/ pitch angle of blades too shallow/ speed slow/ required system discharge high. Blowers: For rotary lobe: when used for pressure pneumatic conveying install a check valve in the blower discharge. Trouble shooting: “Discharge pressure high”: instrument error/ restriction in downstream line/ check valve jammed in closed position/ dirty intake filter. “Discharge pressure low”: instrument error/ slippage of the drive belts/ relief valve stuck open/ increasing air loss at the rotary valve due to larger clearance opening from wear/ loss of air caused by larger lobe clearance in the blower due to wear/ a leak, such as a ruptured hose, in a vacuum system/ a ruptured bag in the downstream bag house. Centrifugal compressors: Good practice: allow safety margins of design speed 5%, design head 10% and design power 15%. The sonic velocity decreases with an increase in gas molar mass. Trouble shooting: “Surging”: insufficient flow/ increased discharge pressure required by the system/ deposit buildup in diffuser. “Discharge pressure low”: instrument error/ compressor not up to speed/ excessive inlet temperature/ leak in discharge system. Provide separate anti-surge system for compressors operating in parallel; need careful design of suction piping for double flow compressors. Reciprocating-piston compressors: Good practice: design velocity through valves < 40 m/s. Trouble shooting: major faults: valves and piston rings. “Knocking”: frame lubrication inadequate/ head clearance too small/ crosshead clearance too high; “Vibration”: pipe support inadequate/ loose flywheel or pulley/ valve LP unloading system defective. “Discharge pressure high”: instrument error/ valve LP unloading system defective / required system discharge high. “Discharge pressure low”: instrument error/valve LP unloading system defective/ LP valve worn/ system leakage. “Discharge temperature high”: instrument error/ LP valve worn/ valve LP unloading system defective/required system discharge pressure high. “Cooling-water temperature high”: instrument error/ water flowrate low/fouled area/LP valve worn. “Valve temperature high”: instrument error/required system discharge pressure high/ run unloaded too long/LP valve worn. “Cylinder temperature high”: instrument error/ required system discharge pressure high/LP valve worn/ wrong speed. “Flow low”:

3.2 Transportation Problems

instrument error/LP valve worn/valve LP unloading system defective/dirty suction filter. 3.2.2

Gas Moving: Vacuum Service

Liquid piston pump, dry vacuum pump and steam ejectors. Liquid piston pump: Good practice: cool the seals with 0.03 L/s of clean cooling water at pressure at least 35 kPa greater than discharge pressure of pump. Trouble shooting: “Noisy”: service liquid level too high/coupling misaligned. “Capacity low”: suction leakage/ service liquid temperature too high/ speed too low/ seal water flowrate < design. “Power excessive”: service liquid level too high/ coupling misaligned. “Service liquid temperature high”: clogged strainer/ partially closed valve/ fouled heat exchanger. Dry vacuum pump: Good practice: size for usual discharge pressure 20–35 kPa gauge to allow for downstream discharge. Vacuum pumps run hot: 50–70 C. Allow 30-min warmup period before putting on-line. Allow 60 min purge before shutdown. Try not to have the pump discharge into a common header. Multistage pumps tend to run cooler than single stage. Install a check valve on the discharge. If the discharge pressure is > 35 kPa, add a positive displacement blower (designed for 6 Ndm3/ s at design conditions for the vacuum pump) with a bypass that is open for startup. Trouble shooting: “Loss of vacuum”: condensation in the suction line/condensation of species from other units connected to a common exhaust header/ increase in discharge pressure from restriction in downstream processing or pressure blowout in other units connected via common discharge header. “Excessive corrosion”: for systems handling acid gas or connected to such systems via common discharge header: warmup period too short/ shutdown purge too short. “Overheating”: low coolingwater flow/ fouled cooling system/ inlet gas temperature > 70 C. “High amps.”: buildup of polymer caused by operating temperature too high/ polymerizable species gain access via common discharge header. Steam ejectors: Good practice: operability of steam ejectors is very sensitive to the stability in the motive fluid (steam) pressure. Prefer vacuum pumps to steam ejectors. Keep diameter of pipes = diameter of inlet and discharge flanges of ejectors. For distillation columns, as the column overhead mass flowrate increases above design, so will the column overhead pressure and vice versa. Compression ratios per ejector: 6:1 to 15:1. If the inlet gas temperature < 0 C or below the triple point of water (0.61 Pa) then add steam jacketing to cope with ice formation. Seal for the hot well: submerge > 30 cm. The volume in the hotwell between the pipe and the overflow weir should be 1.5 times the volume in the down spout sealed. Replace any nozzles or diffusers where the area is >7% larger than design. Trouble shooting: check the last stage first and then move upstream. “Unstable operation or loss of vacuum”: steam pressure < 95% or > 120% of design/ steam superheated > 25 C/ wet steam/ inlet cooling-water temperature hot/ cooling-water flow-

53

54

3 Rules of Thumb for Trouble Shooting

rate low/ condenser flooded/ heat-exchange surface fouled/ 20–30% higher flow of non-condensibles (light end gases, air leaks or leaks from fired furnaces) / seal lost on barometric condenser/ entrained air in condenser water/ required discharge pressure requirement high/ fluctuating water pressure. “Water coming out of discharge”: upstream condenser flooded. 3.2.3

Liquid

The types of pumps include centrifugal, peripheral, reciprocating, rotary, gear and rotary screw. Centrifugal: Good practice: head-capacity curve should not be too flat if pump capacity is controlled by valve positioner. Select pump such that a larger diameter impeller could be installed later. An increase in flowrate causes an increase in required NPSH and a decrease in available NPSH. Trouble shooting: “No liquid delivery”: instrument error/not primed/ [cavitation]*/ supply tank empty. “Liquid flowrate low”: instrument error/ [cavitation]*/ non-condensibles in liquid/ inlet strainer clogged. “Intermittent operation”: [cavitation]*/ not primed/ non-condensibles in liquid. “Discharge pressure low”: instrument error/ noncondensibles in liquid/ speed too low/ wrong direction of rotation (or impeller in backwards if double suction). “Power demand excessive”: speed too high/ density liquid high/ required system head lower than expected/ viscosity high. Peripheral: Trouble shooting: “No liquid delivery”: instrument error/ pump suction problems/ suction valve closed/ impeller plugged. “Liquid flowrate low”: instrument error/ speed too low/ incorrect impeller trim/ loose impeller. “Discharge pressure low”: instrument error/ speed too low/ incorrect impeller trim/ loose impeller. “Power demand excessive”: speed too high/ improper impeller adjustment/ impeller trim error. Reciprocating: Trouble shooting: “No liquid delivery”: instrument error/excessive suction lift / [cavitation]*/ non-condensibles in liquid. “Liquid flowrate low”: instrument error/ excessive suction lift/ [cavitation]*/ non-condensibles in liquid. Rotary: sometimes NPSH is expressed as Net Inlet Pressure Required, NIPR, (or available NIPA), expressed as kPa absolute (kPa abs). Trouble shooting: “No flow”: instrument error/ [pump not turning]*/ [pump not primed]*/ relief valve not adjusted correctly or dirt keeping the relief valve open/ wrong direction of rotation/ [cavitation]*/ excessive suction lift. “Flow < design”: instrument error/ rpm too low/ air leak via bad seals or faulty pipe connections/ [ flow going elsewhere]*/ [high slip]*/ suction line clogged/ insufficient liquid supply/ [air or gas in liquid]*. “Starts but loses prime”: air leakage/ liquid vaporizing in suction line/ insufficient liquid supply. “Noisy operation”: [cavitation]*/ [air or gas in liquid]*/ [mechanical noise related to pump]*/ relief valve chatter/ drive-component noise. “Power > design”: higher viscous losses than expected/ pressure > design/ fluid viscosity > expected/ fluid “sets up” or solidifies in the line or pump during shut down/ fluid builds up on pump

3.2 Transportation Problems

surfaces/ rotating elements bind. “Short pump service life”: [corrosion]*/ abrasives present/ speed and pressures > design/ lack of lubrication/ misalignment. [Air or gas in liquid]*: [ fluid vaporizes]*/ air bleed missing/ fluid gasifies under operating conditions/ leaks in pumps or piping. [Air lock]*: [ fluid vaporizes]*/ air bleed missing/ fluid gasifies under operating conditions. [Cavitation]*: [ fluid vaporizes]* [Flow goes elsewhere]*: relief valve faulty or jammed open/ discharge flow diverted to wrong branch line. [Fluid vaporizes]*: [NPSH supplied too small]*/ fluid viscosity > design/ fluid temperature > design/ vapor pressure of fluid too high. [High slip]*: clearance between rotors > specs/ worn pump/ pressure > design. [NPSH supplied or NIPA supplied too small]*: strainer clogged/ temperature too high/ inlet line clogged/ inlet line diameter too small or length too long/ atmospheric pressure < design. [Mechanical noise related to pump]*: wrong assembly/ pump distortion because of wrong piping installation/ pressure > rating/ worn bearings/ worn gears/ loose gears/ twisted shaft/ sheared keys/ worn splines. [Pump not primed]*: valve on inlet line closed/ inlet line clogged/ air leaks/ pump rpm too low/ liquid drains or siphons out during off-periods/ check-valve missing or faulty/ [air lock]*/ worn rotors. [Pump not turning]*: drive motor stopped/ key sheared or missing/ belt drive broken/ pump shaft broken. Gear: Good practice: the higher the viscosity, the lower the rated rpm. On the discharge install a check valve and an expansion chamber or pulsation dampener on the discharge, the latter to reduce noise. For infrequent operation, operating pressure should be 20–30% < rated pressure. For continual operation, operating pressure << rated pressure and rpm < rated rpm. Never allow them to run dry. For startup, idle off-line for about an hour. Trouble shooting: usually performance does not break down suddenly; instead there is a gradual decrease in performance. Gear pumps are particularly susceptible to cavitation and erosion. “Low discharge pressure”: instrument error/ leakage/ low drive power/ faulty relief-valve setting/ [internal leakage]*/ [abrasion]*. “No liquid delivery”: instrument error/ suction line clogged/ drive not turning shaft/ check-valve fault. “Low liquid delivery”: instrument error/ drive power low/ [internal leakage]*/ [abrasion]*/ [cavitation]*. “Noisy”: entrained air in liquid/ liquid doesn’t drain from grooves/ misaligned drive and pump shafts/ faulty bearings/ loose mountings/ resonance because mating frequency of gears = natural frequency of gear train / rpm too high/ worn parts/ [cavitation]*. “Overheating”: liquid viscosity higher than expected/ liquid feed temperature too low/ faults in drive system such as misaligned drive and pump shafts. “Shaft won’t rotate”: drive system not working/ material in pump not melted/ temperature too low/ seized pump. “Significant oscillation in pump suction pressure”: instrument wrong/ faulty control system/ suction pressure setpoint too low for the process. “Pump discharge pressure oscillates”: instrument wrong/ starved feed to pump/ change in viscosity of feed/ damaged pump internals.

55

56

3 Rules of Thumb for Trouble Shooting

“Inadequate volumetric efficiency”: decrease in viscosity/ pressure in upstream process increase/ worn pump. [Abrasion]*: grit in liquid/[cavitation]*/ pH different from design. [Cavitation]*: pump rpm too high/ suction resistance too high/ clogged suction line/ suction pressure too low/ liquid flowrate higher than design/ entrained gas. [Internal leakage]*: excessive clearance between the gear and sides or end plates/ [abrasion]*. Rotary screw (mono): Good practice: stator has one more lobe or screw than the rotor; Reduce rpm for abrasives; starting torque = 4 initial torque. If particles are present, try to minimize the abrasion by using viscosity > 5000 mPa s. Backflow or “slip” is reduced as viscosity increases. NPSH problems are usually not important except for suction lift, pumping from a vacuum and fluid vapor pressures of > 15 kPa. Trouble shooting: “No liquid delivery”: instrument error/ wrong direction of rotation/ insufficient suction lift/ clogged inlet/air leaks on suction/faulty pressure relief valve/ worn pump. “Rapid wear”: discharge pressure too high/ pump runs dry/ incorrect materials of construction/ speed too high for abrasives/ viscosity too low for abrasives. “Noisy”: insufficient feed flowrate/ air leak in suction/gas in feed liquid/ speed too high/ poor alignment. “Excessive power”: rpm too high/ liquid viscosity > design/ operating pressure > design/discharge line plugged/ stator expanded or swollen. “Failure of the stator”: bond failure (pH > 10 or local hot spot)/ temperature > design. “Initially OK but gradual increase in power needed”: swelling of elastomeric stator coating because of chemical attack. 3.2.4

Solids

Johanson’s1), 2) definitions of terms used to characterize solid particles are given in Section 3.7.3. The important terms are AI, RI, HI, FRI, FDI, BDI, CI, RAS and SBI. Bucket elevators: Trouble shooting: major difficulties are unloading and loading: jamming of materials between the buckets and the side of the boot. Rotary/star valve: Trouble shooting: use amps as guide to solids throughput. Keep air velocity high enough to prevent plugging of the air-vent line. Pneumatic conveying: dilute phase: for vacuum: Trouble shooting: air leakage and powder arching in hopper are the major threats. “No flow or flow < design”: air leaks/ powder arching in feed hopper, see Section 3.10/ low solids flow because of increased air loss in rotary valve/ wrong type of rotary valve used/ insufficient air/ line too long/ vacuum pump problems, see Section 3.2.2 . “Pressure (vacuum) at suction to blower > design (vacuum < design)”: air leaks/ failure of discharge valve to seal on the receiver. “Erratic pressure readings”: irregular feed. “Explosion”: moisture too low/ lines not grounded. “Does not sound “tinny” when listening with stethoscope”: material accumulated inside pipe at this location. 1) J.R. Johanson, 2002, “Troubleshooting bins,

hoppers and feeders,” Chem. Eng. Prog. April pp. 24–36.

2) J.R. Johanson, 2000, “Smooth out solids

blending problems,” Chem. Eng. Prog. April, p. 21.

3.2 Transportation Problems

Pneumatic conveying: dilute phase: for pressure: Trouble shooting: use pressure at the outlet of the blower as prime indicator. “Dp across blower > design or 2:1 ratio”: restriction in downstream conveying line/ check valve jammed closed/ dirty intake filter/ plugged discharge silencer/ increase in feed to the system/ length of pipe > design. “Dp across blower < design”: slipping v-belts/ air loss at the rotary valve. “No flow”: [plugged line]* “No flow or flow < design”: overfed fan system/ insufficient air/ insufficient solids/ line too long/ inlet air pressure too low. “Erratic pressure readings”: irregular feed. “Amps on rotary valve < usual”: solids flow < design/ air loss through the rotary valve/ increased clearances. “Does not sound “tinny” when listening with stethoscope”: material accumulated inside pipe at this location. “Gradual decrease in performance”: wear on the blower caused by dusty air. [Plugged line]*: within the first couple of metres of the beginning of the system: material feed problems/air supply problems. [Plugged line]* after the first couple of metres: air leak with the plug occurring about 10 m downstream of leak/ erosion of rotary valve causing increase in air leakage. Pneumatic conveying: dense phase: Trouble shooting: “No flow or flow < design”: plugged line/ malfunction of line boosters because of stuck check valve/ high humidity. “Solids fed to conveying line < design”: ratio of air to fluidize in the blow tank relative to convey is too small/ fault in control system. “Solids fed to conveying line > design”: ratio of air to fluidize in the blow tank relative to convey is too large/ fault in control system.”Solids flow= 0”: top discharge and the ratio of air to fluidize to convey is too small. “Solids flow gradually decreases”: restriction in the discharge pipe/ blinding of the fluidizing membrane. Feeder: volumetric for extruder: Trouble shooting: “Does not run”: no power/ jammed. “Stalls”: material jam/ current limit set too low. “Erratic speed control”: controller poorly tuned/ sensor malfunction/ material jam. “Feed rate variable”: particles arching in the hopper/ moisture level too high/ overheated polymer (prematurely fused) feed polymer. Feeder: screw conveyor: Trouble shooting: “Shear pins on feeder drive break”: screw diameter < exit hole from bin. “Motor overload on feeder drive”: screw conveyor diameter < exit hole from hopper. “Screw feeder initially OK then motor overloads”: screw flight spacing in the direction of sold flow decreases markedly/ difference between FDI and BDI < 5% suggests a moderately incompressible solid whose flow is very sensitive to screw flight spacing. Feeder from bottom of hopper: Trouble shooting: “Feeder motor overloads immediately:” wrong wiring/foreign material in feeder/ hopper is full and solids give excessive solids pressure because of particle characterization and hopper design / FDI large and large HI. “Feeder exit flowrate suddenly < expected”: blockage in hopper outlet/ lumps of particles forming in hopper/ large RI and small HI possibly caused by temperature cycles. “Feeder exit flowrate gradually < design:” solids builup in the feeder/ large CI, large AI and RI/ wrong materials of construction in feeder. (Often happens with vibrating feeder.) Feeder: belt feeder from the bottom of a hopper: Trouble shooting: “Belt feeder initially starts but suddenly stops with motor overload”: gap between the belt and hopper

57

58

3 Rules of Thumb for Trouble Shooting

interface edge is too small/ belt sags between pulleys/ large FDI and small% difference between FDI and BDI. 3.2.5

Steam

Good practice: Take steam off the top of the steam header; put condensate into the top of the condensate return header.

3.3

Energy Exchange

The fundamentals for thermal energy exchange are that heat flows from a high temperature to a low temperature. Thermal forms of energy are not always available to do work. Overall energy is conserved; often we write expressions for the mechanical energy balance (on the macroscopic level this is Bernoulli’s equation) and the thermal energy balance (on the macroscopic level this is q= UA LMTD). When trouble shooting heat exchangers, usually the fault is fluid dynamical: liquids don’t drain; baffles are placed so that liquid can’t go where you expect; vents are missing that prevent us from bleeding off trapped gases. Fluids to watch are water and hydrogen; both have extremes in thermal properties. Thermal expansion will occur when exchangers are brought up to temperature. This may cause a leak at the head-totube sheet joint if the difference between the temperature on the tubeside less the temperature of the bolts > 50 C. For systems involving steam, scrutinize the steam trapping system: ensure that traps are not flooded, that the appropriate trap has been installed, that the bypass is not left open, and that thermodynamic traps are not fed to a common header. Steam should come from a nozzle on the top of the steam main; condensate should be discharged into the top of the condensate header. More specifically, we list symptoms and possible causes for the following equipment: In this chapter we consider first providing mechanical drives, in Section 3.3.1. Furnaces are considered in Section 3.3.2. Heat exchangers, condensers and reboilers are listed in Section 3.3.3. Sections 3.3.4, 3.3.5 and 3.3.6 consider refrigeration, steam generation and high-temperature heat-transfer fluids, respectively. 3.3.1

Drives

Engines: Good practice: use high efficiency motors when replacing or repairing existing installations. Trouble shooting: “Hammering/knocking”: loose parts/ seized parts. “Pre-ignition”: fuel with unstable hydrocarbons/ incorrect timing. “Detonation”: wet fuel/ incorrect timing/ intake air too hot/ glowing carbon on the piston/ leaking valve stem/ worn valve guides. “Misfiring”: incorrect timing/ faulty ignition elements/ wrong gap in the spark plugs/ wet fuel/ spark-plug gap coated or filled with carbon or oil. “Overheat”: lubrication failure/ inadequate cooling/ poor quality fuel/

3.3 Energy Exchange

fuel to air ratio too lean. “Sooty exhaust”: incorrect fuel/ air- fuel ratio too rich/ inadequate cooling/ wrong valve adjustment. “Valve leaking”: inadequate cooling/ valve angle incorrect/ wrong metallurgy. “Piston blow-by:” over lubrication. “Worn bearings”: misaligned crankshaft. Electric motor: Trouble shooting: “Won’t start”: overload trip/ loose connection/ grounded winding/ grounded stator. “Runs backwards”: reversed phase sequence. “Excessive noise”: 3 phase machine single phased/ unbalanced load between phases. “Synchronous motor fails to come up to speed”: faulty power supply or overload trip/ windings grounded. “Overheat”: unbalanced load between phases/ wrong line voltage/ short circuit in stator winding. DC Motors: “Won’t start”: weak field/ low armature voltage/ open or short circuit in armature or field. “Runs too slow”: low armature voltage/ overload/ brushes ahead of neutral. “Runs too fast”: high armature voltage/ weak field/ brushes behind neutral. “Brushes sparking”: brushes worn/ brushes poorly seated/ incorrect brush pressure/ dirty, rough or eccentric commutator/ brushes off neutral/ short-circuited commutator/ overload/ excessive vibration. “Brush chatter”: incorrect brush pressure/ high mica/ incorrect brush size. “Bearings hot”: belt too tight/ misalignment/ shaft bent/ damaged bearings/ wrong type of bearings. Steam turbine: Good practice: consider extracting energy via a steam turbine for any pressure reduction in steam service. Use high-pressure steam for energy; lowpressure steam for heating. Don’t operate with wet steam. Trouble shooting: “Turbine fails to start:” too many hand valves closed/ nozzles plugged or eroded/ dirt under carbon rings. “Slow startup”: throttle-valve travel restricted/ steam strainer plugged/ load > rating. “Insufficient power”: throttle-valve travel restricted/ too many hand valves closed/ oil relay governor set too low. “Speed increases as load decreases”: throttle-valve travel restricted/ throttle assembly friction/ valve packing friction. “Governor not operating/ excessive speed variation”: governor droop adjustment needed/ governor lubrication problem/ throttle-valve travel restricted. “Overspeed trip on load changes”: trip valve set too close to operating speed/ throttle-valve travel restricted/ throttle assembly friction. “Overspeed trip on normal speed”: excessive vibration/ dirty trip valve/ trip valve set too close to operating speed. “Leaking glands”: dirt under carbon rings/ worm or broken carbon rings/ scored shaft. Steam turbine used for the generation of electricity: Trouble shooting: “Turbine overspeeding”: [load disconnection suddenly]*/ [Trip Throttle Valve stuck]*/ control valve fault// [extraction valve fault]*. “Bearings damaged”: [turbine overspeeding]*/ [lube oil]*/ excessive vibration/ no lube oil/ bearing temperature too high/ flow of parasitic currents/ [clogging]*/ [electronic pin clogging]*. [Clogging]*: [lube oil]*/ long time without operating. [Electronic pin clogging]*: [lube oil]*/ long time without operating. [Extraction valve fault]*: wear on valve bearing/ loss of hermetic seal. [Load disconnection suddenly]*: operator error/ automatic bus bar protection because of downstream changes in electric system. [Lube oil]*: low pressure/ oil temperature too high/ oil too old/ oxidation/ water contaminates oil.

59

60

3 Rules of Thumb for Trouble Shooting

[Solenoid valve malfunction]*: [electronic pin clogging]*/ [clogging]*/ solenoid shorted coil/ faulty control signal/ sensor error. [Trip Throttle Valve stuck]*: [clogging]*/ [solenoid valve malfunction]* Gas turbine: consists of a compressor, combustor and turbine sections. Trouble shooting: “Combustion noise:” fouled or clogged combustor/ loose or cracked lining in combustor. “Vibration”: bearing failure in compressor or turbine/ blade damage in compressor or turbine/ surging compressor/ fouled turbine. “Exhaust temperature > design”: combustor fouling. “Exhaust temperature < design”: combustor clogged. “Thermal efficiency < design”: fouled turbine/ turbine blade damage/ turbine nozzle distortion. “Mass flow < design”: compressor fouling/ compressor filter clogged/ compressor blades damaged. 3.3.2

Thermal Energy: Furnaces

Multi-use including heating, boiling, reactions. Related topics distillation, Section 3.4.2. For fired furnaces: monitor CO and excess air to reduce rejected energy and improve efficiency, consider the installation of economizers and air preheaters to recover additional heat from the flue gas. For steam generation: preheat boiler feed water with available low-temperature process streams, maximize the use of heat-transfer surfaces by optimizing sootblowing frequency and decoking of tubes, flash blowdown to produce low-pressure steam if required. Trouble shooting: “Gas temperature > design”: instrument wrong/ insufficient excess air/ process side coking of tubes/ leak of combustible material from process side/ overfiring because of high fuel-gas pressure. “Gas temperature < design”: instrument fault/ fouling/ too much excess air/ insufficient area/ fuel-gas pressure < design. For convection furnace: “Exit process gas temperature < design”: excess air/ decrease in flame temperature/ damper has failed closed. “Pressure inside furnace > design”: instrument wrong/ fouling on the outside of the tubes in the convection section/ exhaust fan failure. “Faint blue-gray smoke rising from top of furnace”: fouling outside tubes in the convection section/ pressure in furnace > atmospheric. “Puffing, rhythmic explosions”: burners short of air for short period causing minor over-firing/ wind action/ start up too fast. “Tube failure”: localized overheating/ burning acid gases as fuel/ free caustic in water and dryout/ dry out and attack by acid chloride carried over from water demineralization/ breakthrough of acid into water from demineralizer. “High fuel-gas pressure”: failure of pressure regulator. “Tube dryout”: tubeside velocity too low. “Low furnace efficiency”: high combustion air flow/ air leak into the firebox/ high stack temperature/ heat leaks into the system. “Equipment suddenly begins to underperform”: fouling/ bypass open. “Temperature-control problems”: missing or damaged insulation/ poor tuning of controller/ furnace not designed for transient state/ unexpected heat of reaction effects/ contaminated fuel/ design error.

3.3 Energy Exchange

3.3.3

Thermal Energy: Fluid Heat Exchangers, Condensers and Boilers

For not truly countercurrent, if the correction factor for the LMTD drops below 0.75 we run the risk of temperature crossover. Provide pressure relief to allow for systems where block valves could isolate trapped fluids. Include impingement baffles at shell inlet nozzles to prevent erosion of tubes and flow-induced vibration. Account for the larger heat exchange that occurs for clean tubes/surfaces; the design was based on reduced heat-transfer coefficients that accounts for ultimate dirty film resistance. Ensure the air is vented. Liquids being heated should leave at the top of the exchanger to prevent the buildup of gases coming out of solution and vice versa for liquids with suspended solids or viscous fluids. Orient baffle windows to facilitate drainage. Slope condensers to remove the condensed phase. Maximum cooling-water temperature is 45 C. Prefer water or other nonflammable heat-transfer media. For flammable heat transfer fluids, select operating temperature below its atmospheric boiling temperature. If refrigeration is required, prefer less hazardous refrigerant even if this means operating at higher pressures. Use pinch analysis, identify inefficient exchanges and retrofit heat-exchanger networks to maximize heat recovery. Optimize cleaning schedule. Consider “on-line” mechanical cleaning where fouling is a problem. Use turbulence promoters in laminar flow and gas services and where turndown has significantly reduced the heat-transfer coefficient. For air-cooled systems include a trim cooler with water as coolant. Shell and tube heat exchangers: Good practice: to provide lower inventory and intensify, prefer plate exchangers to shell and tube exchangers with the highest surface compactness. Trouble shooting: “Thermal underperformance on both streams (coolant exit temperature < design; hot exit > design temperature):” instrument fault/ not enough area/ thermal load reduced via flowrate or change in thermal properties (eg, less hydrogen than design)/ inerts blinding tubes/ [ fouling]* more than expected/ tube flooded with condensate (see faulty steam trap, Section 3.5.1) or trap in backwards or insulated inverted bucket steam trap. “Equipment suddenly begins to underperform”: fouling/ bypass open. Temperature-control problems”: missing or damaged insulation/ poor tuning of controller/ not designed for transient state/ unexpected heat of reaction effects/ contaminated feeds/ design error/ unexpected heat of solution effects/ changes in properties of the fluids. “Heat transfer to shell side fluid < design and Dp < design”: instrument/ increase in viscosity/ fluid bypasses baffles (baffle cut > 20%, no sealing strips, excessive baffle clearance, shell side nozzles too far from tube bundle)/ stratification/ faulty location of exit nozzles/ faulty baffling/ inlet maldistribution. “Heat transfer to tubeside < design and uneven (and uneven tubeend erosion at inlet)”: maldistribution to the tubes (axial nozzle entry velocity > tube velocity, for radial nozzle entry velocity > 1.9 tube velocity). “Heat transfer to one fluid < design and Dp= design”: instrument fault/ oil contamination of water. “Thermal overperformance both fluids, and usually Dp > design on hot fluid side, perhaps charring of cold stream and freezing of hot stream:” instrument fault/ cocurrent piped incorrectly as countercurrent/ area too large/ hydrogen concentration in gas stream > design/ clean tubes but design area selected on dirty service. “Thermal overperformance one

61

62

3 Rules of Thumb for Trouble Shooting

stream: cold exit temperature > expected”: instrument/ plugged tubes/ inlet velocity < design, fouled screen on pump suction/ pump problems, see Section 3.2.3/ increased heat load. “Poor control of outlet temperatures (–5 C)”: poor tuning of control/ instrument fault/ oversized area combined with multipass with local changes in effective MTD with fluid velocity. “Rapid tube failure or glass or karbate tubes break”: inlet gas velocity too high and directed onto tubes/ gas velocity > 5 m/s causing tube vibration/ surges in cooling water/ surges cause by syphon without vent break. “Higher Dp when operating at design flows and temperatures”: underdesign/ design for 2-phase stratified flow but slug flow occurs/ gas service but the operating pressure < design. “Leaks”: erosion/ corrosion/ vibration/ improper tube finishing/ cavitation/ lack of support for tube bundle/ tube end fatigue. “Leaks from the gasket at the tube sheet joint”: sudden upsets that cause the DT between the flange and the bolt to be > 50 C. “Noise/ vibration”: excessive clearance between baffles and tubes/ inlet gas velocity too high and directed onto tubes/ gas velocity > 5 m/s causing tube vibration/ surges in cooling water/ surges cause by syphon without vent break. “Gradual reduction in heat transfer and increase in Dp”: small tube leaks. [Fouling]*: change in pH/ water temperature high and invertly soluble compounds precipitate/ water temperature high and algae and fungi form/ corrosion products/ sublimation/ process condensate freezes/ coolant fouling/ silt deposits/ aggregation and destablization of colloids causing wax and asphaltenes to deposit from hydrocarbons. For more on [ fouling] see Section 3.11. Shell and tube condensers: Trouble shooting: “Condensation duty < design; exit vapor temperature > design, high flowrate of vapor out vent”: instrument fault/ undersized condensers/ change in process gas pressure/ inward leakage of non-condensibles/ change in feed composition/ [ fouling]* on the process side/ vapor binding/ vapor pockets/ inert blanketing (usually near the condensate outlet for condensers operated flooded for pressure control)/ condensate flooding, see steam traps, Section 3.5.1/ baffle orientation horizontal not vertical/ excessive entrainment in vapor feed/ baffle window > 45%/ drain line too small/ leakage between the tubesheet and baffles/ bowed tubesheet/ condenser designed for horizontal service installed vertically. “Condensation duty > design:” excess condenser area/ clean tubes/ condenser designed for vertical service installed horizontally/ liquid entrainment in feed. “Condensation duty < design and Dp process > design and excessive flow out vapor vent”: undersized condenser. “Coolant water temperature > design”: instrument fault/ low coolant flowrate/ high coolant inlet temperature/ cooling tower fault/ excess condenser area. “Cooling water exit temperature > design and higher steam usage in distillation column reboiler and uneven column operation:” excess condenser area via overdesign or clean surfaces. Heat transfer drops off > rate than expected and Dp increases faster than expected”: [ fouling]* because of oversized kettle reboiler on distillation column or change in pH or flow regime laminar when design was turbulent or higher level of contamination in fluids or crud carryover from upstream equipment (e.g. silica from catalyst in upstream reactor) or compensation for oversize by reduced coolant flowrate. “Loss of volatile vapor out vent, high vent-gas temperature, degree of subcooling < design and unusual temperature profile between vapor inlet and condensate outlet:”

3.3 Energy Exchange

instrument error/ underdesign. “Loss of volatile vapor out vent, apparent undersized area for condensation of immiscible liquids”: lack of subcooling of condensate/ condenser installed horizontally instead of vertically. “Fog formation”: high DT with noncondensibles present/ high DT with wide range of molar mass of the vapor. “Equipment suddenly begins to underperform”: [ fouling]*/ bypass open. Temperature-control problems”: missing or damaged insulation/ poor tuning of controller/ not designed for transient state/ unexpected heat of reaction effects/ contaminated feeds/ design error/ unexpected heat of solution effects/ changes in properties of the fluids. [Fouling]*: change in pH/ water temperature high and invertly soluble compounds precipitate/ water temperature high and algae and fungi form/ corrosion products/ sublimation/ process condensate freezes/ coolant fouling/ silt deposits. For more see Section 3.11. Shell and tube reboilers: Good practice: to provide lower inventory and intensify, prefer thermosyphon reboilers to kettle reboilers. If DT > 25 C, probably the boiling mechanism is film boiling. If DT < 25 C, usually the boiling mechanism is nucleate boiling. Trouble shooting, general:”Insufficient boilup”: [ fouling on process side]*/ condensate flooding, see steam trap malfunction, Section 3.5.1 including higher pressure in the condensate header/ inadequate heat supply, steam valve closed, superheated steam/ boiling point elevation of the bottoms/ inert blanketing/ film boiling/ increase in pressure for the process side/ feed richer in the higher boiling components/ undersized reboiler/ control system fault/ for distillation, overdesigned condenser. “Equipment suddenly begins to underperform”: [ fouling]*/ bypass open. Temperature-control problems”: missing or damaged insulation/ poor tuning of controller/ not designed for transient state/ unexpected heat of reaction effects/ contaminated feeds/ design error/ unexpected heat of solution effects/ changes in properties of the fluids. “Insufficient boilup and gradual increase in steam pressure to maintain boilup:” [ fouling]*/ inerts in steam. “Insufficient boilup and gradual decrease in steam pressure to maintain boilup:” steam blowing, see steam trap malfunction, Section 3.5.1. “Water contamination”: leak. “Cycling (30 s–several minutes duration) steam flow, cycling pressure on the process side and, for columns, cycling Dp and cycling level in bottoms”: instrument fault/ condensate in instrument sensing lines/ surging/ [ foaming]* in kettle and thermosyphon/ liquid maldistribution/ steam-trap problems, see Section 3.5.1, with orifice Dp across trap < design/ temperature sensor at the feed zone in a distillation column/ collapsed tray in a distillation column. “Level high in reboiler”: instrument/ inlet or exit pipe nozzle too small/ wrong nozzle orientation/ steam trap fault, see Section 3.5.1/ steam trap is above the reboiler. “Breathing: puffs of vapor and entrained liquid leave reboiler:” overdesign/ clean tubes when designed for fouled conditions. [Inadequate heat supply]*: wet steam/ too great a Dp across steam valve gives wiredrawing and superheat/ steam valve closed/ control system fault. Kettle: Good practice: rarely underdesigned and should not be used for foams. Trouble shooting: general plus the following symptoms and causes unique to this type of reboiler:

63

64

3 Rules of Thumb for Trouble Shooting

“Surging”: poor liquid distribution/ [ fouling]*. “Low boilup rate and gradual increase in steam”: film instead of nucleate boiling/ too high a DT/ clean tubes/ conservative overdesign/ [ fouling]*/ flooding with condensate because of steam-trap problems, see Section 3.5.1/ bottom temperature elevation/ increase in column pressure/ feed concentration of light components < design/ not enough heating medium. “Low boilup rate and decrease in steam pressure”: steam trap blowing, see Section 3.5.1. “Low boilup rate, pressure increase in reboiler and surges”: [ foaming]*/ inerts/ leaks/ undersized reboiler/ diameter of vent line too small/ top tubes not covered with liquid/ high liquid level that floods the vapor disengagement space/ inlet feed maldistribution/ inadequate vapor disengagement. [Foaming]*: see Section 3.11. [Fouling on the process side]*: low liquid level causing vapor-induced fouling/ solids in feed that are trapped by the overflow baffle. For more on fouling see Section 3.11. Thermosyphon: Good practice: vertical thermosyphon reboilers are usually not used for vacuum or extremely high pressure service. Trouble shooting: general plus the following symptoms and causes unique to this type of reboiler: “Insufficient boilup”: [ fouling on the process side]*/ insufficient steam flow/ condensate flooding/ low liquid level in distillation column gives low liquid circulation and increased fouling/ high liquid level in distillation column (static head > design) or higher density of feed liquid gives higher boiling temperature and circulation and insufficient vaporization for vertical thermosyphon/ pipe lengths < design/ pipe diameter > design/ process fluid in vertical thermosyphon drops below 30–40% of the tube length. For horizontal thermosyphon, appears to be undersized but the cause is liquid feed maldistribution. “Surges in boilup”: process fluid circulation rate too low/ [ fouling on the process side]*/ wide boiling range/ overdesign. “Cycling (30 s– several minutes duration) steam flow, cycling pressure on the process side and, for columns, cycling Dp and cycling level in bottoms”: in addition to general, all natural circulation systems are prone to surging/ feed contains high w/w% of high boilers/ vaporization-induced [ fouling]*/ constriction in the vapor line to the distillation column. For horizontal thermosyphon: maldistribution of fluid temperature and liquid. [Fouling]*: insufficient static head/ excess friction in the pipes/ on the tubeside the outlet nozzle area < total tube area/ on tubeside the inlet nozzle area < 0.5 total tube area/ rate of vaporization > 25% of circulation rate/ mass rate of vaporization > mass rate of circulation/ natural circulation rate < 3 vaporization rate/ vaporization induced solids. For more see Section 3.11. Forced circulation: operates with sensible heat mode in the tubes. Trouble shooting: general plus the following symptoms and causes unique to this type of reboiler: “Unstable”: insufficient NPSH in pump, see Section 3.2.3. “Insufficient boilup and rapid fouling”: insufficient circulation/ pump fault, see Section 3.2.3/ plugged circulation lines. “Insufficient boilup”: [ fouling]*/ circulation rate low/ pump problems, see Section 3.2.3/ no vortex breaker/ excessive circulation and a wide spread in boiling temperatures in bottoms. “Excessive vapor in flash chamber, unstable distillation column operation and apparent underdesign of overhead condenser”: overdesign.

3.3 Energy Exchange

Vertical falling-film evaporator: see also Absorbers, Section 3.4.8 and Evaporators, Section 3.4.1 and falling-film reactors, Section 3.6.6. Good practice: always check the liquid feed distribution with water before putting on line. Trouble shooting: “Boilup < design”: [liquid maldistribution]*. [Liquid maldistribution]*: tubes not vertical/ inadequate calming of feed/ variations in weir height. Spiral plate exchanger: Trouble shooting: “Heat transfer < design”: stratification caused by faulty inlet and exit nozzle location/ baffling/ maldistribution. Plate exchanger: Good practice: put regulating and control valves on the inlet lines, never on the outlet lines, to minimize pressure in the exchanger. Never allow the exchanger to be under a vacuum. Keep temperature < 120 C; pressure < 2.5 MPa. Trouble shooting: “Leaking gaskets”: temperature too high/ temperature spike/ pressure too high/ cold fluid stopped but hot fluid continues/ superheated steam/ under vacuum. Air cooled: Good practice: induced draft preferred to forced draft to minimize hot-gas recirculation. Include a water-cooled “trim cooler”. Ensure the exit tubes are “flooded” so that the vapor doesn’t bypass condenser. If extreme cold conditions are expected, allow for fan to operate in reverse to counteract the overcooling by the natural circulation of cold air. Trouble shooting: “Insufficient condensation”: instrument fault/ maldistribution along either feed or exit headers/ buildup of non-condensibles in bottom tube rows/insufficient area/ ambient temperature too high/ fan not working/ blades wrong pitch/ baffles stuck/ [ fouled tubes]*/ hot-gas recirculation/ tubes not sealed. “Cycling”: control system/ vent for syphon-break is missing on exit manifold. “Outlet temperature on tube-side is high”: undersized/ tube [ fouling]* on inside or outside/ flow maldistribution on process or air side/ hot air recirculation/ air flowrate too low. “Dp on process side high”: [ fouled]* tube side/ increased liquid viscosity/ overcooling/ vapor not condensed. “Exit air temperature > expected”: low air flowrate/ flow maldistribution on tube side/ ambient air temperature > expected/ unexpected hot air recirculation. “Exit air temperature < expected”: high air flowrate/ flow maldistribution on tube side/ ambient air temperature < expected. “Sluggish control”: the use of fan pitch variation as the control variable. [Fouling]*: see Section 3.11. 3.3.4

Thermal Energy: Refrigeration

Trouble shooting: use the p-H diagram for the refrigerant as a basis for trouble shooting. “Compressor discharge pressure < design”: turbine drive problem, power limited/ overloaded centrifugal compressor or valve problem for reciprocating compressor/ wrong composition for the speed/ not enough refrigerant/ compressor fault, see Section 3.2.1. “Compressor discharge pressure > design:” fouled condenser/ insufficient air to the cooling tower/ low flowrate of water to the condenser/ air in the refrigerant/ too much refrigerant, level too high. “Compressor discharge pressure > design and condensing temperature normal”: poor drainage from the condenser/ non condensibles in the refrigerant/ refrigerant letdown valve plugged. “Suction pressure

65

66

3 Rules of Thumb for Trouble Shooting

< design”: process feedrate < design/ low circulation of refrigerant/ not enough refrigerant in chiller, level too low/ leak causing a loss of refrigerant/ compressor problems, see Section 3.2.1. “Suction pressure > design”: process coolant load > design/ throttle valve incorrectly adjusted/ low level refrigerant/ compressor problems, see Section 3.2.1. “Process exit temperature > design, refrigerant temperature from the chiller > design and the approach temperature = design”: chiller pressure > design/ not enough refrigerant/ heavy ends impurities in refrigerant/ level of refrigerant in chiller < design/ expansion valve plugged/ restriction in refrigerant suction line/ unit too small/ process fluid velocity too slow/ too much refrigerant in chiller causing flashing in the compressor suction. “Process exit temperature > design, refrigerant temperature in chiller = design and approach temperature > design”: fouling on refrigerant side/ fouling on the process side/ chiller unit too small. “Condenser temperature > design”: [ fouled]* condenser [Fouling]*: see Section 3.11. 3.3.5

Thermal Energy: Steam Generation

See thermal energy furnaces/ boilers, Section 3.3.2. See Section 3.2.5 for steam distribution Trouble shooting steam generation: “Tube failure”: feed water contains impurities/ for forced circulation: water circulation rate too low/ dry spots in tubes/ vibration induced tube failure. “Wet steam”: rate of steam generation > design causing inadequate demisting in the steam drum. “Steam production < design”: fuel-gas pressure < design/ soot in flue-gas passages/ thermostat incorrect and burners cut out too soon/ wrong type of fuel-gas burner. “Stack temperature > design”: not enough air/ overfiring. 3.3.6

High-Temperature Heat-Transfer Fluids

Good practice: usually a portion of the liquid is purged and replaced with fresh makeup. Trouble shooting: “Rapid cycling of the furnace or heating elements”: [ fluid velocity low]*. “Vapor pressure increased”: [thermal cracking of fluid]*. “Noisy pump”: contaminants such as water/ [ fluid velocity low]*/ [thermal cracking of fluid]*. “Pump discharge pressure fluctuates”: contaminants such as water/ [ fluid velocity low]*/ [thermal cracking of fluid]*. “Startup of cold unit takes longer than usual”: [oxidation of fluid]*. “Heater cannot achieve setpoint”: [oxidation]*/ [thermal cracking]*. “Poor control”: [control valve plugged]*/ [heater cannot achieve setpoint]*/ control design faulty/ controller not well tuned. [Control valve plugged]*: [thermal cracking]*/ [oxidation]*/ filter plugged/ filter missing/ filter not working. [Fluid velocity low]*: pump problems, see Section 3.2.3/ filter plugged/ controller not well tuned/ wrong location for filter/ crud left in the lines during maintenance.

3.4 Homogeneous Separation

[Oxidation]*: temperature of air in expansion tank > 60 C/ for higher temperatures in expansion tank, dry inert gas blanket not used in the expansion tank. [Thermal cracking]*: fluid velocity in the furnace or heater < design. See also trouble-shooting suggestions related to gas-liquid separators, Section 3.5.1, furnaces, Section 3.3.2, and pumps, Section 3.2.3.

3.4

Homogeneous Separation

The fundamentals upon which most of these processes are based include: mass is conserved; mass transfers because of bulk movement and diffusion. The rate of mass transfer is proportional to the concentration driving force of the target species and the surface area across which the transfer occurs. Phase equilibrium is a useful starting approximation but usually it is the rate at which the system moves toward equilibrium that is important. Surface phenomena effects, especially foaming and fouling, wetting and dispersed phase stability are issues to consider. In this section we consider the separation of species contained in a homogeneous phase, such as a liquid or gas. The separation is based on exploiting a fundamental difference that exists between the species. Methods that exploit differences in vapor pressures are evaporation, in Section 3.4.1 and distillation, in Section 3.4.2. Methods exploiting solubility are solution crystallization, Section 3.4.3; absorption, Section 3.4.4, and desorption, Section 3.4.5. Solvent extraction, Section 3.4.6, exploits differences in partition coefficient. Methods based on exchange equilibrium and molecular geometry include adsorption of species from a gas, Section 3.4.7, and of species from a liquid, Section 3.4.8. Ion exchange, Section 3.4.9, exploits differences in surface activity and exchange equilibrium. Membrane separations described include reverse osmosis, Section 3.4.10; nanofiltration, Section 3.4.11; and ultra and micro-filtration, Section 3.4.12. Separation of larger sized species are considered “heterogeneous systems” and are considered in Section 3.5. 3.4.1

Evaporation

Good practice: keep the pressure drop between the last effect and the inlet to the vacuum device < 3 kPa. Consider vapor recompression for conventional low DT evaporators such as falling film, forced circulation and horizontal tube falling film. Vapor recompression is rarely used on high DT systems such as rising film, calandria and submerged tubes. Trouble shooting: “Product contamination”: leaking valves/ crud left in storage tanks/ crud left in dead legs in piping/ [corrosion]* products/ unexpected chemical reactions/ sampling fault/ analysis fault/ unexpected solubility effects. [Corrosion]*: see Section 3.1.2.

67

68

3 Rules of Thumb for Trouble Shooting

Vapor recompression evaporators: “Evaporation rate < design”: [ fouled]* heat-transfer surface/ uneven movement of liquid over heat-transfer surface/ feed property changes/ excessive noncondensibles from leaks or present in feed/ flooded condensate, trap malfunction, Section 3.5.1/ feed temperature < design/ water leakage into the system/ lower compressor suction pressure, see also Section 3.2.1.”Steam economy low”: instrument fault/ excessive venting especially the first, second and third effects/ vapor exiting through condensate, trap problems, Section 3.5.1 / vapor blowing into product flash tank through the liquor lines/ [ foaming]*/ internal afterheaters leaking/ afterheater scaled so that liquor from the colder effect is not correctly preheated for the next effect/ [entrainment]*/ excessive vacuum/ [ fouling]*. “Recovery-boiler efficiency low”: [ fouling]*. “Vibration”: vapor velocity high through the first row of tubes. “Vacuum problems”: see vacuum, Section 3.2.2. [Corrosion]*: see Section 3.1.2. [Entrainment]*: poor design of deflector/ liquid level above the tubes/ [ foaming]*. [Foaming]*: see Section 3.11. [Fouling]*: sodium suflate precipitates especially in the first effect/ lignin precipitates especially in the first and second effect/ vapor sulfurization and condensation in third and fourth effects/ velocity too small. Falling-film evaporator: Trouble shooting. “Evaporation rate < design”: [liquid maldistribution]*/ steam trap malfunction, see Section 3.5.1/ steam flowrate too small/ [ foaming]*/ [ fouling]*. [Foaming]*: see Section 3.11. [Fouling]*: tubular velocity too small: for 5-cm diameter tubes, recommended velocities are: for viscous liquids use 3 m/s; for the finishing effect, 2–2.7 m/s; for the intermediate effects, 1.5–1.8 m/s; for the initial effects, 1.2–1.5 m/s/ pump problems, see Section 3.2.3. [Liquid maldistribution]*: not vertical/ inadequate calming of feed/ variations in weir height. Forced-circulation evaporator: Trouble shooting: usual problems are fouling/scaling and high liquid viscosity. [Fouling]*: tubular velocity too small: for 5-cm diameter tubes, recommended velocities are: for viscous liquids use 3 m/s; for the finishing effect, 2–2.7 m/s; for the intermediate effects, 1.5–1.8 m/s; for the initial effects, 1.2–1.5 m/s/ pump problems, see Section 3.2.3. For a more general consideration of fouling see Section 3.11. [Liquid maldistribution]*: not vertical/ inadequate calming of feed/ variations in weir height. Multiple-effect evaporator: Good practice: capacity of one or more effects in series is proportional to (condensing temperature of the steam supplied – temperature of the liquid boiling in the last effect) and the overall heat-transfer coefficient. If foaming occurs, reduce the liquid level in the effect. Trouble shooting: “Reduced flowrate from last stage to maintain target strength”: water temperature to contact condenser too high/ insufficient condensing area/ [decreased UA]*/ [ foaming]*. “DT higher than usual before stage “x” and DT lower than usual after stage “x””: [decreased UA in stage “x”/ [ foaming]*. “Steam usage higher than normal”: steam leak into an effect/ bleed

3.4 Homogeneous Separation

rate too high/ poor trap performance, see 3.5.1. “Cycling (30 s–several minutes duration) steam flow, cycling pressure on the process side and, for columns, cycling Dp and cycling level in bottoms”: instrument fault/ condensate in instrument sensing lines/ surging/ [ foaming]* in kettle and thermosyphon/ liquid maldistribution/ steamtrap problems, see 3.5.1, with orifice Dp across trap < design/ temperature sensor at the feed zone in a distillation column/ collapsed tray in a distillation column/ unsteady vacuum see Section 3.3.2. [Decreased UA]*: inadequate condensate removal/ liquid level too low in the effect/ [ fouling]*/ inadequate removal of non-condensible gas. [Foaming]*: natural occurring surfactants/ pH far from the zpc/ naturally occurring polymers/ solids particles/ corrosion particles/ mechanical foam breaker not rotating/ baffle foam breaker incorrectly designed or damaged/ antifoam ineffective (wrong type or incorrect rate of addition)/ gas velocity too high/ rate of evaporation too fast/ overhead disengaging space insufficient height/ liquid downflow over foam too low, see also Section 3.11. [Fouling]*: tubular velocity too small: for 5-cm diameter tubes, recommended velocities are: for viscous liquids use 3 m/s; for the finishing effect, 2–2.7 m/s; for the intermediate effects, 1.5–1.8 m/s; for the initial effects, 1.2–1.5 m/s. 3.4.2

Distillation

Good practice: for trays, add 10% more trays or two trays to improve operability. Weir height: 5 cm with length 75% of the tray diameter to provide a liquid weir overflow rate > 5 and < 20 L/s m of weir into the downcomer. Usually use 15 L/s m. For lower flows use a picket weir. Overall downcomer area should be > 5% total tray area. For foaming liquids increase downcomer area by 50%. The downcomer exit should be at least 1.2 cm below the top edge of the outlet weir. Include four, 6-mm diameter weep holes in each tray for shutdown drainage. For packing, water test the liquid distributor for good liquid distribution before startup. [Surface tension negative]*: If the surface tension of the distillate > surface tension of the bottoms (surface tension negative) prefer the use of trays to packings to minimize potential for liquid film breakup. [Surface tension positive]*: If the surface tension of the distillate < surface tension of the bottoms (surface tension positive), the foam above trays might be unexpectedly stable. Trouble shooting: The relationship between the symptom and the causes partly depends on the control system used. Check the auxiliaries to see if they are at fault: reboilers and condensers, see Section 3.3.3; vacuum, see Section 3.2.2; pumps, see Section 3.2.3. For packed towers, 80% of the causes are liquid maldistribution. “Dp across the column » design (> 12 the column height), reflux flowrate » usual; DT across column < design, overhead composition contains heavies > design; surges in the liquid overhead, bottoms level low or fluctuates, bottoms pressure > design, higher column pressure and higher temperature profile below the flooded portion of the column the temper-

69

70

3 Rules of Thumb for Trouble Shooting

ature profile > design and all trays below the flood are dry and bottoms composition off spec”: [ jet flooding]*. “Dp across the column » design, reflux flowrate gradually increasing; DT across column < design, overhead composition contains heavies > design; bottoms level low or fluctuates, bottoms pressure > design, and higher temperature profile below the flooded portion of the column the temperature profile > design and “all trays below the flood are dry””: [downcomer flooding]*. “Dp across the column > design”: instrument fault/ high boilup rate/ steam flow to reboiler > design. “Dp across the column < design”: instrument fault/ [low boilup rate]*, see Section 3.3/ dry trays/ low feedrate/ feed temperature too high. “Feed flowrate < design”: instrument fault/ pump problems, see Section 3.2.3/ filter plugged/ column pressure > design/ feed location higher than design. “Temperature of feed > design”: instrument fault/ preheater fouled/ feed flowrate low/ heating medium temperature < design, see heaters, Section 3.3.3. “Temperature of bottoms < design”: instrument fault/ [low boilup]* see Section 3.3.3/ loss of heating medium/ steam trap plugged, see 3.5.1/ feedrate to column > design/ feed concentration of low boilers (overheads) > design/ feed distributor fouled. “Temperature of bottoms > design”:instrument fault/ [column pressure > design]*/ high boilup/ overhead condenser vent plugged/ insufficient condensing, see Section 3.3.3. “Temperature at top > design”: instrument fault/ bottom temperature > design/ reflux too low/ distillate feed forward too high/ column pressure high/ [ flooding]*. “Temperature at top > design and overhead composition contaminated with too many heavies”: vapor bypassing caused by excessive vapor velocities (high boilup) or not enough liquid on tray or packing, or downcomers not sealed, or sieve holes corroded larger than design and tray weeps/ reflux too low/ feed contains excessive heavies. “Temperature at top < design”: instrument fault/ control temperature too low/ [low boilup]* see Section 3.3.3. “All temperatures falling simultaneously”: [low boilup]*. “All temperatures rising simultaneously:” pressure rising. “Overhead off spec”: poor tray or packing efficiency/ [maldistribution]*/ not enough trays or packing/ loss of efficiency/ high concentration of non-condensibles/ missing tray/ collapsed tray/ liquid entrainment/ liquid bypass and weeping/ liquid or gas maldistribution. “Overhead contaminated with heavies and excessive reflux rate and high boilup rate”: inadequate gas-liquid contact/ insufficient liquid disengagement from vapor/ presence of non-condensibles in feed. “Overhead and bottoms off spec and decreases across column in both DT and Dp “: [dry trays]*. “Overhead and bottoms off spec”: bypass open on reflux control valve. “Overhead and bottoms off spec, decrease in DT across column and perhaps Dp increase and cycling of liquid in the bottoms”: [damaged tray]*. “Distillation overhead off spec”: excessive inerts from upstream/buildup of trace/ purge not sufficient from recycle. “Separation performance of column decreases”: trace amounts of water/ trace amounts of water trapped in column/ [bumping resulting in plate damage]*. “Level of bottoms > design”: bottoms pump failure, see Section 3.2.3/ bottoms line plugged. and see implications for reboiler, Section 3.3.3. “Level in bottoms > design” and [column pressure > design]*/ high boilup/ overhead condenser vent plugged. “Level in bottoms > design and pressure increase in kettle reboiler and surges”: [ foaming]*, inerts/ leaks in kettle reboiler/ undersized reboiler. See also Section 3.3.3. “Bottoms off spec”:

3.4 Homogeneous Separation

loss of tray efficiency/ contamination of bottoms from pump, (from light oil lubricant in bottoms pump or forced circulation reboiler)/ transient vapor puff from horizontal thermosyphon reboiler, see Section 3.3.3. “Distillate flow too low”: feedrate low/ feed composition of overhead species low/ [low boilup]*/ reflux too high/ overhead control temperature too low. “Distillate flow too high:” feedrate high/ feed composition of overhead species high/ reflux ratio too low. “Bottoms and overhead flowrates < design”: [ flooding]*/ excessive entrainment/ [ foaming]*/ excessive Dp but not flooded/ plugging and fouling/ [maldistribution]*. “Water hammer in column”: process fluid above the tube sheet of a thermosyphon reboiler. “Cycling of column temperatures:” controller fault. “Product contamination”: leaking valves/ crud left in storage tanks/ crud left in dead legs in piping/ corrosion products/ unexpected chemical reactions/ sampling fault/ analysis fault/ unexpected solubility effects. “Cycling (30 s–several minutes duration) steam flow, cycling pressure on the process side and, for columns, cycling Dp and cycling level in bottoms”: instrument fault/ condensate in instrument sensing lines/ surging/ [ foaming]* in kettle and thermosyphon/ liquid maldistribution/ steam-trap problems, see Section 3.5.1, with orifice Dp across trap < design/ temperature sensor at the feed zone in a distillation column/ collapsed tray in a distillation column. [Bumping resulting in plate damage]*: trace amounts of water. [Column pressure > design]*: [high boilup]*/ overhead condenser vent plugged. [Damaged trays]*: leak of water into high molar mass process fluid/ large slugs of water from leaking condensers or steam reboilers/ startup with level in bottoms > design/ attempt to overcome flooding by pumping out bottoms at high rate/ too rapid a depressurization of column/ unexpected change in phase. [Downcomer flooding]*: excessive liquid load/ restrictions/ inward leaking of vapor into downcomer/ wrong feed introduction/ poor design of downcomers on bottom trays/ unsealed downcomers/ [ foaming]*. [Dry trays}*: flooded above/ insufficient reflux/ low feedrate/ high boilup / feed temperature too high. [Foaming]*: surfactants present/ surface tension positive system/ operating too close to the critical temperature and pressure of the species/ dirt and corrosion solids/ natural occurring surfactants/ pH far from the zpc/ naturally occurring polymers/ solids particles/ corrosion particles/ antifoam ineffective (wrong type or incorrect rate of addition/ gas velocity too high/ vapor velocity too high/ tray spacing too small/ asphaltenes present. A more generic listing of the causes of foaming is given in Section 3.11. [Jet flooding]*: excess loading/ fouled trays/ plugged holes in tray/ restricted transfer area/ poor vapor distribution/ wrong introduction of feed fluid/ [ foaming]*/ feed temperature too low/ high boilup/ entrainment of liquid because of excessive vapor velocity through the trays/water in a hydrocarbon column. [High boilup]*: see Section 3.3.3. [Low boilup]*: see Section 3.3.3. [Maldistribution]*: weirs not level/ low liquid load/ backmixing/ faulty design. [Premature flooding]*: internal damage/[ fouling]*/ change in feed composition or temperature/ unexpected entrainment/ [ foaming]*/ incorrect design for downco-

71

72

3 Rules of Thumb for Trouble Shooting

mers/ unstable control system/ level control problems/ instrument error/ second liquid phase in the column. 3.4.3

Solution Crystallization

Good practice: to prevent plugging, avoid having natural sumps for suspension type crystallizers. Control the degree of supersaturation. For crystallizers operated with cooling or evaporative crystallization, the supersaturation occurs near the heat exhange surface. For antisolvent or reaction crystallizers, the key control of supersaturation is local (often the mixing). Differentiate among the different types of product or impurities to solve problems: surface contamination, agglomeration traps impurities, inclusions, polymorphism. Check that impurities are soluble for the end point conditions of crystal growth= condition of separation. Trouble shooting: base approach on mass and energy balances, population or number balances. Follow the population density of number versus size. Must know the type of crystals and the mode of operation. “Yield < design”: initial concentration < design. “Impure product because of surface contamination”: poor solid-liquid separation/ poor washing, see Sections 3.5.12, 3.5.13, 3.5.14. “Impure product because of agglomeration trapped impurities”: wrong pH/ wrong magma electrolyte concentration/ wrong mixing. “Impure product because of inclusions”: supersaturation driving force too large. “Impure product because of polymorphism”: change in crystal habit during the crystallization process/ isoelectric point/ mixing problem. “Crystal habit (shape and aspect ratio) differs from specs”: wrong temperature during growth/ impurities especially surfactants/ supersaturation level too high. “Size distribution > design”: supersaturation too close to metastable limit. “Filtration rate slow”: crystals too small/ large size distribution/ fault with filtration, see Sections 3.5.12 and 3.5.13. “Incrustation, fouling, deposits”: cold spots/ missing insulation/ low suspension density/ protrusions and rough areas on the process surface/ local supersaturation too high/ cooling surfaces too cold. “Product contamination”: leaking valves/ crud left in storage tanks/ crud left in dead legs in piping/ [corrosion]* products/ unexpected chemical reactions/ sampling fault/ analysis fault/ unexpected solubility effects. [Corrosion]* see Section 3.1.2. Vacuum and circulating systems: Trouble shooting: “Crystal size too small”: low suspension density/ high circulation rate/ solids in feed causing nucleation sites/ feed flowrate > design/ excessive turbulence/ local cold spots/ subsurface boiling/ supersaturation too high or too close to the metastable limit. “Insufficient vacuum”: see Section 3.2.2/ obstruction in vapor system/ insufficient cooling water to condenser/ temperature of the cooling water to the condensers > design/ air leaks. For steam ejectors: steam pressure < design. For mechanical vacuum pumps: seal water flowrate < design/ rpm < design. “Liquid level in crystallizer fluctuates wildly”: check out the vacuum system, see Section 3.2.2/ low steam pressure to the steam ejectors/ fluctuation in the flow of cooling water to the condensers. “Circulation rate differs from design”: see pumps, Section 3.2.3. [Foaming]*: air leaks in pump packing/ air in

3.4 Homogeneous Separation

feed/ air leak in flanges or valve stems. A generic listing of the causes of foaming is given in Section 3.11.

3.4.4

Gas Absorption

Good practice: the more selective the absorbent the more difficult it is to regenerate the absorbent. Prefer the use of low holdup internals. Select materials of construction to promote wetting: select critical surface tension of the solid is > the surface tension of the liquid. If the surface tension of the feed liquid > 2 mN/m larger than the surface tension of the bottom exit liquid or the absorption of the solute lowers the surface tension (surface tension negative) prefer the use of trays to packings to minimize potential for liquid film breakup. If the surface tension of the feed liquid > 2 mN/m smaller than surface tension of the bottom exit liquid (surface tension positive), the foam above trays might be unexpectedly stable; stable films on packing. Trouble shooting: for multitube cocurrent falling film absorber: “Concentration of product acid < design, inadequate absorption”: liquid maldistribution/ gas maldistribution. “Low heat-transfer coefficient”: liquid or gas maldistribution. “Hydraulic instability”: no vent break on the syphon. “Product contamination”: leaking valves/ crud left in storage tanks/ crud left in dead legs in piping/ corrosion products/ unexpected chemical reactions/ sampling fault/ analysis fault/ unexpected solubility effects. For amine absorption of sour gas, Good practice: keep inlet amine solvent temperature at least 5 C hotter than inlet gas temperature to minimize condensation of volatile hydrocarbons in the inlet gas stream. Trouble shooting: “Insufficient absorption or off-specification for exit scrubbed gas”: feed gas concentration off spec/ feed gas temperature or pressure outside operating window: for amine absorbers: > 50 C for H2S and < 24 C for CO2/ feed gas pressure has decreased/ [solvent flowrate too low]*; for glycol dehydration: 12.5 to 25 L TEG per kg water removed / [solvent incorrect]* / incorrect feed try location/ [column operation faulty]*/ absorber operating conditions differ from design/ [absorber malfunction]*. “Dp across absorber > design”: gas flowrate > design/ pressure < design/ [ foaming]*/ plugged trays/ plugged demister pads/ collapsed tray or packing. “Dp on column fluctuating”: [ foaming]*. Solvent carryover from the top of the column”: [ foaming]*. “Liquid level in vessels fluctuates”: [ foaming in column]*. “Change in absorption rate”: for amine absorption: decrease in removal of H2S and increase in removal of CO2/ [ foaming]*. “Overloaded liquid in downstream gaseous processing equipment”: [ foaming in absorber]*. “Solvent losses high”: [physical losses]*/ [entrainment]*/ [solubility]*/ [vaporization]*/ [degradation]*/ [loss elsewhere]*/ for glycol dehydration typical losses = 0.015 mL/ m3 gas treated. [Amine concentration too high or too low]*: if too high, lack of equilibrium driving force/ if too low, insufficient moles of amine for the feed concentrations. [Column operation faulty]*: plugged tray or packing/ poor distribution for packing/liquid flowrate < minimum required for loading/ [gas velocity too high]*/ col-

73

74

3 Rules of Thumb for Trouble Shooting

lapsed trays or packing/ plugged or broken distributors/ [ foaming]*/ solvent–stripper overhead temperature too low. see also Section 3.4.2 [Corrosion]*: see Section 3.1.2. [Degradation]*: chemical reaction; for amine: reacts with CO2 and O2; forms stable salts: for glycol: reacts with O2/ thermal decomposition; for amine: surface temperatures > 175 C ; for glycol: surface temperatures > 205 C. [Entrainment: GL]*: demister plugged, missing, collapsed, incorrectly designed/ [ flooding]*/ [ foaming]*/ inlet liquid line or distributor undersized of plugged/ poor distribution for packing/liquid flowrate < minimum required for loading/ [gas velocity too high]*/ solvent feed temperature > specifications/ [column operation faulty]*/ tray spacing < design. see also GL separators Section 3.5.1 [Entrainment: L-L]*: fluid velocity too high; example > 10 L/s m2/ liquid distributor orifice velocity > design; for amine: for amine > 0.8 m/s; for hydrocarbon > 0.4 m/s/ faulty location of exit nozzles/ interface level wrong location/ faulty control of interface/ no vortex breaker/ exit fluid velocities > design/ insufficient residence time/ [stable emulsion formation]*. see also decanters, Section 3.5.3. [Foaming]*: [ foam-promoting contaminants]*/ [gas velocity too high]*/ [liquid residence time too low in GL separator]*/ antifoam addition faulty/ faulty mechanical foam breaker/ [liquid environment wrong]*. A generic listing of causes for foaming is given in Section 3.11. [Foam promoting contaminants: soluble]*: naturally occurring or synthetic polymers/ naturally occurring or synthetic organics >C10; example lube oils/ naturally occurring or synthetic surfactants; for amine systems: the surface active contaminants include condensed hydrocarbons, organic acids, water contaminants, amine-degradation products/ faulty cleaning before startup; surfactants left in vessels. [Foam promoting contaminants: solid]*: [corrosion products]*; for amine systems: iron sulfides; amine salts formed from organic acids + hydrocarbons/ faulty cleanup before startup; rust left in vessel/ dust/ dirt/ particulates. [Flooding]*: see Section 3.4.2. [Gas velocity too high]*: vessel diameter too small for gas flow/ column pressure < design/ trays or packing damaged or plugged giving excessive vapor velocity/ temperature too high/ upstream flash separator passing liquids: feed contaminated with excessive volatile species/ stripping gas fed to column too high/ flowmeter error/ design error. [Liquid environment wrong]*: pH far from the zpc/ electrolyte concentration too low. [Physical losses]*: leak to atmosphere/ purges for sampling/ sampling/ heat exchanger leak/ pump seal flushes/ filter changes/ piping, fitting, valve stems, gaskets, pumps. [Solubility losses]*: liquid-liquid systems: system pressure < design/ for amine: concentrations > 40% w/w/ system temperatures too high. [Solvent contaminated]*: carryover from upstream equipment; example oil from compressor; brines, corrosion inhibitors, sand, [corrosion products]* / oxygen leaks into storage tank/ inadequate corrosion control, example low pH causing corrosion/ degradation via overheating, ex-hot spots in reboiler tubes or fire tubes/ ineffective

3.4 Homogeneous Separation

filters/ ineffective cleaning before startup/ for amine absorbers: corrosion products/ FeS/ chemicals used to treat well [Solvent feed temperature too high]*: fouled exchanger/ undersized heat exchanger/ ambient temperature too high. [Solvent flowrate too low]*: flowmeter or sensor error/ absorber pressure > design/ plugged strainer, lines of filters/ low liquid level in pump feed tank/ [cavitation]*/ air-locked pump and see Section 3.2.3 for trouble shooting pumps. [Solvent incorrect]*: incorrect concentration of active ingredient: for amine absorbers: [amine concentration too high or too low]*; for glycol dehydration: solvent concentration TEG < specifications / [solvent stripping inadequate]*/ [solvent feed temperature too high]*/ [solvent contaminated]*. [Solvent loss elsewhere]*: upstream units, for example for glycol dehydration: glycol dumped with hydrocarbons separated in upstream flash drum/ loss in downstream solvent stripper. [Solvent stripping inadequate]*: not enough steam in stripper/ incorrect pressure in stripper/ [ foaming]*/ [contaminated solvent]*/ contaminated feed: for amine strippers: other sulfur species causing high partial pressure/ leak in the feed preheater contaminating feed with stripped solvent. [Vaporization losses]*: system pressure < design/ for amine: concentrations > 40% w/w/ system temperatures too high. 3.4.5

Gas Desorption/Stripping

Trouble shooting: “Solvent or stripped liquid concentration > design”: boilup rate or steam stripping rate too low/ feed concentration > expected/ feed contamination; for sour-water stripper: acid in feed may be chemically bonded with NH3 and prevent adequate stripping of NH3/ [ foaming]*/leak in preheater exchanger/ [column malfunction]*. “Overhead from stripper < specifications”: insufficient flowrate of stripping gas/ for glycol dehydration: reboiler temperature < 175–200 C or reboiler too small for required duty or fouling of reboiler tubes/ [ foaming]*/ dirty or broken packing or plates/ [ fouled or scaled internals]*/ [ flooding]*/ top pressure > design/ leak in preheater exchanger/ [ feed concentration off specification]*. “Overhead temperature on stripper > design”: reflux flowrate too low/ [ flooded]*/ [ foaming]*/ feed contaminated with light hydrocarbons. “For sour-water strippers or glycol dehydration: Pressure at reboiler > design”: instrument error/ top pressure > design/ [Dp across column > design]*/ overhead line plugged/ [ flooding]*/ for stripper for glycol dehydration: slug of hydrocarbon in feed is flash vaporized at reboiler and blows liquid out of stripper. “For sour-water strippers: odor or H2S problems at the storage tank”: 0.6 to 1 m layer of oil on top of water missing/ oil layer exceeds 0.6 to 1 m depth/ faulty inert gas operation. “Plugging of overhead system”: top temperature not within the operating window; for sour-water strippers: temperature < 82 C at which ammonium polysulfides form but temperatures too high give excessive water in overhead vapor causing problems for downstream operation / overhead lines not insulated/ insufficient steam tracing

75

76

3 Rules of Thumb for Trouble Shooting

on overhead vapor lines. “Feed flowrate and composition to the stripper varies”: [instrument error]*/ sampling error/ analysis error/ [ faulty separation in flash drum]*/ [ foaming in upstream absorber]*/ no intermediate storage tank between the flash drum and the stripper/ storage tank faulty operation or design: for SWS: residence time < 3 to 5 days; stratification occurs, bypassing occurs, insufficient mixing in tank/ oil layer on top of water in storage tank exceeds 0.6 to 1 m depth. [Column malfunction]*: [ feed concentration off specification]*/ excessive stripping gas or steam velocity/ too much cooling or condensation/ top temperature > design/ insufficient reflux cooling/ packing broken, damaged/ [ fouled or scaled internals]*/ [ foaming]*/ [ flooding]*, see also Section 3.4.2. [Feed concentration off specification]*: [ foaming in upstream absorber]*/ for glycol dehydration: upstream flash separator passing water; for oil or hydrocarbon in feed to SWS”: residence time for sour water in flash drum is < 20 min. [Flooding]*: see Section 3.4.2. [Foaming]*: see Section 3.11. [Fouling]*: see Section 3.11 [Fouled or plugged internals]*: for SWS: cooling water leak/ pH of feed water too basic/ calcium ion concentration too high causing precipitation when temperatures in stripper exceed 122 C/ temperature < 82 C at which ammonium polysulfides form/ overhead lines not insulated. [Instrument error]*: calibration fault/ sensor broken/ sensor location faulty/ sensor corroded/ plugged instrument taps: for sour-water strippers: water or steam purge of taps malfunctioning or local temperatures < 82 C at which ammonium polysulfides form. [Dp across column > design]* see Section 3.4.2. 3.4.6

Solvent Extraction, SX

Good practice: the dispersed phase should not preferentially wet the materials of construction. If unexpected rapid coalescence occurs, suspect [Marangoni effects]* and change the dispersed phase. Treat the buildup of the “rag” at the interfaces based on the cause: corrosion products or stabilizing particulates, surfactants, or amphoteric precipitates of aluminum or iron. Consider adjusting the pH. Solid particles tend to accumulate at the liquid–liquid interface. Trouble shooting: “Poor separation”: level control fault/ phase velocities too high/ contaminant gives stable dispersion/ smaller drop size than design/ rag formation/ temperature change/ pH change/ decrease in electrolyte concentration. [Rapid coalescence]*: wrong phase is the continuous phase/ [Marangoni instabilities]*/ pH at the zpc/ high electrolyte concentration in the continuous phase. [Marangoni effects]*: non-equilibrated phases/ local mass transfer leads to local changes in surface tension and stability analysis yields stable interfacial movement. For column extractors: “Decrease in extraction efficiency”: agitator speed to fast/ excessive backmixing/ flooding.

3.4 Homogeneous Separation

[Flooding]*: agitator speed too fast/ feed sparging velocity too high/ drop diameter smaller than design. For centrifugal extractors: for bioprocessing/proteins: Good practice: the partition coefficient sensitive to pH, electrolyte type and concentration. “Product contamination”: leaking valves/ crud left in storage tanks/ crud left in dead legs in piping/ corrosion products/ unexpected chemical reactions/ sampling fault/ analysis fault/ unexpected solubility effects. Other suggestions for trouble-shooting decanters are given in Section 3.5.3. More about stable emulsion formation is given in Section 3.11. 3.4.7

Adsorption: Gas

Trouble shooting: ”Wet gas”: steam leak/ leaky valves/ inadequate regeneration/ wrong adsorbent/ adsorbent damaged by excessive regeneration temperature/ adsorption cycle too long/ [early breakthrough]*. “Dp high”: fine particulates in feed/ breakdown of adsorbent/ high gas feedrate. “Product contamination”: leaking valves/ crud left in storage tanks/ crud left in dead legs in piping/ [corrosion]* products/ unexpected chemical reactions/ sampling fault/ analysis fault/ unexpected solubility effects. [Corrosion]*: see Section 3.1.2. [Early breakthrough]*: gas short-circuiting bed/ faulty regeneration/ increased concentration in feed/ other contaminants in feed. 3.4.8

Adsorption: Liquid

Good practice: carbon regeneration by multiple hearth furnaces. For edible oils prevent contact with air. Trouble shooting: “Early breakthrough”: liquid short-circuiting bed/ faulty carbon regeneration/ increased concentration in feed/ other contaminants in feed. “Pressure drop high”: fine particulates in feed/ breakdown of carbon/ high liquid feedrate. 3.4.9

Ion Exchange

Good practice: use an upstream degasser to remove carbonic acid. Trouble shooting: usual sources of trouble are change in ions in the feed, the multiport valves improperly seat so that feed or regenerant bypass into the effluent, clogged liquid distributors, clogged underdrains; degradation of the resin and faulty backwash. Organic fouling mainly affects anionic exchangers. “Throughput capacity < design”: instrument error/ increase in feed concentration/ less resin volume than design/ regenerant concentration < design/ regenerant volume < design, 0.5–3.5 L/s L of resin/ regenerant flowrate < design/ wrong regeneration ion/ contamination of regenerant with high valence ions. “In the spring, reduced flowrate through the unit

77

78

3 Rules of Thumb for Trouble Shooting

demineralizing river water”: high concentration of particulates, clay in spring river water. “In the summer for a unit demineralizing water, throughput of the cationic exchanger decreeses, exchange capacity for calcium and magnesium decreases but the anionic exchanger is unaffected”: suspect ferric or high-valency cation present in the feed. “In the summer for a unit demineralizing river water, throughput of the anionic exchanger decreases, exchange capacity decreases but the cationic exchanger is unaffected”: suspect fertilizer runoff with phosphate, carbonic acid and high sulfate anions as contamination. “Contamination in exit liquid > design”: instrument error/ sampling error/ online too long/ faulty regeneration/ [ fouled]*/ [poisoned]*/ high feed concentration of target ions/ feed concentration of high valence co-ions. “Dp > design”: dirt in feed/ water from river in springtime/ instrument error/ temperature/ resin void volume changes/ inlet distribution system blocked/ [resin degradation]* and backwashes into inlet/ backwash rate too high/ underbed blocked. [Fouling of the resin]*: iron and high-valence ions/ oil/ mud/ polyelectrolyte/ calcium sulfate precipitate/ silica/ barium sulfate/ carbonic acid/ sulfate or phosphate/ organics/ algae and bacterial fouling. [Poison resin]*: cobalticyanide/ polythionate/ ferricyanides/ complex humic acid/ color bodies in sugar juices. [Resin degradation]*: ingress of oxidants/ free chlorine in feed/ temperature increase/ [ fouled]*/ [poisoned]*/ corrosion products/ [resin fines]*. [Resin fines]*: thermal or physical shock/ freeze/thaw. WAC: “Alkalinity leakage during exhaustion cycle”: inadequate regeneration. “Hardness leakage during exhaustion cycle”: regeneration fault with calcium sulfate precipitation (if sulfuric acid is the regenerant). SAC: “Sodium leakage”: inadequate regeneration: wrong concentration, wrong flowrate, wrong length of time. “Hardness leakage during exhaustion cycle”: regeneration fault with calcium sulfate precipitation (if sulfuric acid is the regenerant). WBA: “Mineral acid leakage”: under regenerated/ upstream SAC malfunctioning. “Sodium leakage, high pH and high conductivity”: SAC resins contaminated the bed. “Silica problems”: series regeneration with SBA with pH falling below the isoelectric point of silica in the resin bed. SBA: “Increase in silica leakage”: [resin degradation]*. “Leakage of target ions”: organic fouling. “Low pH and high conductivity”: organic fouling. “Increase rinse quantities”: organic fouling. “Low pH (5.5), increased conductivity, increase silica leakage, increase rinse volumes, loss of throughput capacity:” organic fouling. “Product contamination”: leaking valves/ crud left in storage tanks/ crud left in dead legs in piping/ corrosion products/ unexpected chemical reactions/ sampling fault/ analysis fault/ unexpected solubility effects.

3.5 Heterogeneous Separations

3.4.10

Membranes: Reverse Osmosis, RO

Good practice: consider pretreating hydrophobic membranes for aqueous use. Trouble shooting: “Permeate flow < design:” physical fouling (incorrect/incomplete pretreatment/ scaling/ biofouling)/ chemical fouling: (pH shift/ incorrect anti-scalant dosage). “Permeate quality degradation:” failure of mechanical seal/ chemical attack of membrane by pH, chlorine or biodegradation / concentration polarization / post-contamination. 3.4.11

Membranes: Nanofiltration

Trouble shooting: “Permeate flow < design:” physical fouling (incorrect/incomplete pretreatment/ scaling/ biofouling)/ chemical fouling: (pH shift/ incorrect anti-scalant dosage). “Permeate quality degradation:” failure of mechanical seal/ chemical attack of membrane by pH, chlorine or biodegradation / concentration polarization / post-contamination. 3.4.12

Membranes: Ultrafiltration, UF, and Microfiltration

Good practice: for membranes that are not hydrophobic; check the isoelectric or zero point of charge point of the species in solution compared with the charge on the membrane and consider changing the pH of operation so that the surface charges are the same. For hydrophobic membranes treating aqueous feeds, consider pretreating the membrane to make the membrane surfaces hydrophilic. Trouble shooting: “Permeate flux < design:” physical clogging (inadequate prescreening/ backwash problems/ aeration/ recirculation/ increase in influent solids loading); chemical fouling (change in water quality/ inadequate cleaning). “Permeate quality < design:” failure in mechanical seal, breakage of the membrane or hollow fibres / post contamination via regrowth/ degradation of membrane by pH or chlorine.

3.5

Heterogeneous Separations

In heterogeneous phase separation we start with at least two phases. Sections 3.5.1 to 3.5.4 address the separation of gas from liquid, gas from solid, liquid from liquid and gas from liquid, respectively. Liquid-solid separators include drying, Section 3.5.5; screens, Section 3.5.6; settlers, Section 3.5.7; hydrocyclones, Section 3.5.8; thickeners, Section 3.5.9; sedimentation centrifuges, Section 3.5.10; filtering centrifuges, Section 3.5.11; and filters, Section 3.5.12. Trouble shooting screens to separate solids is discussed in Section 3.5.13.

79

80

3 Rules of Thumb for Trouble Shooting

3.5.1

Gas–Liquid

Good practice: install a demister. In this section knockout pots and steam traps are considered. Knockout pots: Trouble shooting: “Poor separation”: [ foaming]*/ insufficient residence time/ feed and exit nozzles at wrong location/ faulty design. [Foaming]*: surfactants present/ dirt and corrosion solids/ natural occurring surfactants/ pH far from the zpc/ naturally occurring polymers/ insufficient disengaging space above the liquid/ antifoam ineffective (wrong type or incorrect rate of addition)/residence time insufficient/ designed for a vertical vessel but a horizontal vessel installed/ vapor velocity too high/ mechanical foam breaker not rotating/ baffle foam breaker incorrectly designed or damaged/ asphaltenes present/ liquid downflow velocity through the foam is too low.3) See also Section 3.11. Steam traps: Good practice: install trap below condensate exit (or with a water seal if the trap is elevated), use a strainer before most traps, use a check valve for bucket traps. Slant pipes to the trap. Use a downstream check valve for each trap discharging to a common header. Pipe diameter should be greater than or equal to the trap inlet pipe diameter. Prefer to install auxiliary trap in parallel instead of a bypass. Do not group trap thermodynamic traps because of their sensitivity to downstream conditions. Float and thermostatic: usually discharges continuously, low pitched bubbling noise. High pitch noise suggests live steam is blowing. Balanced thermostatic: leave about 0.6 m of uninsulated pipe upstream of trap. Diagnostics: when bellows placed in boiling water the expansion should be 3 mm. Inverted bucket: use initial prime to prevent steam blowing. Diagnostics sounds: when it is functioning well: loud initially, then lower pitch bubbling and then silence. Discontinuous discharge. When steam is blowing through the trap, the sound is a steady bubbling if primed with a light load or constant rattling; or continuous high pitched whistling. Diagnostic for loss of prime: close outlet valve for several minutes, then open valve slowly and operation should return to normal. If this fails then check seat and valve. Thermodynamic: about 6 cycles/minute. Trouble shooting: the major faults are wrong trap, dirt, steam locking in the trap, group trapping, air binding and water hammer. Too large a trap gives sluggish response and wastes steam. Too small a trap gives poor drainage, backup of condensate. There is a DT across all traps. “No condensate discharge”: strainer or line plugged/ steam off/ valves plugged/ no water or steam to the trap/ trap clogged/ wrong trap selected/ worn orifice/ steam pressure too high (inverted bucket)/ orifice enlarged by erosion (bucket trap)/ incorrect Dp across the orifice (inverted bucket)/ air vent clogged (inverted bucket or thermostatic air vent on float trap)/ valve seat choked (inverted bucket)/ flabby or elongated bellows (thermostatic)/ superheated 3) Turner, J. et al., 1999, HP June p. 119.

3.5 Heterogeneous Separations

steam caused burst joints or scale (thermostatic). “Cold trap + no condensate discharge”: strainer or line plugged/ steam off/ valves plugged/ no water or steam to the trap/ trap clogged. “Hot trap + no condensate discharge”: bypass open or leaking/ trap installed at high elevation/ broken syphon/ vacuum in heater coils/ pressure too high (inverted bucket)/ orifice too large (inverted bucket)/ vent hole plugged (inverted bucket)/ defective trap parts (inverted bucket)/ clogged orifice (thermodynamic). “Live steam blowing, and inlet and exit temperatures are equal”: bypass open or leaking/ worn trap components/ scale in orifice/ valve fails to seat/ trap lost prime (inverted bucket)/ sudden drops in pressure/ [backpressure too high]* (thermodynamic)/ faulty air release (float)/ trap too large (thermodynamic). “Continuous discharge when it should be discontinuous”: trap too small/ dirt in trap/ high-pressure trap installed incorrectly for low pressure service (bucket trap)/ valve seat clogged with dirt/ excessive water in the steam/ bellow overstressed (thermostatic)/ one trap serves > one unit/ strainer clogged. “OK when discharging to the atmosphere but not when to a backpressure condensate header”: condensate line diameter too small/ wrong orifice/ interaction with other traps connected to a common header/ condensate line partially plugged/ [backpressure too high]*. “Slow and uneven heating of upstream equipment”: trap too small/ insufficient air handling capacity/ short-circuiting when units are group trapped. “Inverted bucket trap loses prime:” sudden drop in pressure/ faulty seat/ faulty valve. “Upstream process cycling”: defective float/ multiple sources of condensate to a single trap/ trap flooded from condensate header/ condensate discharged into the bottom of the condensate header/ Dp across the orifice is incorrect for the orifice (inverted bucket). [Backpressure too high and trap is hot]*: return line too small/ other traps blowing steam / obstruction in return line/ bypass open/ pressure in header too high. [Backpressure too high and trap is cold]*: obstruction in return line/ excess vacuum in return line. 3.5.2

Gas–Solid

Bag filters and dry cyclones are discussed in this section. Bag filters: Good practice: replace a complete set of bag filters annually. Install a bypass. Limit the number of parallel rows of bags on either side of the walkway to 3–4 rows for 20-cm diameter bags and 2–3 rows for 30-cm diameter bags. For cleaning, use 0.5–0.7 kPa clean, dry air with an air: cloth ratio of 2: 1 for reverse jet and 2.5: 1 for shaking. Trouble shooting: “Excessive particle emissions”: cleaning too often/ pressure used to clean it too high/ bag breaks/ gas temperature too high and particles crust on movable blowrings and tear bag. “Dp across bags > design”: faulty cleaning/ improper bag tension/ excessive moisture causing blinding/ poor air distribution/ hopper plugged, see Section 3.10.3/ gas velocity > design. “Short bag life”: excessive cleaning/ high inlet gas velocity/ fines > design/ blinding because of condensation, improper cleaning, excessive dust load or high cake density.

81

82

3 Rules of Thumb for Trouble Shooting

Dry cyclone: Trouble shooting: “Increase in catalyst losses”: [poor separation in cyclone]*. “Opaque flue gas from the vessel”: [poor separation in cyclone]*. “Particulate carryover that affects operation of downstream equipment”: [poor separation in cyclone]*. “Temperature hot spots in upstream reactor”: [maldistribution]*/ local exothermic reactions. [Attrition of the particles]*: local velocities upstream of cyclone > 60 m/s/ particle too fragile. [Change in size of particles in the feed]*: [generation of fines]*/ [coarse particles]*. [Coarse particles (diameter > design)]*: agglomeration of catalyst/ [sintered particles]*/ wrong specifications for catalyst. [Dipleg unsealed]*: solids level does not cover end of dipleg/ Dp indicator for catalyst level faulty/ Dp indicator for catalyst level OK but bed density incorrect. [Generation of fines]*: [attrition of the catalyst]*/ fines in the new catalyst. [Maldistribution]*: feed distributor poorly designed/ feed distributor plugged. [Plugged dipleg]*: spalled refractory plug/ level of catalyst in bed too high / Dp indicator for catalyst level faulty/ Dp indicator for catalyst level OK but bed density incorrect. air out periods with a lot of water of steam in vessel. [Plugged grid holes]*: foreign debris entering with fresh catalyst/ faulty grid design. [Poor separation in cyclone]*: [stuck or failed trickle valve]*/ [plugged dipleg]*/ [dipleg unsealed]*/ gas velocity into cyclone too low or too high/ faulty design of cyclone/ solids concentration in feed too high/ cyclone volute plugged/ hole in cyclone body/ pressure surges/ [change in size of particles in feed]*. [Sintered particles]*: high temperature upstream/ [temperature hot spots in the upstream reactor]*. [Stuck or failed trickle valve]*: binding of hinge rings/ angle incorrect/ wrong material/ hinged flapper plate stuck open/ flapper plate missing. 3.5.3

Liquid–Liquid

Decanters and hydrocyclones are discussed in this section. Decanter: Good practice: contamination can interfere with the operation. Traditionally this contamination is surfactants, or particulates. The particulates can be corrosion products, amphoteric precipitates of aluminum or iron. Try changing the pH of the water to alter the surface charge on the dispersed drops. The separation capacity of a settler/decanter doubles for every 20 C increase in temperature. Caution, if, to ease this separation, the temperature is increased, such an increase in temperature will increase the bulk-phase contamination because of the increased cross-contamination by the mutual solubility. Trouble shooting: “Entrained droplets in liquid effluent”: sensor error/sampling error (immiscible drops are not being entrained)/ faulty design of separator/ improper cleaning of vessel after shutdown, e.g, rust left in vessel/ pressure fluctuation/ pressure too low causing flashing/[inaccurate sensing of interface]*/ [drop doesn’t settle]*/ [drop settles and coalesces but is re-entrained]*/ [drop settles but doesn’t

3.5 Heterogeneous Separations

coalesce]*/ [stable emulsion formation]*. “Fluctuation in liquid level”: no vacuum break on syphon line for bottoms/ level sensor error/ poorly tuned controller/ surges in feed. [Coalescer pads ineffective]*: temperature too high/ pH incorrect/ fibers have the same charge as the droplets/ surface tension negative system/ wetting properties of fibers changed/ fibers “weathered” and need to be replaced/ flowrate too slow through fibers/ wrong mix of fibers/ prefiltering ineffective/ surface tension < 1 mN/m for fluoropolymer fibers or < 20 mN/m for usual fibers/ wrong design/ included in decanter but should be separate horizontal coalescer promoter unit/ faulty design. see Section 3.9.2. [Density difference decrease]*: dilution of the dense phase/ reactions that dilute the dense phase; for sulfuric acid alkylation: if acid strength < 85% w/w the olefins polymerize with subsequent oxidation of the polymers by sulfuric acid. As a self-perpetuating continuing decrease in acid strength. Alkylate-acid separation is extremely difficult when acid concentration is 40% w/w. [Drop doesn’t settle]*: [density difference decrease]*/ [viscosity of the continuous phase increases]*/ [drop size decreases]*/ [residence time for settling too short]*/ [phase inversion or wrong liquid is the continuous phase]*/ pressure too low causing flashing and bubble formation. [Drop settles and coalesces but is re-entrained]*: faulty location of exit nozzles for liquid phases/ distance between exit nozzle and interface is < 0.2 m/ overflow baffle corroded and failure/ interface level at the wrong location/ faulty control of interface/ liquid exit velocities too high/ vortex breaker missing or faulty on underflow line/ no syphon break on underflow line/ liquid exit velocities too high. [Drop settles but doesn’t coalesce]*: [phase inversion]*/ pH far from zpc/ surfactants, particulates or polymers present/ electrolyte concentration in the continuous phase < expected/ [coalescer pads ineffective]*/ [drop size decrease]*/ [secondary haze forms]*/ [stable emulsion formation]*/ [interfacial tension too low]*/ [Marangoni effect]*. [Drop size decrease]*: feed distributor plugged/ feed velocity > expected/ feed flows puncture interface/ local turbulence/ distributor orifice velocity > design; for amine units: for amine > 0.8 m/s; for hydrocarbon > 0.4 m/s/ [Marangoni effects]*/ upstream pump generates small drops/ [secondary haze forms]*/ poor design of feed distributor. [Inaccurate sensing of the interface]*: instrument fault/ plugged sight glass. [Interfacial tension too small]*: temperature too high/ [surfactants present]* at interface. [Marangoni effects]*: non-equilibrated phases/ local mass transfer leads to local changes in surface tension and stability analysis yields stable interfacial movement. [Phase inversion]*: faulty startup/ walls and internals preferentially wetted by the dispersed phase. [Rag buildup]*: collection of material at the interface: [surfactants present]* / particulates: example, products of [corrosion see Section 3.1.3]*, amphoteric precipitates of aluminum/ naturally occurring or synthetic polymers. [Residence time for settling too short]*: interface height of the continuous phase decreases/ [inaccurate sensing of interface]*/ turbulence in the continuous phase/

83

84

3 Rules of Thumb for Trouble Shooting

flowrate in continuous phase > expected; for example > 3 L/s m2 / sludge settles and reduces effective height of continuous phase/ [phase inversion]*/ inlet conditions faulty. [Secondary haze forms]*: small secondary drops are left behind when larger drop coalesces, need coalescer promoter, see Section 3.9.2. [Stable emulsion formation]*: [surfactants present]* / contamination by particulates: example, products of [corrosion products. see Section 3.1.3]*, amphoteric precipitates of aluminum or iron/ pH far from the zpc/ contamination by polymers/ temperature change/ decrease in electrolyte concentration/ the dispersed phase does not preferentially wet the materials of construction/ coalescence-promoter malfunctioning/ improper cleaning during shutdown/ [rag buildup]*. [Surfactants present]*: formed by reactions/ enter with feed, example oils, hydrocarbons >C10, asphaltenes/ left over from shutdown, example soaps and detergents/ enter with the water, example natural biological species, trace detergents. [Viscosity of the continuous phase increases]*: temperature too low, for alkylate-acid separation, temperature < 4.4 C/ [phase inversion]*/ contamination in the continuous phase/ unexpected reaction in the continuous phase causing viscosity increase. Hydrocyclones: Good practice: control on pressure drop. May be operated as open or flooded underflow. Trouble shooting: “Incorrect separation”: faulty design/ inlet pressure too low/ wrong Dp from feed to overflow/ interfacial tension < 10 mN/m/ wrong volume split/ feed drop size too small. 3.5.4

Gas–Liquid–Liquid Separators

Horizontal drum: Good practice: separates gas, oil and water; as for example as an early separation of natural gas upstream of drying or to handle sour water. Typically, a relatively small load of hydrocarbon. Often called a “flash drum”. Often follow the flash drum with a storage tank to allow further separation of water and hydrocarbon. Contamination from naturally occurring or synthetic surfactants or polymers, or corrosion products from upstream processing can cause stable foam or emulsion formation. Trouble shooting: “Entrained liquid in overhead gas”: sensor error/ [entrainment: GL]*. “Incomplete separation of oil from water”: faulty design of separator/ residence time of liquid phases too short/ liquid velocity in the decant phases too fast/ [Marangoni instabilities]*/ liquid feed velocity too high/ poor distribution of liquid feeds/ faulty location of exit nozzles for liquid phases/ overflow baffle corroded and failure/ interface level at the wrong location/ faulty control of interface/ no vortex breaker at water and heavy oil exit nozzles/ liquid exit velocities too high/ [emulsification]*/ contaminant gives stable dispersion/ smaller drop size than design/ rag formation/ temperature change/ pH change/ decrease in electrolyte concentration. See Sections 3.5.1 and 3.5.3 for more details. [Entrainment: GL]*: vessel diameter too small for gas flow/ no demister or demister malfunctioning/ vessel pressure < design/ [ foaming]*/ inlet liquid line or distributor undersized or plugged.

3.5 Heterogeneous Separations

[Entrainment: L-L]*: liquid velocity too high; example > 10 L/s m2/ liquid distributor orifice velocity > design; for amine: for amine > 0.8 m/s; for hydrocarbon > 0.4 m/s / faulty location of exit nozzles/ interface level wrong location/ faulty control of interface/ no vortex breaker/ exit fluid velocities > design/ insufficient residence time/ [stable emulsion formation]*. [Foaming]*: see Section 3.11 for generic causes. [Marangoni instabilities]*: non-equilibrated phases/ local mass transfer leads to local changes in surface tension and stability analysis yields stable interfacial movement. [Stable emulsion formation]*: see Section 3.11 for generic causes; Section 3.5.3 for more specific causes. The dispersed phase should not preferentially wet the materials of construction. If unexpected rapid coalescence occurs, suspect Marangoni effects and change the dispersed phase. Treat the buildup of the “rag” at the interfaces based on the cause: corrosion products or stabilizing particulates, surfactants, or amphoteric precipitates of aluminum or iron. Consider adjusting the pH. Solid particles tend to accumulate at the liquid–liquid interface. 3.5.5

Dryer for GS Separation

Trouble shooting: work with an overall mass and energy balance. For continuous rotary steam-tube dryer: “Product moisture content high”: insufficient steam flow/ rotational speed too fast/ insufficient area/ particles clump/ moisture content of the feed too high/ faulty design of dryer/ flights damaged/ incorrect angle of inclination/ upstream batch centrifuge gives periodic wet cake. For fluidized-bed dryer: “Product moisture content high”: solids buildup on gas sparger in fluidized bed (caused because inlet gas temperature too high). For spray dryer: “product wet and clumps form inside spray dryer”/ insufficient gas flow/ inlet gas temperature too low, instrument fault/ feed solids concentration lower than design/ liquid drops larger than design. For fixed bed-hopper to dry polymer feed for extruder (hot gas < 120 C): “Feed material not dry”: incorrect drying temperature/ solids throughput > design/ instrument error/ input air too moist/ ambient air leaking into drying air circuit/ adsorbent for drying air incorrectly regenerated/ air dryer (gas adsorber) fault. See related unit adsorption: gas, Section 3.4.4. Hopper dryer for polymer feed to extruder. “Polymer pellets leaving hopper are not dry”: incorrect drying temperature/solids throughput > design/ instrument error/ input drying air too moist/ ambient air leaking into drying air circuit/ adsorbent for drying air incorrectly regenerated/ air dryer (adsorber) fault see Adsorption: gas, Section 3.4.7. 3.5.6

Screens for Liquid Solid Separation or Dewatering

Batch Screen pack downstream of extruder: Good practice: install standby screen pack with diverter valve to bring standby on line when on–line filter blinds. Trouble

85

86

3 Rules of Thumb for Trouble Shooting

shooting: “Solid contamination of product”: mesh size too large/ contamination downstream of screen pack. “Gels in final product”: gels form in extruder/ Dp across screen pack excessive/ screen area too small/ size and type of screen cannot retain gel/ gels form downstream of screen pack/ downstream temperature promotes gel formation. “Dp excessive”: filter media too fine/ screen area too small/ screen temperature too low/ gel formation in extruder. “Filter media pushes through back support plate in screen pack”: screens on the downstream side of the pack not coarse enough or not rigid enough/ support plate holes too large. 3.5.7

Settlers for LS Separation

Grit chamber: Trouble shooting: “Floating sludge”: sludge decomposing and buoyed to the surface/ infrequent sludge removal/ sludge not removed from the hoppers. “Excessive sedimentation at the inlet”: fluid velocity too slow. “Intermittent surging”: intermittent pumping rates/ liquid maldistribution of feed. “Sludge hard to remove from hopper”: high feed concentration of grit, clay/ low velocity in the sludge withdrawal lines. 3.5.8

Hydrocyclones for LS Separation

Good practice: control on pressure drop. Trouble shooting: “Underflow too dilute, underflow appears as smooth inverted cone”: inlet velocity low/ inlet feed pressure low. “Underflow appears as slow, vertical rope of coarse solids”: underflow opening too small/ feed concentration of solids higher than design. “No discharge from the underflow”: plugged inlet/ plugged underflow. “Underflow unsteady and variable inlet pressure”: air-gas in the feed. 3.5.9

Thickener for LS Separation

Good practice: consider the use of flocculants or use a deep cone. Flocculant dosage should be related to feed inlet concentration, see also Section 3.9.3. Include highpressure water purge lines for both forward and reverse flow at > 1 m/s. Raise and lower the rake once per shift. For startup, pump feed into the empty tank and recycle underflow until the design underflow densities are achieved. Trouble shooting: “Stalled rake”: uneven central feed distribution/ excessive flocculant causing island formation/ underflow concentration > design/ unpumpable underflow/ trying to maintain the underflow concentration when the feed contains fines > design/ storing too many solids in the thickener/ “sanding out”/ particle shape differs from design. “Plugged underflow lines”: insufficient fines/ targetting underflow concentration > design/ temperature change/ pump problems, see Section 2.2.3 / suction velocity < 0.6–2.5 m/s. “Underflow concentration of solids too low”: removal of too much underflow/ flocculation problems that give islands that lead to the feed concentra-

3.5 Heterogeneous Separations

tion ratholing directly to underflow. “Supernatant cloudy”: feed velocity excessive causing breakup of colloidal flocs/ changes in pH or electrolyte concentration causing floc breakup/ excessive feed turbulence/ excessive vertical drops of feed/ insufficient flocculant added or flocculant feedrate constant instead of proportional to solids concentration in the feed. “Sanding out”: too high an underflow concentration/ feed concentration of dense particles > 200 mm is > design/ power failure. “Torque > expected:” feed concentration suspended solids > expected. 3.5.10

Sedimentation Centrifuges

Horizontal scroll discharge decanter. Trouble shooting: “Centrifuge won’t start:” vibration switch triggered/ no power/ motor or starter failure/ overheated motor/ [broken shear pin]* / lubrication oil flowswitch tripped. “Centrifuge shuts down”: blown fuse/ overload relays tripped/ motor overheated/ [broken shear pin]* / lube oil flowswitch tripped. “Excessive vibration”: broken isolators/ motor on flexible mounts/ motor bolts loose/ flexible piping not used/ misalignment/ bearing failure or damaged/ loss of plows/, damaged conveyor hub/ solid product buildup in conveyor hub/ conveyor or bowl not balanced/ conveyor flights worn or portion of blade missing/ trunions cracked or broken/ conveyor bowl cracked or broken/ leaking effluent weirs/ plugged solids in the effluent hopper/ not level. “High moisture in exit solids”: liquid dams not set alike or set incorrectly/ feed temperature too low/ feedrate too high/ effluent hopper plugged or not vented/ conveyor flights worn. “High solids in liquid effluent”: feedrate excessive/ effluent dams set wrong/ no strip installed/ incorrect feed temperature. [Shear pins breaks]*: feedrate too high/ solids concentration too high/ foreign material stuck in bowl. [Solid and screen bowl shear pins break]*: plugged discharge hopper/ conveyor blades bent or rough/ worn bowl strips/ loose or broken trunion bolts/ bowl inadequately washed/ clearance too large for blade tip to bowl wall/ bowl inside rough/ wrong size shear pin. 3.5.11

Filtering Centrifuge

Good practice: monitor pH and temperature. Pusher: Trouble shooting: “Machine floods”: feed concentration < design/ feedrate > design/ irregular feedrate/ change in size distribution or particle diameter. “Unstable cake formation”: feedrate > design.

87

88

3 Rules of Thumb for Trouble Shooting

3.5.12

Filter for LS Separation

Good practice: Precoat: 0.75 kg/m2 to give a precoat thickness of 1.6 mm. Rate for precoat: concentration between 0.3 and 5% w/w and at a rate of 0.7–1.4 L/s. m2. This should give a Dp= 14 kPa. For leaf or rotary filters, maintain consistent pressure differential across cake once the cake is formed. Consider adding body feed continuously when filtering gelatinous species. Trouble shooting: “Poor clarity”: leak/ cracks in cake/ partially blinded septa/cake washing too fast/ flashing of filtrate/ air in filter of feed liquid/ changes in liquid properties/ incorrect filter aid/ change in temperature of pH/ small diameter particles in feed than design/ process upset. “Short cycle/ high pressure/ low flow”: flow lines too small/ obstruction in outlet line/ pump sucking air/ pressure differential too low/ wide fluctuations in feedrate/ air trapped in filter/ too high a filtration rate. 3.5.13

Screens for Solid–Solid Separation

Screen, vibrating: Good practice: if damp or sticky, predry or use heater above the screen to reduce moisture to < 3%. Avoid resonance frequencies. Usual angle of operation is 12 to 18 ; for wet, inclined vibrating screen to 7 to 110. Capacity decreases if the angle of inclination is too high. Blinding is mainly caused by material that is 1 to 1.5 times the hole size. Feed thickness should not exceed 4 aperture size for 1.6 Mg/m3; and not exceed 2 to 3 aperture size for 0.8 Mg/m3. Trouble shooting: “Capacity decreases”: angle of inclination too high/ blinding.

3.6

Reactor Problems

Temperature is usually a key variable. Increasing the temperature by 10 C, doubles the rate of reaction. Operating temperature should be at least 25 C less than the maximum temperature for a catalyst. For PFTR: increasing the temperature may break down the catalyst into a powder that causing dusting/ contamination/ plugging problems downstream and increase the pressure drop in the reactor; may cause the catalyst to agglomerate and deactivate with a drop off in conversion and increase in pressure drop; may lead to tube failure; may promote coking. Increasing the velocity of reactants through the reactor decreases the time available for reaction and exit concentration of product should decrease, pressure drop should increase. Here are typical trouble shooting symptoms and causes for plug flow tubular reactors, stirred tank reactors and reactive extrusion. For PFTR, we consider multitube fixed bed catalyst, nonadiabatic, Section 3.6.1; fixed bed, adiabatic, Section 3.6.2; bubble reactors, Section 3.6.3; packed column reactors, Section 3.6.4; trickling

3.6 Reactor Problems

bed, Section 3.6.5 and thin film, Section 3.6.6. For STR, we considered batch STR, Section 3.6.7; semibatch, Section 3.6.8; CSTR, Section 3.6.9; fluidized bed, Section 3.6.10. In Section 3.6.11 we consider a mix of CSTR, PFTR with recycle. Finally reactive extrusion is considered in Section 3.6.12. 3.6.1

PFTR: Multitube Fixed-Bed Catalyst, Nonadiabatic

Trouble shooting: “Pressure surge”: possible shutdown?/[runaway reactor]*. “Dp increases dramatically, top of tubes hot, less conversion than expected”: possible shutdown?/ contamination in feed / [poisoned catalyst]*. “Rapid decline in conversion”: unfavorable shift in equilibrium at operating temperature, for exothermic reactions/ [sintering]*/ [agglomeration]*/ poison in new feed . “Gradual decline in conversion”: sample error/ analysis error/ temperature sensor error/ [catalyst activity lost]*/ [maldistribution]*/ [unacceptable temperature profiles]*/ [inadequate heat transfer]*/ wrong locations of feed, discharge or recycle lines/ faulty design of feed and discharge ports/ wrong internal baffles and internals/ faulty bed-voidage profiles. “Gradual decline in conversion and axial temperature constant with depth of region increasing with time”: [poisoned catalyst]*. “Gradual decline in conversion and axial temperatures < usual”: [poisoned catalyst]*. “Gas exit concentration of reactants high”: sample error/ analysis error/ catalyst selectivity low/ [catalyst activity lost]*. “Exit concentration of product higher than design”: reactor leaking. “Change in product distribution”: [maldistribution]* / [poisoned catalyst]*/ feed contaminants/ change in feed/ change in temperature settings. “Temperature runaways”: [temperature hot spots]*/ [reactor instability]*. “Pressure and bed temperature and reactor unsteady”: water in feed/ [maldistribution]*. “Local high temperature/hot spot with T > 100 C above normal”: [maldistribution of gas flow]*/ instrument error/ extraneous feed component that reacts exothermically. “Local low temperature within the bed”: [maldistribution of gas flow]*/ instrument error/ extraneous feed component that reacts endothermically. “Exit gas temperature too high”: instrument error/ control-system malfunction/ fouled reactor coolant tubes. “Temperature varies axially across bed”: [maldistribution]*. “Soon after startup, temperature of tubewall near top > usual and increasing and perhaps Dp increase and less conversion than expected or operating temperatures > usual to obtain expected conversion”: inadequate catalyst regeneration/ contamination in feed; for steam reforming sulfur concentration > specifications/ wrong feed composition; for steam reforming: steam/CH4 < 7 to 10. “Soon after startup, temperatures over full length of some tubes > usual and perhaps Dp > or < usual and may increase with time”: faulty loading of the catalyst/ [maldistribution]*. “Hot bands or stripes; perhaps Dp increase”: low ratio of steam to methane/ [carbon formation; whisker type]*/ wrong feed composition: for steam reforming steam/methane < 7 to 10: 1. “Hot bands or stripes near top and perhaps over all tube and rapidly increasing Dp and conversion < specifications”: [deactivated catalyst by pyrolytic coke formation]*/ feed concentration wrong: for steam reforming high concentration of heavier hydrocarbons/ steam to hydrocarbon ratio low/ [catalyst poisoned]* by sulfur. “Temperature at inlet high and high Dp”: [ for steam

89

90

3 Rules of Thumb for Trouble Shooting

reforming: steam contaminated with inorganic solids]*. “Hot bands in top 1/3 of tubes and methane > usual in exit gas and perhaps Dp increase”: contamination in feed / [poisoned catalyst]*. “Dp higher than design”: catalyst degradation/ instrument error/ high gas flow/ sudden coking/ crud left in from construction or revamp. “Dp increasing gradually yet flowrate constant”: [coke formation]*/ [dust or corrosive products from upstream processes]*. “Startup after catalyst regeneration, conversions < standard”: [regeneration faulty]*. “Startup after catalyst replacement, poor selectivity”: bad batch of catalyst/ preconditioning of catalyst faulty/ temperature and pressures incorrectly set/ instrument error for pressure or temperature. “Startup after catalyst replacement, Dp < expected and conversion < standard”: [maldistribution]* and axial variation in temperature/ larger size catalyst. “Startup after catalyst replacement, conversion < standard and Dp increasing”: [maldistribution and axial temperatures different]*/ feed precursors present for polymerization or coking. “Startup after catalyst replacement, Dp for this batch of catalyst > previous batch”: catalyst fines produced during loading/ poor loading. “Startup after catalyst replacement, conversion < specifications per unit mass of catalyst and more side reactions”: [maldistribution]*/ faulty inlet distributor/ faulty exit distributor. “Startup after catalyst replacement, poor selectivity”: bad batch of catalyst/ preconditioning of catalyst faulty/ [tube walls not passified]*/ temperature and pressures incorrectly set/ instrument error for pressure or temperature. “Startup after catalyst replacement, increased side reactions and conversion < specification”: catalyst loading not the same in all tubes. [Active species volatized]*: [regeneration faulty]*/ faulty catalyst design for typical reaction temperature/ [hot spots]*. [Agglomeration of packing or catalyst particles]*: [temperature hot spots]*. [Attrition of the catalyst]*: flowrates > expected/ catalyst too fragile. [Carbon buildup]*: [inadequate regeneration]*/ [excessive carbon formed]*. [Catalyst selectivity changes]*: [poisoned catalyst]*/ feed contaminants/ change in feed/ change in temperature settings. [Catalyst activity lost]*: [carbon buildup]*/[regeneration faulty]*/ [sintered catalyst]*/ excessive regeneration temperature/ [poisoned catalyst]*/ [loss of surface area]*/ [agglomeration]*/ [active species volatized]*. [Excessive carbon formed]*: operating intensity above usual/ feed changes/ temperature hot spots. [Dust or corrosive products from upstream processes]*: in-line filters not working or not installed/ dust in the atmosphere brought in with air/ air filters not working or not installed. [Loss of surface area]*: [sintered catalyst]*/ [carbon buildup]*/ [agglomeration]*. [Maldistribution]*: faulty flow-distributor design/ plugging of flow distributors with fine solids, sticky byproducts or trace polymers/ [sintered catalyst particles]*/ [agglomeration of packing or catalyst particles]*/ fluid feed velocity too high/ faulty loading of catalyst bed/ incorrect flow collector at outlet. [Poisoned catalyst]*: poisons in feed/ flowrate of “counterpoison” insufficient/ poison formed from unwanted reactions.

3.6 Reactor Problems

[Poisons in feed]*: depends on reaction/ contamination in feed/ upstream process or equipment upsets/ changes in feed. Poisons for platforming include high sulfur in feed and high feed end point with upstream equipment failure being compressor failure/ water upset/ chloride upset. Poisons for steam reforming: sulfur, arsenic and alkali metals in the hydrocarbon or steam. [Reactor instability]*: control fault/ poor controller tuning/ wrong type of control/ insufficient heat transfer area/ feed temperature exceeds threshold/ coolant temperature exceeds threshold/ coolant flowrate < threshold/ tube diameter too large. [Regeneration doesn’t remove all carbon from the catalyst]*: regeneration temperature not hot enough/ regeneration time not long enough/ [maldistribution]*. [Regeneration faulty]*: temperatures too high/ oxygen concentration < standard/ oxygen concentration > standard causing too rapid a burn/ incorrect temperature and time so that coke left on catalyst. [regeneration doesn’t remove all carbon from the catalyst]*/ excessive temperature during regeneration. [Runaway reactor]*: feed temperature too high/ [temperature hot spot]*/ cooling water too hot/ feed temperature too high. [Sintered catalyst]*: temperature sensor error/ [temperature hot spots]*/ [maldistribution]*/ temperature in reactor too high/ regeneration temperature too high. [Temperature hot spots]*: bed too deep/ [maldistribution]*/ flowrate < design/ instrument error/ extraneous feed component that reacts exothermically. [Tube walls not passified]*: walls activated unwanted side reactions and faulty passivation treatment/ wrong passivation treatment/ no passivation treatment. 3.6.2

PFTR: Fixed-Bed Catalyst in Vessel: Adiabatic

Trouble shooting:4) gas-catalytic reactions. Temperature and pressure drops across bed are usually key variables. When a hot spot develops, it usually develops at the front end of the bed and gradually moves through the bed. It may take three to four weeks to travel through the full bed. If the hot spot is 100–200 C above normal, then usually carbon is deposited and the catalyst is irrevocably damaged. Temperature control is critical for exothermic reactions. “Dp rapidly increases”: emergency shutdown? “Pressure surge”: possible shutdown?/[runaway reactor]*. “Rapid decline in conversion”: unfavorable shift in equilibrium at operating temperature, for exothermic reactions/ [sintering]*/ [agglomeration]*/ poison in new feed . “Gradual decline in conversion”: sample error/ analysis error/ temperature sensor error/ [catalyst activity lost]*/ [maldistribution]*/ [unacceptable temperature profiles]*/ wrong locations of feed, discharge or recycle lines/ faulty design of feed and discharge ports/ wrong internal baffles and internals/ faulty bed-voidage profiles. “Gradual decline in conversion and axial temperature constant with depth of region increasing with time”: [poisoned catalyst]*. “Gradual decline in conversion and axial 4) Based on R.B. Anderson, person communica-

tion; H.F. Rase “Fixed bed reactor design and diagnostics”, 1990, Wiley and Dutta, S. and

R. Gauly, Hydrocarbon Processing, 1999, Sept, 43–50.

91

92

3 Rules of Thumb for Trouble Shooting

temperatures < usual”: [poisoned catalyst]*. “Gas exit concentration of reactants high”: sample error/ analysis error/ catalyst selectivity low/ [catalyst activity lost]*. “Exit concentration of product higher than design”: reactor leaking. “Change in product distribution”: [maldistribution]* / [poisoned catalyst]*/ feed contaminants/ change in feed/ change in temperature settings. “Temperature runaways”: [temperature hot spots]*/ [reactor instability]*. “Pressure and bed temperature and reactor unsteady”: water in feed/ [maldistribution]*. “Local high temperature/hot spot with T > 100 C above normal”: [maldistribution of gas flow]*/ instrument error/ extraneous feed component that reacts exothermically. “Local low temperature within the bed”: [maldistribution of gas flow]*/ instrument error/ extraneous feed component that reacts endothermically. “Exit gas temperature too high”: instrument error/ control-system malfunction. “Temperature varies axially across bed”: [maldistribution]*. “Dp higher than design”: catalyst degradation/ instrument error/ high gas flow/ sudden coking/ crud left in from construction or revamp. “Dp increasing gradually yet flowrate constant”: [coke formation]*/ [dust or corrosive products from upstream processes]*. “Startup after catalyst regeneration, conversions < standard”: [regeneration faulty]*. “Startup after catalyst replacement, poor selectivity”: bad batch of catalyst/ preconditioning of catalyst faulty/ temperature and pressures incorrectly set/ instrument error for pressure or temperature. “Startup after catalyst replacement, Dp < expected and conversion < standard”: [maldistribution]* and axial variation in temperature/ larger size catalyst. “Startup after catalyst replacement, conversion < standard and Dp increasing”: [maldistribution and axial temperatures different]*/ feed precursors present for polymerization or coking. “Startup after catalyst replacement, Dp for this batch of catalyst > previous batch”: catalyst fines produced during loading/ poor loading. “Startup after catalyst replacement, conversion < specifications per unit mass of catalyst and more side reactions”: [maldistribution]*/ faulty inlet distributor/ faulty exit distributor. [Active species volatized]*: [regeneration faulty]*/ faulty catalyst design for typical reaction temperature/ [hot spots]*. [Agglomeration of packing or catalyst particles]*: [temperature hot spots]*. [Attrition of the catalyst]*: flowrates > expected/ catalyst too fragile. [Carbon buildup]*: [inadequate regeneration]*/ [excessive carbon formed]*. [Catalyst selectivity changes]*: [poisoned catalyst]*/ feed contaminants/ change in feed/ change in temperature settings. [Catalyst activity lost]*: [carbon buildup]*/[regeneration faulty]*/ [sintered catalyst]*/ excessive regeneration temperature/ [poisoned catalyst]*/ [loss of surface area]*/ [agglomeration]*/ [active species volatized]*. [Excessive carbon formed]*: operating intensity above usual/ feed changes/ temperature hot spots. [Dust or corrosive products from upstream processes]*: in-line filters not working or not installed/ dust in the atmosphere brought in with air/ air filters not working or not installed. [Loss of surface area]*: [sintered catalyst]*/ [carbon buildup]*/ [agglomeration]*.

3.6 Reactor Problems

[Maldistribution]*: faulty flow-distributor design/ plugging of flow distributors with fine solids, sticky byproducts or trace polymers/ [sintered catalyst particles]*/ [agglomeration of packing or catalyst particles]*/ fluid feed velocity too high/ faulty loading of catalyst bed/ incorrect flow collector at outlet. [Poisoned catalyst]*: poisons in feed/ flowrate of “counterpoison” insufficient/ poison formed from unwanted reactions. [Poisons in feed]*: depends on reaction/ contamination in feed/ upstream process or equipment upsets/ changes in feed. Poisons for platforming include high sulfur in feed and high feed end point with upstream equipment failure being compressor failure/ water upset/ chloride upset. [Reactor instability]*: control fault/ poor controller tuning/ wrong type of control/ feed temperature exceeds threshold. [Regeneration doesn’t remove all carbon from the catalyst]*: regeneration temperature not hot enough/ regeneration time not long enough/ [maldistribution]*. [Regeneration faulty]*: temperatures too high/ oxygen concentration < standard/ oxygen concentration > standard causing too rapid a burn/ incorrect temperature and time so that coke left on catalyst. [regeneration doesn’t remove all carbon from the catalyst]*/ excessive temperature during regeneration. [Runaway reactor]*: feed temperature too high/ [temperature hot spot]*. [Sintered catalyst]*: temperature sensor error/ [temperature hot spots]*/ [maldistribution]*/ temperature in reactor too high/ regeneration temperature too high. [Temperature hot spots]*: bed too deep/ [maldistribution]*/ flowrate < design/ instrument error/ extraneous feed component that reacts exothermically. 3.6.3

PFTR: Bubble Reactors, Tray Column Reactors

Wide variety of configurations ranging from tube loop, jet loop, air lift loop and sparger “bubble reactor”. Related reactors include CSTR, Section 3.6.9; STR, Section 3.6.7. Trouble shooting: Carryover”: [ foaming]*. [Foaming]*: surfactants present/ dirt and corrosion solids/ natural occurring surfactants/ pH far from the zpc/ naturally occurring polymers/ insufficient disengaging space above the liquid/ antifoam ineffective (wrong type or incorrect rate of addition)/ bubble rate too high/ mechanical foam breaker not rotating/ baffle foam breaker incorrectly designed or damaged/ asphaltenes present/ liquid downflow velocity through the foam is too low. See Section 3.11 for generic causes of [ foaming]*. See Trouble shooting: section STR, Section 3.6.7 for more on trouble shooting aerobic bioreactors.

93

94

3 Rules of Thumb for Trouble Shooting

3.6.4

PFTR: Packed Reactors

These include trickling filters and gas-liquid-solid packed-column bioreactor. Trickling filter: Trouble shooting: “Plugged: interstitial voids become filled with biological growth”: packing too small/ packing of variable diameter/ organic to liquid loading > design. “Ice formation on top filter surface”: liquid maldistribution/ feed liquid temperature too low/ air temperature too low. “Odors”: loss of aerobic conditions/ accumulation of sludge and biological growth/ lack of chlorine in influent/ high organic loadings in feed especially from milk processing and canneries. [Foaming]*: surfactants present/ dirt and corrosion solids/ natural occurring surfactants/ pH far from the zpc/ naturally occurring polymers/ insufficient disengaging space above the liquid/ antifoam ineffective (wrong type or incorrect rate of addition)/ vapor velocity too high/ mechanical foam breaker not rotating/ baffle foam breaker incorrectly designed or damaged/ asphaltenes present/ liquid downflow velocity through the foam is too low. Gas-liquid-solid packed-column bioreactor: Trouble shooting: Carryover”: [ foaming]*. [Foaming]*: surfactants present/ dirt and corrosion solids/ natural occurring surfactants/ pH far from the zpc/ naturally occurring polymers/ insufficient disengaging space above the liquid/ antifoam ineffective (wrong type or incorrect rate of addition)/ bubble rate too high/ mechanical foam breaker not rotating/ baffle foam breaker incorrectly designed or damaged/ asphaltenes present/ liquid downflow velocity through the foam is too low. See Section 3.11 for generic causes of [ foaming]*. See Trouble shooting: section STR, Section 3.6.7 for more on trouble shooting aerobic bioreactors. 3.6.5

PFTR: Trickle Bed

Good practice: gas-liquid flow cocurrently down through a packed bed of catalyst. Porosity 0.38–0.42. Ensure operation in the correct flow regime. The effectiveness of the solid catalyst and of the gas-liquid mass transfer decreases if solid catalyst is non-wet. For good wetting of the solid keep the surface tension of the solid > surface tension of the liquid. Prevent foaming. The efficiency depends on the skill in initially distributing the gas and the liquid. Use liquid distribution plate similar to design used for packed towers. The liquid distribution plate should have at least 50 holes/m2 of catalyst bed.

3.6 Reactor Problems

Trouble shooting:5) For trickle bed reactors with specific applications to hydrotreating. “Low conversion”: feed composition change/ wrong catalyst for feed/ sample error/ flowrate error/ feedrate higher but reactor temperature not increased/ temperature profile wrong/ thermocouple fault/ controller fault/ feed bypassing reactor through leak in heat exchanger/ [channeling]*/ [catalyst]*/ [ foaming]*/ For hydrotreating: [hydrogen starvation]*/ catalyst not presulfided/ [incomplete presulfiding of catalyst]*. “Sudden loss of activity of catalyst”: heat exchanger leak/ change in feed composition/ For hydrotreating: [hydrogen starvation]*. “Dp across the catalyst bed > design”; [channeling]*/ cracked hydrocarbon feed stored without effective nitrogen blanket/ solids in feed/ corrosion products from upstream operations/ bypass on feed filter open/ feed distributor fault/ top catalyst support tray has holes that are too small/ bottom catalyst bed support tray holes are too large/ pugged or partially plugged outlet/ crush strength of catalyst exceeded and fines plug bed/ excessive recycle compressor surge causing breakdown of top layer of catalyst. For hydrotreating: “Rapid breakthrough of H2S during catalyst sulfiding”: [channeling]*. “Nonuniform bed temperatures across the diameter during sulfiding”: [channeling]*. “Color > specifications”: composition change in feed/ catalyst aged.. [Catalyst]*: regeneration failed to remove carbon from catalyst/ excessive regeneration temperature > 540 C causing sintering, > 760 C molybdenum sublimation, > 820 C reduction in crush strength and change in alumina/ poisons in feed/ aged catalyst. [Channeling]*: nonuniform catalyst bed density/ low superficial flowrate < 1.4 kg/ s m2/ offset, tilted or faulty feed distributor/ thermal shock to upstream pipes or equipment causes scale to dislodge and buildup on bed/ internal vessel obstructions such as thermowells or supports. [Foaming]*: surfactants present/ dirt and corrosion solids/ natural occurring surfactants/ pH far from the zpc/ naturally occurring polymers/ insufficient disengaging space above the liquid/ antifoam ineffective (wrong type or incorrect rate of addition)/residence time insufficient/ designed for a vertical vessel but a horizontal vessel installed/ vapor velocity too high/ mechanical foam breaker not rotating/ baffle foam breaker incorrectly designed or damaged/ asphaltenes present/ liquid downflow velocity through the foam is too low/ operating in the wrong flow regime. [Hydrogen starvation]*: change in feed composition without corresponding change in hydrogen/ leaks/ dissolution of hydrogen in liquid product/ lower concentration of hydrogen in treat gas/ flowrate of treat gas < expected because of recycle compressor fault. 5) Based on Koros, R.M. “Engineering Aspects

of Trickle Bed Reactors”, pp. 579–630 in “Chemical Reactor Design and Technology” H. de Lasa (Ed), Martinus Nijhoff Publishers,

1986 and M.D. Edgar, D.A. Johnson, J.T. Pistorius and T. Varadi “Trouble Shooting Made easy” HP May 1984 p. 65.

95

96

3 Rules of Thumb for Trouble Shooting

[Incomplete presulfiding of catalyst]*: contact with hydrogen at high temperature for too long a time/ maximum temperature of 150–175 C exceeded/ use of cracked feed/ excessive addition of presulfiding agents. [Rapid coking of catalyst]*: [hydrogen starvation]*/ temperatures too high. 3.6.6

PFTR: Thin Film

Related topics evaporation, Section 3.4.1 for gravity and agitated falling films, absorbers, Section 3.4.8, and shell and tube heat exchangers, Section 3.3.3.See Section 3.3.3 for trouble shooting Vertical falling-film evaporator. 3.6.7

STR: Batch (Backmix)

Trouble shooting: Batch STR used for polymerization and, to a lesser extent, nitration, sulfonation, hydrolysis, neutralization and, to a much lesser extent, dehydrogenation, oxidation and esterification can pose potentially unsafe operation. Key indicators of such potential hazards include “Sudden increase in pressure”, “Unexplained increase in temperature”, “Failure of the mixer”, “Power failure”, and “Loss of cooling water”. For any of these conditions our first question should be: emergency shut down? Our knowledge of the MSDS information for the species and their interaction with each other and with the environment is critical. Aerobic bioreactors: Trouble shooting: “inoculation cannot be used for the reactor/ fermenter”: [contamination]*. “product formation is inhibited”: [contamination]*. “target product cannot be separated from contaminating species”: [contamination]*. “fermentation broth cannot be filtered”: [contamination]* .”steam out of air filter yields dark brown liquid”: media blowback. “reduction in cell volume, no further product production, no oxygen uptake, no heat production”: [contamination by bacteriophage]* . “foaming”: air leaks through gaskets, coils, jacket, hatch/ pH shifted away from zpc/ particles present. [contamination in the first 24 hours]*: contaminated inoculum/ poor sterilization of tank accessories and content/ unsterile air. [contamination comes in after 24 hours]*: air supply/ nutrient recharges/ antifoam feed/ loss of pressure during the run/ lumps in the media/ media blowbacks. [contamination]*: [stock culture contaminated]*/ [raw materials contaminated]*/ [inoculation tank contaminated]*/ [ fermenter contaminated]* / [incorrect procedures]*/ [ faulty maintenance]*/ [contamination by bacteriophage]* [stock culture contaminated]*: foreign microorganisms in culture stock/ contaminated inoculation flash/wrong sterilization procedure/ temperature and pressure instruments wrong/ air left in sterilization chamber/ sterile area contaminated/ cotton plugs contaminated/ [ faulty sterilization]*/ raw material contaminated with spores combined with inadequate germination-sterilization. [raw materials contaminated]*: dry materials not finely ground/ lumps not removed/ insoluble solids not suspended in solution well because of lumping or

3.6 Reactor Problems

inadequate mixing/ lumps too big to be sterilized in the time–temperature available/ mixing inadequate to keep particles suspended/ particles enter air sparger during filling operation/ starches or proteins not prehydrolyzed with enzymes. [inoculation tank contaminated]*: wrong procedures/ tank dirty/ air leak/temperature and pressure instrument fault/ sample line, inoculation fitting dirty/dirty dead spots, debris, corrosion/ media blow back into air filter/ unsterilized air filter/faulty antifoam or pH additive lines. [ fermenter contaminated]*: [inoculum tank contaminated]* / inoculum line contaminated/ procedure wrong/ tank dirty/ air leak/ leak from the coil or jacket/faulty sensors/ antifoam is not sterile/ dirty gaskets, bottom valve, sample line and valve, vent line valve, vacuum breaker/ nutrient feed tank or line not sterile/ all lines were not up to sterilization temperature/ steam condensate left in lines/ the humidity of the fermenter air upstream of the “sterile filter” is > 90%/ pH and DO probes were not cleaned between runs/ probe holders were not brushed and cleaned with a hypochlorite or formaldehyde solution/ for a previously contaminated vessel the valves and gaskets were not replaced, instrument sensors were not removed and cleaned; high boiling germicide, such as sodium carbonate or sodium phosphate was not used. [ faulty maintenance]*: braided packing (on agitator shafts for sterile vessels) not receiving enough germicide/ mechanical seals (on agitator shafts for sterile vessels) not lubricated with sterilizing liquid/ instruments faulty/ bolts on flanges not tightened after heat up to 120 C/ packed bed air filters not packed to correct density of 200–250 kg/m3. [ faulty sterilization]*: particles too coarse and dry/ particles not wetted/ particles not suspended/raw material contaminated with spores plus inadequate germination-sterilization. [contamination by bacteriophage]*: source usually difficult to trace/ substitute an immune strain/ develop strain resistant to the phage. “Carryover”: [ foaming]* [Foaming]*: surfactants present/ dirt and corrosion solids/ naturally occurring surfactants/ pH far from the zpc/ naturally occurring polymers/ insufficient disengaging space above the liquid/ antifoam ineffective (wrong type or incorrect rate of addition)/ bubble rate too high/ mechanical foam breaker not rotating/ baffle foam breaker incorrectly designed or damaged/ asphaltenes present/ liquid downflow velocity through the foam is too low. See Section 3.11 for generic causes of [ foaming]*. Anaerobic digesters: Trouble shooting: “Sludge temperature fluctuates”: instrument fault/ fluctuating feedrate. “Poor heat transfer with the hot water coils, exit water temperature < design: “ sludge solids adhere to heat-transfer surface. “Temperature constant but production of methane gas < design”: increased accumulation of scum or grit/ excessive acid production with lower pH and volatile acid > 500 mg/L/ organic overload/ toxic metals in feed/ highly acidic feed/ overdigested sludge. “Foaming”: incomplete digestion/feedrate > design/ inadequate mixing/ temperature too low/ withdrawal of too much product (digested sludge)/ rate of reaction > design/ large

97

98

3 Rules of Thumb for Trouble Shooting

quantities of organics in feed/ insufficient reaction volume for the high organic feedrate/ pH shift away from zpc. 3.6.8

STR: Semibatch

Trouble shooting: Semi-batch STR used for polymerization and, to a lesser extent, nitration, sulfonation, hydrolysis, neutralization and, to a much lesser extent, dehydrogenation, oxidation and esterification can pose potentially unsafe operation. Key indicators of such potential hazards include “Sudden increase in pressure”, “Unexplained increase in temperature”, “Failure of the mixer”, “Power failure”, and “Loss of cooling water”. For any of these conditions our first question should be: emergency shut down? Our knowledge of the MSDS information for the species and their interaction with each other and with the environment is critical. Polymerizer: Trouble shooting: “Temperature increases suddenly: “ emergency shutdown?/ mixer stopped/ fouling of heat exchanger/ “gel effect” in polymerization reaction/ coagulation and product fouls the walls of the reactor. “Particle product size < design”: too many nucleation sites/ lower level of oxygen than design/ too much emulsifier/ too much initiator. “Particle product size > design”: coagulation/ too few initial nucleation sites/ too much oxygen in the feed/ too little emulsifier/ too little initiator/ emulsifier post feed is too late. “Temperature increases > design”: emergency shutdown?/ coagulation/ emulsifier post feed too late. “Batch times < design”: too many nucleation sites/ lower level of oxygen than design/ too much emulsifier/ too much initiator. “Batch times > design”: too few initial nucleation sites/ too much oxygen in the feed/ too little emulsifier/ too little initiator. Agitated bubble reactors: Trouble shooting. Consider increasing the impeller diameter or using a disk turbine to increase mass transfer. Trouble shooting: “foaming”: mixer tip speed too high/ linear gas velocity too high/ surfactant contaminants/ decrease in electrolyte concentration in the liquid/ change in pH/ use of turbine impeller/ lack of a gas sparger/ mechanical foam breaker not rotating/ disengagement space not high enough/ mechanical baffle foam breakers faulty/ antifoam ineffective. “flooded impeller”: too small a diameter impeller/ speed too slow. Gas-liquid-solid bioreactor: Trouble shooting: Carryover”: [ foaming]*. [Foaming]*: surfactants present/ dirt and corrosion solids/ natural occurring surfactants/ pH far from the zpc/ naturally occurring polymers/ insufficient disengaging space above the liquid/ antifoam ineffective (wrong type or incorrect rate of addition)/ bubble rate too high/ mechanical foam breaker not rotating/ baffle foam breaker incorrectly designed or damaged/ asphaltenes present/ liquid downflow velocity through the foam is too low. See Section 3.11 for generic causes of [ foaming]*. See trouble shooting: section STR, Section 3.6.7 for more on trouble shooting bioreactors.

3.6 Reactor Problems

3.6.9

CSTR: Mechanical Mixer (Backmix)

Consider complications because of catalyst deposition and erosion. Trouble shooting: CSTR used for polymerization and, to a lesser extent, nitration, sulfonation, hydrolysis, neutralization and, to a much lesser extent, dehydrogenation, oxidation and esterification can pose potentially unsafe operation. Key indicators of such potential hazards include “Sudden increase in pressure”, “Unexplained increase in temperature”, “Failure of the mixer”, “Power failure”, and “Loss of cooling water”. For any of these conditions our first question should be: emergency shut down? Our knowledge of the MSDS information for the species and their interaction with each other and with the environment is critical. See Semibatch and STR Sections 3.6.8 and 3.6.7 for more. Liquid–liquid: Typically reactor is a CSTR followed by a decanter to separate the phases and recycle the “catalyst” phase to the reactor. Trouble shooting: see decanter: Section 3.5.3a. “Alkylate is purple”: [stable emulsion formation]*/[density difference decrease]*/ [drops don’t settle]*/ [acid runaway]*. “Dp across the alkylate cooler > design”: [stable emulsion formation]*/ [density difference decrease]*/ [drops don’t settle]*/ [acid runaway]*/ acid recirculation rate too fast. [Acid runaway]*: excessive contaminants in feed to reactor/ feedrate too fast/ poor contact or mixing between isobutane, olefin and acid/ fresh acid makeup feedrate stopped/ faulty control/ faulty meter/ ratio of acid: hydrocarbon outside range 45– 60% v/v/ ratio of isobutane: olefin < 8: 1/ initial reactor temperature too high or > 18 C/ poor mechanical design for fresh acid addition. “temperature of the recycled acid is > 1.7 C hotter than feed entering the reactor”: [acid runaway]*/ alkyl sulfates polymerize in the decanter/ acid recirulation rate too fast. [Density difference decrease]*: dilution of the dense phase/ reactions that dilute the dense phase; for sulfuric acid alkylation: if acid strength < 85% w/w the olefins polymerize with subsequent oxidation of the polymers by sulfuric acid. as a self-perpetuating continuing decrease in acid strength. Alkylate-acid separation is extremely difficult when acid concentration is 40% w/w. [Drop doesn’t settle]*: [density difference decrease]*/ [viscosity of the continuous phase increases]*/ [drop size decreases]*/ [residence time for settling too short]*/ [phase inversion or wrong liquid is the continuous phase]*/ pressure too low causing flashing and bubble formation. [Drop settles and coalesces but is re-entrained]*: faulty location of exit nozzles for liquid phases/ distance between exit nozzle and interface is < 0.2 m/ overflow baffle corroded and failure/ interface level at the wrong location/ faulty control of interface/ liquid exit velocities too high/ vortex breaker missing or faulty on underflow line/ no syphon break on underflow line/ liquid exit velocities too high. [Drop settles but doesn’t coalesce]*: [phase inversion]*/ pH far from zpc/ surfactants, particulates or polymers present/ electrolyte concentration in the continuous phase < expected/ [coalescer pads ineffective]*/ [drop size decrease]*/ [secondary haze forms]*/ [stable emulsion formation]*/ [interfacial tension too low]*/ [Marangoni effect]*.

99

100

3 Rules of Thumb for Trouble Shooting

[Drop size decrease]*: feed distributor plugged/ feed velocity > expected/ feed flows puncture interface/ local turbulence/ distributor orifice velocity > design; for amine units: for amine > 0.8 m/s; for hydrocarbon > 0.4 m/s/ [Marangoni effects]*/ upstream pump generates small drops/ [secondary haze forms]*/ poor design of feed distributor. [Inaccurate sensing of the interface]*: instrument fault/ plugged sight glass. [Interfacial tension too small]*: temperature too high/ [surfactants present]* at interface. [Marangoni effects]*: non-equilibrated phases/ local mass transfer leads to local changes in surface tension and stability analysis yields stable interfacial movement. [Phase inversion]*: faulty startup/ walls and internals preferentially wetted by the dispersed phase. [Rag buildup]*: collection of material at the interface: [surfactants present]* / particulates: example, products of [corrosion see Section 3.1.2]*, amphoteric precipitates of aluminum/ naturally occurring or synthetic polymers. [Residence time for settling too short]*: interface height of the continuous phase decreases/ [inaccurate sensing of interface]*/ turbulence in the continuous phase/ flowrate in continuous phase > expected; for example > 3 L/s m2 / sludge settles and reduces effective height of continuous phase/ [phase inversion]*/ inlet conditions faulty. [Secondary haze forms]*: small secondary drops are left behind when larger drop coalesces, need coalescer promoter, see Section 3.9.2. [Stable emulsion formation]*: [surfactants present]* / contamination by particulates: example, products of [corrosion products, see Section 3.1.2]*, amphoteric precipitates of aluminum or iron/ pH far from the zpc/ contamination by polymers/ temperature change/ decrease in electrolyte concentration/ the dispersed phase does not preferentially wet the materials of construction/ coalescence–promoter malfunctioning/ improper cleaning during shutdown/ [rag buildup]*. [Surfactants present]*: formed by reactions/ enter with feed, example oils, hydrocarbons >C10, asphaltenes/ left over from shutdown, example soaps and detergents/ enter with the water, example natural biological species, trace detergents. [Viscosity of the continuous phase increases]*: temperature too low, for alkylate-acid separation, temperature < 4.4 C/ [phase inversion]*/ contamination in the continuous phase/ unexpected reaction in the continuous phase causing viscosity increase.

3.6 Reactor Problems

3.6.10

STR: Fluidized Bed (Backmix)

Trouble shooting: First we consider fluidized-bed reactors in general, then fluidized combustors or regenerators and then provide specifics for a fluid catalyst cracking unit, FCCU, which consists of a riser or fluidized-bed reactor, cyclone separator, steam stripper, spend catalyst transport, air-oxidizing regenerator, cyclone separator and a regenerated catalyst return.6) General fluidized-bed reactor: “Gradual change in yield”: [carbon buildup]*. “Poor yield”: [Loss of catalyst activity]*/ [maldistribution]*/ [unacceptable temperature profiles]*/ [inadequate heat transfer]*/ wrong locations of feed, discharge or recycle lines/ faulty design of feed and discharge ports/ [inadequate mixing]*/ [excessive backmixing]*/ wrong internal baffles and internals/ [poor bubbling hydrodynamics]*/ [inadequate solids circulation rates in reactor]*. “Change in product distribution”: [maldistribution]* / poisoned catalyst/ feed contaminants/ change in feed/ change in temperature settings.”Temperature hot spots”: [maldistribution]*/ local exothermic reactions. “Temperature runaways”: temperature hot spots. “Pressure and bed temperature and reactor unsteady”: water in feed/ reactor grid hole erosion/ [maldistribution]*/ for FCCU: surging regenerator holdup/ unsteady reactor-regenerator differential pressure controller operation/ rough circulation/ incorrect aeration of U-bend/ incorrect aeration of standpipe/ sticky stack slide valves/ sensor control performance for stack slide valve unsatisfactory.”Particulate carryover that affects operation of downstream equipment”: [poor separation in cyclone]*. “Shifts in yield distribution”: [Feed contaminated with light hydrocarbons]*/ [sintered catalyst]*/ coarse particles. “Dp increase across the grid”: [plugged grid holes]*/ fluid flow > usual. “Dp across grid < expected”: air flowrate < design/ [eroded grid holes]*/ for FCCU: [Failure of internal seals in regenerator]*. “Erratic or cycling pressures”: [surging of the catalyst bed]*. “Catalyst losses increase”: [poor separation in cyclone]*/ insufficient head space above bed/ fluidization velocity too high/ increase in volume of product through unexpected side reactions/ change in feed flowrate/ flowrate instrument error/ velocity through reactor too high/ pressure surges/ attrition of catalyst. [Attrition of the catalyst]*: steam flowrate > expected/ air flowrate > expected/ local velocities into the dense phase > 60 m/s/catalyst too fragile. [Carbon buildup]*: [inadequate regeneration]*/ [excessive carbon formed]*. [Coarse particles (diameter > design)]*: [generation of fines]*/[loss of catalyst fines]*/ [poor separation in cyclone]*/ agglomeration of catalyst/ [sintered catalyst]*/ wrong specifications for catalyst. [Eroded grid holes]*: hole velocity too high / materials of construction/ contaminants in fluid. [Excessive backmixing]*: [maldistribution]*/ [poor bubbling hydrodynamics]*. 6) Based on Luckenbach, E.C. et al. “Encyclopedia

Processing and Design”, Marcel Dekker 1981 p. 89; Dutta, S. and R. Gualy, “Overhaul pro-

cess reactors”, HP 1999 Sept pp. 43–50, Lieberman, N. P. “Troubleshooting Process Operations”, 2nd ed. 1985 Pennwell Books.

101

102

3 Rules of Thumb for Trouble Shooting

[Excessive carbon formed in cracker]*: cracker operating intensity above usual; for FCCU excess aromatics in feed / changes in feed/ poor catalyst stripping/ heavier recycle/ leakage of fractionator bottoms into the feed/ [sintered catalyst]*/ [ feed contaminated with metals]*/ [ feed contaminated with heavy hydrocarbons, especially aromatics]*. [Failure of internal seals in regenerator for FCCU]*: pressure bump during startup/ regenerator pressure too high/ velocity through the grid too low/ low flow of air to the grid/ stresses too high/ erosion/ abnormal conditions with the auxiliary burner on startup. [Gas bubbles too big]*: particles heavier than design/ particles larger than design/ sintered particles/ single fluidized bed too deep. [Gas bypassing in fluidized bed]*: particles heavier than design/ particles larger than design/ agglomerated particles/ single fluidized bed too deep instead of multiple beds in series. [Gas velocity too high]*: [increase in production of light ends in reactor]*. [Generation of fines]*: [attrition of the catalyst]*/ fines in the new catalyst. [Inadequate heat transfer]*: [maldistribution]*/ insufficient heat exchanger area/ design error/ fouled exchanger. see Section 3.3.8. [Inadequate mixing]*: [maldistribution]*/ [poor bubbling hydrodynamics]*. see Section 3.7.1. [Inadequate regeneration]*: [regenerator doesn’t remove all carbon from the catalyst]*/ excessive temperature during regeneration/ coarse particles. [Loss of catalyst activity]*: [carbon buildup]*/[inadequate regeneration]*/[sintered catalyst]*/ excessive regeneration temperature/ [poisoned catalyst]*/ [loss of surface area]*. [Loss of catalyst fines]*: insufficient disengaging space above the top of the bed/ agglomeration of catalyst/ [poor separation in cyclone]*/ Dp indicator for catalyst level faulty/ Dp indicator for catalyst level OK but bed density incorrect. [Loss of surface area]*: [sintered catalyst]*/ [carbon buildup]*. [Maldistribution]*: faulty feed-distributor design/ plugging of fluid distributors with fine solids, sticky byproducts or trace polymers/ [temperature hot spots]*/ [sintered catalyst particles]*/ [poor bubbling hydrodynamics]* / [poor circulation]*. [Plugged dipleg]*: spalled refractory plug/ level of catalyst in bed too high / Dp indicator for catalyst level faulty/ Dp indicator for catalyst level OK but bed density incorrect. air out periods with a lot of water or steam in vessel. [Plugged grid holes]*/ foreign debris entering with fresh catalyst/ faulty grid design/ lumps of coke or refractory in catalyst/ failure of grid hole inserts/ [sintered catalyst]*/ bits of refractory. [Poisoned catalyst]*: poisons in feed/ flowrate of “counterpoison” insufficient/ poison formed from unwanted reactions. [Poisons in feed]*: depends on reaction: for FCCU poisons in the feed include nickel, vanadium and sodium; the counterpoison is a solution of antimony. [Poor bubbling hydrodynamics]*: [Gas bypassing in fluidized bed]*/ [gas bubbles too big]* particles heavier than design/ [particles larger than design]*/ [sintered catalyst]*/ fluid feed velocity too high/ too deep a bed of catalyst/ [maldistribution]*.

3.6 Reactor Problems

[Poor circulation]*: coarse particles/ [maldistribution]*. [Poor separation in cyclone]*: [stuck or failed trickle valve]*/ [plugged dipleg]*/ dipleg unsealed/ solids level does not cover end of dipleg/ gas velocity into cyclone too low or too high/ faulty design of cyclone/ solids concentration in feed too high/ cyclone volute plugged/ hole in cyclone body/ Dp indicator for catalyst level faulty/ Dp indicator for catalyst level OK but bed density incorrect/ pressure surges. “Dp > design”: fines in packed beds/ fines in distributors/ fines in exit nozzles/ crud left in from construction or revamp. [Reactor instability]*: control fault/ poor controller tuning/ wrong type of control/ insufficient heat transfer area. [Regenerator doesn’t remove all carbon from the catalyst]*: damaged air grid/ insufficient air/ excessive regenerator velocity/ poor spent catalyst initial distribution/ coarse particles. [Sintered catalyst]*: local high temperatures/ [maldistribution]*/ for FCCU [afterburn in regenerator]*/ [Feed contaminated]*/ high temperature in the regenerator/ [temperature hot spots in the reactor]*. [Solids conveying lines flow capacity < design]*: sticky fines buildup in lines/ wrong Dp across line. [Stuck or failed trickle valve]*: binding of hinge rings/ angle incorrect/ wrong material/ hinged flapper plate stuck open/ flapper plate missing. [Surging of the catalyst bed]*: water in the feed/ [plugged grid holes]*/ faulty grid design/ [grid holes eroded]* /[ for FCCU: failure of internal seals in regenerator]*/ for FCCU: [seal failures]*/ hole in the overflow well/ [reactor instability]*/ control fault in Dp between cracker and regenerator. [Unacceptable temperature profiles]*: fluctuating temperature/ unsteady bed temperatures. Specific for a fluidized bed combustion/catalyst regenerator: “Increase in catalyst losses”: [poor separation in cyclone]*/ [ failure in regenerator plenum]*/ for FCCU [ failure of internal seals in regenerator]*. [Failure in regenerator plenum]*: faulty cyclone design/ catalyst feed too high/ regenerator velocity too high/ faulty spray nozzles causing impingement of plenum sprays/ temperatures too high causing failure in plenum. Specific for fluid cat cracker unit – including regenerator system: “Overloaded wet gas compressor”: for FCCU high hydrogen production/ increase in production of light ends. “Gas compressor flow reversal”: [poisoned catalyst]*. “Gas compressor surge”: [poisoned catalyst (that causes production of lower MM species)]*. “Gas compressor flow reversal”: [poisoned catalyst]*. “Wet gas compressor surge”: [poisoned catalyst (that causes production of lower MM species)]*. “Hydrogen concentration in wet gas increases”: [poisoned catalyst, especially with nickel and vanadium]*/ [ feed contaminated with metals, especially nickel and vanadium]*/ [loss of catalyst activity]*/ feed concentration high in hydrogen/ loss of antimony solution addition. “Increase in the production of light ends”: for FCCU [ feed contaminated with metals]*/ feed concentration high in light ends. “Erratic or cycling instrument records on holdup, density and the overflow well”: [surging of the catalyst bed]*. “Opaque flue gas from the regenerator”: [poor separation in

103

104

3 Rules of Thumb for Trouble Shooting

cyclone in regenerator]/ fluidization velocity too high/ increase in volume of product through unexpected side reactions/ change in feed flowrate/ flowrate instrument error. “Vibration in the preheat system”: [ feed contaminated with water]*. “Dp increase between reactor and fractionator inlet”: [coking in overhead lines]*. “Dp lower on the regenerator slide valve”: [poisoned catalyst]*. “Dp between cracker and regenerator incorrect”: fault with the input air blower/ fault with the flue-gas slide valve on the regenerator/ fault with the regenerated catalyst slide valve/ fault with the spent-catalyst slide valve/ fault with the wet gas compressor/ fault downstream of the wet-gas compressor, such as plugged fractionator overhead condensers (with ammonium chloride salts)/ changes in environment air conditions. Regenerator should be about 20 kPa higher than the cracker for Dp across the regenerated catalyst slide valve. “Dp between cracker and regenerator fluctuating”: fluctuating temperature in cracker/ fluctuating pressure in regenerator/ fluctuating catalyst circulation rate/ fluctuating level in the overflow well/ shift in catalyst between cracker and regenerator/ incorrect aeration of U-bend/ incorrect aeration of standpipe/ sticky stack slide valves/ sensor control performance for stack slide valve unsatisfactory/ moisture in aeration medium/ unsteady control of air/ U-bend vibration. “Dp across cylcone > expected”: steam flowrate > expected/ air flowrate > expected. “Pressure fluctuating in regenerator”: incorrect aeration of U-bend/ incorrect aeration of standpipe/ sticky stack slide valves/ sensor control performance for stack slide valve unsatisfactory. “Plugged pump on the bottoms of the fractionator”: [poor separation in cyclone]*/ velocity through reactor too high/ faulty cyclone design. “Overflow well level low”: [eroded grid holes]*. “Overflow well level high”: [plugged grid holes]*. “Overflow well level fluctuating”: incorrect aeration of U-bend/ incorrect aeration of standpipe/ sticky stack slide valves/ sensor control performance for stack slide valve unsatisfactory/ hole in the overflow well. “Catalyst loss from the regenerator increased”: [plugged grid holes]*/ [eroded grid holes]*/ foreign debris entering with fresh catalyst/ faulty grid design/[poor separation in cyclone]*/ steam flowrate > design/ air flowrate > design. “Catalyst circulation fluctuates”: [Dp between cracker and regenerator fluctuating]*/ fluctuating temperature in cracker/ fluctuating temperature in regenerator/ fluctuating level in the overflow well/ shift in catalyst between cracker and regenerator/ incorrect aeration of U-bend/ incorrect aeration of standpipe/ sensor control performance for air system unsatisfactory/ moisture in aeration medium/ unsteady control of air/ U-bend vibration/ coarse particles/ hole in the overflow well/ incorrect aeration of U-bend/ incorrect aeration of standpipe/ sticky stack slide valves/ sensor control performance for stack slide valve unsatisfactory/ [surging of the catalyst bed]*. “Catalyst becomes lighter in regenerator gradually”: [afterburn in regenerator]*. “Catalyst in fractionator bottoms”: [poor separation in cyclone]*/ velocity through reactor too high/ faulty cyclone design. “Catalyst has salt and pepper appearance after regeneration”: air grid deficiency/ [Failure of internal seals in regenerator]*. “Reduced rates of spent catalyst withdrawal”: [poor separation in cyclone in regenerator]*. “Temperature difference between bed and cyclone inlet in regenerator”: [ failure of internal seals in regenerator]*/ [afterburn in regenerator]* and [inadequate regeneration]*. “Temperatures of bed and cyclone are uneven in the regenerator”: hole in the over-

3.6 Reactor Problems

flow well/ [plugged grid holes]*/ foreign debris entering with the fresh catalyst/ faulty grid design. “Temperature on regenerator shell or U-bend high”: damaged refractory. “Temperature increase in the dilute phase relative to the dense phase”: [afterburn in regenerator]*. “Temperature of the regenerator cannot be lowered”: [low catalyst circulation rate]*. “Temperatures of regenerator too high > 750 C”: excessive heat release. “Temperature in dilute phase decreases relative to temperature of the dense bed in the regenerator”: [regenerator doesn’t remove all carbon from the catalyst]*. “Feed preheat requirements > usual”: [low catalyst circulation rate]*. “Unexplained increase in coke”: [poor catalyst stripping]*. “High bottom sediment and water levels in the slurry oil product”: [poor separation in cyclone in the cracker]*. “Higher H/C ratio”: [poor catalyst stripping]*. “Excess oxygen in regenerator high”: [afterburn in regenerator]*/ [plugged grid holes]*/ [eroded grid holes]*/ faulty grid design. “Ratio of carbon dioxide to carbon monoxide is higher than usual”: [afterburn in regenerator]*. “Uneven oxygen distribution in the dilute phase”: [Failure of internal seals in regenerator]*. “Unsteady heat balance”: [surging of the catalyst bed]*. “Stripping steam flowrate > expected”: flowmeter error/ steam traps faulty/ partially opened valves/ missing restrictive orifice. “Air flowrate > expected”: flowmeter error/ partially opened valves/ missing restrictive orifice. “Flow reversal with feed going incorrectly to the regenerator”: [Dp across the regenerator slide valve is < design]*. [Afterburn in regenerator]*: for FCCU [ failure of internal seals in regenerator]*/ too much excess air/ oxygen recorder reading incorrect/ meter error for feed and recycle flowmeters/ meter error for cyclone flowmeter/ [insufficient carbon production on catalyst during cracking]*/ air flowrate to regenerator too high/ [plugged grid holes]*/ [eroded grid holes]*/ faulty grid design causing localized air-distribution problem. [Coking in overhead lines]*: insulation missing or damaged on transfer line/ extremely cold/ increase in heavies and condensibles in reactor products. [Control air flowrate too low]*: controller for air faulty or poorly tuned. [Failure in regenerator plenum]*: faulty cyclone design/ catalyst feed too high/ regenerator velocity too high/ faulty spray nozzles causing impingement of plenum sprays/ temperatures too high causing failure in plenum. [Feed contaminated with metals]*: abnormal operation in the upstream atmospheric and vacuum units. [Feed contaminated with heavy hydrocarbons]*: leak in heat exchangers/ partly open valves. [Feed contaminated with light hydrocarbons]*: leak in heat exchangers/ partly open valves. [Feed contaminated with sodium]*: seawater leak in upstream equipment/ treated boiler feedwater leaks into feed/ upset in upstream caustic unit. [Feed contaminated with water]*: water in feed tanks/ leaks from steam-out connections/ steam leaks in tank heaters/ water not cleaned out of the lines at startup/ moist air not removed from lines at startup. [Higher reactor velocities]*: [ feed contaminated with metals]*. [Higher regenerator holdup]*: hole in the overflow well.

105

106

3 Rules of Thumb for Trouble Shooting

[Increased air requirements in regenerator for the same conversion in the cracker]*: [ feed contaminated with heavy hydrocarbons]*/ [inadequate regeneration]*/ [coke on catalyst > usual]. [Insufficient coke production on catalyst during cracking]*: cracking operation intensity is lower than usual/ higher quality of feed to the cracker than usual for FCCU fewer aromatics in feed. [Low catalyst circulation rate]*: partial blockage of the U-bends/ excessive stripping steam/ insufficient aeration/ [control air flowrate too low]*/ differential pressure between cracker and regenerator set incorrectly or fluctuating. [Poor catalyst stripping]*: insufficient steam stripping flowrate/ faulty flow controller on steam flow/ faulty design of stripper/ reactor temperature too low/ faulty contacting between steam and catalyst/ circulation rate too high/ coarse particles. [Dp across the regenerator slide valve is < design]*/ sudden drop in regenerator pressure/ regenerator slide valve sticking partly open/ compressor surge (see Section 3.3.2). [Regenerator doesn’t remove all carbon from the catalyst]*: [excessive coke formed in cracker]*/ low excess oxygen/ oxygen sensor error/ flowmeter error for air/ [poor air distribution]*/ flowmeter error for feed and recycle/ air flowrate too small. [Sodium on catalyst]*: carryover of sodium from upstream units (caustic)/ treated boiler feedwater used in regenerator sprays/ [ feed contaminated with sodium]*. [Uneven oxygen distribution in the regenerator]*: hole in the overflow well/ [plugged grid holes]*/ foreign debris entering with fresh catalyst/ faulty grid design. [Unstable catalyst bed]*: airflow too low/ grid holes eroded/ faulty grid design. 3.6.11

Mix of CSTR, PFTR with Recycle

Conventional activated sludge: Trouble shooting: “Increase in sludge volume index”: high-density inerts in feed and usually caused few operating problems. “Decrease in sludge volume index and “bulking”: high concentration of dissolved organics in feed. “Sludge rises”: excessive nitration. “Frothing”: decrease in aeration suspended solids/ increase in surfactants in feed/ aeration > design/ increase in temperature. 3.6.12

Reactive Extrusion

Trouble shooting: “Inadequate mixing of liquid reactant with polymer”: liquid flowrate too high/ screw channel under injection not full of polymer/ faulty screw design. “Residuals in final polymer > design”: vent temperature too low/ screw speed too low/ polymer feedrate too high/ screw design does not provide enough shear/ vent pressure too high. “Polymer has crosslinked or degraded”: screw rpm too high/ degree of fill too low/ feedrate too low/ heat-zone temperatures set too high/ screw-design fault giving excessive shear/ [screw tip pressure too high]*. “Extruder torque excessive”: throughput too high/ screw speed too low/ heat-zone temperatures set too low/ faulty screw design. “Unable to melt material”: throughput too high/ screw

3.7 Mixing Problems

speed too low/ faulty screw design/ material too slippery. “Residence time too short”: throughput too high/ [degree of fill too high]* / faulty screw design. “Residence time too long”: throughput too low/ [degree of fill too low]* / faulty screw design. “Gels or crosslinked materials”: localized initiator concentration too high/ [melt temperature too high]*. [Degradation of melt in extruder]*: [RTD too wide]*/ barrel temperature too high/ screw speed too high (causing overheating and shear damage)/ oxygen present/[oxidation]*/ nitrogen purge ineffective/ wrong stabilizer/ wrong screw/ flows not streamlined/ stagnation areas present/ extruder stopped when temperatures > 200 C/ copolymer not purged with homopolymer before shutdown/[residence time too long]*. [Degree of fill too high]*: feedrate too high/ screw speed too slow. [Melt temperature too high]*: screw speed too high/exit barrel zone temperatures too high/ screw tip pressure too high/ degree of fill too low/ [shear intensity too high]*/ heat-zone temperatures set too high/ [screw tip pressure too high]*. [RTD too narrow]*: [degree of fill too high]* [Screw tip pressure too high]*: screens plugged/ die or adapter or breaker plates too restrictive and give too much Dp/ [polymer viscosity too high]*/ temperatures in die assembly too low/ barrel temperature too low/ screw speed too high/ [shear intensity too low]*/ lubricant needed/ flow restriction/ throughput too high/ die land too short/ cold start/ [degradation of melt in extruder]*. [Shear intensity too low]*: screw speed too low/ faulty screw design. For other symptoms see Section 3.9.6.

3.7

Mixing Problems

Here we consider mechanical agitation of liquids and liquid-solid systems. Operating information is given for solids blenders, in Section 3.7.3.

3.7.1

Mechanical Agitation of Liquid

Good practice: prefer motionless mixers to intensify. For systems where the viscosity increases with time (ex. polymer reactors) prefer turbines to propellers because turbines are power self-limiting. Check shaft wobble to ensure that impeller will not hit vessel walls if turned on in an empty tank. Consider a foot bearing. Trouble shooting propeller/ impeller mixers: “Shaft wobble/vibration”: impeller speed too close to the 1st critical speed/ shaft runout at the impeller and impeller eccentricity too large/ insufficient support. “Excessive gear-reducer maintenance”: excessive load/ high shock loads/ excessive shaft bending.

107

108

3 Rules of Thumb for Trouble Shooting

“Excessive packing wear”: insufficient lubrication/ excessive shaft wobble/ shaft is out of round. “Failure of the mechanical seal”: dirty lubricant/ not enough lubricant/ excessive shaft wobble. 3.7.2

Mechanical Mixing of Liquid–Solid

Trouble shooting stirred tanks: “Solids floating on the surface”: solids not wetted by liquid/ insufficient vortex. 3.7.3

Solids Blending

Use the Johanson indices to characterize particles: (see also related topic storage bins, Section 3.10) Arching index [m], AI, = diameter of the circular exit hole from a hopper that will ensure that an arch collapses in a conical bin or circular mixer, values range 0–1.2 m; Ratholing index [m], RI, = diameter of the circular exit hole from a hopper that will ensure rathole failure and cleanout in a funnel-flow bin or mixer, values range from 0–9 m. (If RI > 3 then likely “lumps”.) Hopper index [degrees], HI = the recommended conical half-angle (measured from the vertical) to ensure flow at the walls. Usually add 3 to account for variability. Values range 14–33 with 304 s/s. Flow ratio index [kg/s], FRI = maximum solids flowrate expected after deaearation of a powder in a bin. (measures consistency: small FRI for fine, highly compressible particles; Large FRI for particles > 400 mm, incompressible, very permeable.) Values range 0–90 kg/s. Bin density index [Mg/m3], BDI = bulk specific mass expected in a container full of solids or, in a mixer, when mixer stops and solid is allowed to deaerate. Values 0.3–1.6 Mg/m3. Feed density index [Mg/m3], FDI = bulk specific mass at the conical hopper or mixer’s discharge outlet. Values are 1–60% < BDI. Chute index [degrees], CI, = recommended chute angle (with the horizontal) at points of solids impact. Values = angle of slide. values = 20–90. High values suggest particles stick to sides of mixer or bins. Rough wall angle of slide [degrees], RAS = angle (relative to the horizontal) that causes continual sliding on a solid on an 80-grit sandpaper surface with a pressure of 140 kPa gauge. Values 20–35. Approximately equal to the angle of repose. Adhesion angle index [degrees], AAI = difference between angle of slide (with horizontal) after an impact pressure of 7 MPa or (CI–10) and the angle of slide without impact pressure. Spring back index [%], SBI = percentage of solids that spring back after consolidation.

3.8 Size-Decrease Problems

Good practice: usually 300 s blend time. We can invest much effort into mixing solids, but we must prevent demixing after the blends are mixed. The four mechanisms of demixing are 1) sifting, 2) angle of repose, 3) fluidization and 4) air current. Here are the details: 1. demixing via sifting: occurs if the particles are free flowing with mix of particle size with one size > 3 the diameter of the other. 2) demixing via angle of repose: this occurs with moderately free-flowing particles (AI < 0.18 m) with different angles of repose or two different RAS. For particles characterized in this way, the only blender that seems to prevent demixing is the air-pulse blender. 3) Demixing by fluidization: tends to occur if the blend contains > 20% fluidizing fines characterized by AI < 0.18 m; RI < 1.5 m; FRI < 0.76 kg/s plus coarser material with AI < 0.012; RI < 0.6 m and FRI > 7.6 kg/s. This type of demixing tends to occur if the action of the mixer induces air. 4) Demixing by air current: air carriers superfine, easy-flow, non-agglomerating particles into voids. This is a problem if AI < 0.18 m; RI < 1.5 m; FRI < 0.4 kg/s. Again try to avoid mixers whose action induces air. Trouble shooting for polymer blenders of feedstock for extruder: “Material does not flow”: bridging/ see also hoppers, Section 3.10. “Components do not feed”: jammed valve or auger/ solids blockage or bridging/ power fault in feeder. “Inconsistent flowrate”: bridge or block in blender/ jammed discharge mechanism/ inconsistent feedrates to blender. “Wrong blend compositions:” calibration error in feeder.

3.8

Size-Decrease Problems

The focus here is on gas liquid systems that include bubble columns, Section 3.8.1, in packed columns, Section 3.8.2, and in agitated tanks. 3.8.1

Gas Breakup in Liquid: Bubble Columns

Good practice: electrolytes in the liquid alter the bubble diameter, the holdup, the interfacial area per unit volume in mechanically agitated devices and affects the kL a for bubble columns. 3.8.2

Gas Breakup in Liquid: Packed Columns

Good practice: the critical surface tension of the solid packing should be greater than the surface tension of the liquid to ensure that the liquid film remains intact in a packed contactor.

109

110

3 Rules of Thumb for Trouble Shooting

3.8.3

Gas Breakup in Liquid: Agitated Tanks:

Good practice: consider increasing the impeller diameter or using a disk turbine to increase mass transfer. Trouble shooting: “Foaming”: mixer tip speed too high/ linear gas velocity too high/ surfactant contaminants/ decrease in electrolyte concentration in the liquid/ change in pH/ use of turbine impeller/ lack of a gas sparger. “Flooded impeller”: too small a diameter impeller/ speed too slow.

3.9

Size Enlargement

The fundamentals used to solve troubles for size-increase operations are surface phenomena. Surfaces are attracted to each other by van der Waals forces; surfaces are repelled by the electrochemical double layer or by steric hindrance. Surface energies, contact angles, and wetting are important. Specific symptoms and causes for equipment to coalesce drops in gas are listed in Section 3.9.1; for coalesce liquid drops in a liquid environment, Section 3.9.2, and create solid aggregates or flocs in a liquid environment, Section 3.9.3. Then we consider the creation of larger-size particle clusters by tabletting, Section 3.9.4; and pelleting, Section 3.9.5. The last two processes considered in this section focus on change in shape by injection molding and extrusion, Section 3.9.6; and coating, Section 3.9.7. 3.9.1

Size Enlargement: Liquid–Gas: Demisters

Good practice: consider using mesh pads upstream of impact filter beds to reduce the load on the filter bed. Consider installing mesh pads vertically to facilitate drainage and minimize re-entrainment. For non-corrosive and non-fouling, consider installing vane separators downstream of mesh pads to collect larger drops sheared off from the mesh pad. Cannot be used for up to 25% turnup capacity; avoid the use of inertial devices for up to 25% turndown capacity. Trouble shooting: “Demisters ineffective”: temperature too high/ fibers have the same charge as the droplets/ wetting properties of fibers changed/ fibers “weathered” and need to be replaced/ flowrate too slow through fibers/ wrong mix of fibers/ prefiltering ineffective/ [ foaming]*/ wrong design/ re-entrainment. [Foaming]*: see Section 3.11. 3.9.2

Size Enlargement: Liquid–Liquid: Coalescers

Good practice: consider decreasing the temperature to decrease the solubility and increase the surface tension. Adjust pH for water flowing through fibrous and mesh beds so that drop and fiber have the opposite surface charge. Promote coalescence in solvent-extraction systems by using surface tension positive configurations. Trou-

3.9 Size Enlargement

ble shooting: “Coalescer pads ineffective”: temperature too high/ pH incorrect/ fibers have the same charge as the droplets/ surface tension negative system/ wetting properties of fibers changed/ fibers “weathered” and need to be replaced/ flowrate too slow through fibers/ wrong mix of fibers/ prefiltering ineffective/ surface tension < 1 mN/m for fluoropolymer fibers or < 20 mN/m for usual fibers/ wrong design/ included in decanter but should be separate horizontal coalescer promoter unit/ faulty design/ [stable emulsion formed]*. [Stable emulsion formed]*: see Section 3.11. 3.9.3

Size Enlargement: Solid in Liquid: Coagulation/Flocculation

Coagulation and flocculation in general: Trouble shooting: “Supernatant not clear”: [coagulation doesn’t occur]*/ flocculation doesn’t occur]*/ [ floc doesn’t settle out]*/ [ floc forms but breaks up]*. [Coagulation doesn’t occur]*: wrong dosage of coagulant-flocculant/ wrong counterion/ pH different from expectations/ pH far from zpc/ faulty mixing in the rapid mix/ valence on the counterion too small/ charge on the dispersed particles or drops reversed from expectations. [Flocculation doesn’t occur]*: faulty fluid dynamics into the basin/ reel at wrong rpm/ residence time too short/ mixing not tapered/ unexpected turbulence/ too short a residence time between coagulant and subsequent flocculant dosage. [Floc doesn’t settle out]*: floc formed is too loose/ settler fault. [Floc forms but breaks up]*: local turbulence > shear strength of floc. For water treatment: Trouble shooting: “Coagulation-flocculation ineffective, supernatant murky”: pH > 10/ wrong dosage of alum or coagulant / pH < 4/ increase in concentration of particles in feed/ rpm of reels in flocculation basin too slow/ feed temperature < 12 C/rpm of reels in flocculation basin too fast. For latex: Trouble shooting: “Exit crumb too small”: brine concentration too high/ temperature too low/ power input too high/ wrong pH. “Excessive amount of fines in supernatant”: brine concentration too high/ wrong pH/ temperature too low. “Strength of the resulting crumb < specifications”: pH too high and brine concentration too high. 3.9.4

Size Enlargement: Solids: Tabletting

Trouble shooting: “Product tablet weight > design”: sample error/ lab error/excessive fines. 3.9.5

Size Enlargement: Solids: Pelleting

Trouble shooting: For strand pelletizer for polymer resin. “Pellet diameter too small”: hole too small for the desired throughput/ extruder output too low. “Pellet diameter

111

112

3 Rules of Thumb for Trouble Shooting

too large”: output too high/ speed too low/ feedroll speed too low/ output too high for die-size. “Pellet too short or too long”: mismatch ratio of feedroll speed versus rotor teeth speed. “Strands dropping”: feedroll pressure too small/ throughput too low/dieplate has too many holes. “Pellet cuts are angled”: feed not perpendicular to strands/ strands overlapping. “Pellet oval shaped”: feedroll pressure too high/ inadequate cooling before cutting. “Pellet has tails”: incorrect clearance between rotor and cutters. For water-ring pelletizer for polymer resin: “Pellet diameter too small”: hole too small for the desired throughput/ throughput too large. “Pellet diameter too large”: output too low. “Pellet too short or too long”: mismatch throughput versus cutter speed. “Blocked holes”: nonuniform pressure on the die face/ throughput too low/die-plate has too many holes. “Pellet oval shaped”: cutter speed too high/ inadequate cooling. “Pellet has tails”: incorrect clearance between die and cutters/ worn cutter blades. 3.9.6

Solids: Modify Size and Shape: Injection Molding and Extruders

Consider first injection molding machines, then extruders. Injection molding machines Good practice: resin should be dried < 0.02%. Do not use resin that has been out of the dryer for > 20 min. Cold molds are difficult to fill and require higher injection pressures. Hot molds, generally, give better finish and less molded-in stress. Melt temperature is very sensitive to very small changes in rpm or backpressure despite sensor or controller set point. Measure with hand-held pyrometer or laser sensor. Need slower fill rate for sprue-gated parts to prevent blush, splash or jetting. If the walls are > 5 mm, then slow fill helps reduce sinks and voids. Backpressures of 0.35 to 0.7 MPa help ensure homogeneous melt and consistent shot size. As backpressure increases, melt temperature increases. Holding or backpressures that are 0.4 to 0.8 of injection pressures are typical. To purge a machine, acrylic is recommended. Trouble shooting: injection molding: Basically the cause can be with the material, the machine, the operator, the operating conditions, the mold or the part design. To check on the material, try material from another supplier; to check the machine, use same material, conditions, mold on another machine; if the trouble is random, then it is probably the machine; try a different operator on the machine; trouble appears in same location in the product, then flow conditions and look for problems from the front of the piston to the gate. The symptom-cause information is presented as issues related to appearance (color, surface finish and transparency), strength and shape defects and operation and symptoms. 3.9.6.1

.

Appearance (Color, surface finish, transparency)

Color: “Discoloration” (typically appears before burn marks appear; location appears at the weld line or where air is trapped in the mold): [contamination in heating cylinder]*/ sensor error/ control error/ [degradation, mechanical]*/ [degradation, thermal]*/ [melt too hot]*/ [melt not homogeneous]*/ overall cycle too long/ [contamination in

3.9 Size Enlargement

hopper and feed zones]*/ incorrect cooling of ram and feed zone/ [venting in mold insufficient]*/ [residence time too long]*/ cooling time too short/ dryer residence time too long / excessive clearance between screw and barrel / clamp pressure too high/ injection forward time too long/ gate too small/ runner-sprue-nozzle too small. “Black specs inside transparent product”: faulty cleanout of machine from previous molding operations/ failure to purge when not running for extended times/ nozzle too hot/ barrel temperature in the feed area is too low combined with high screw speed or high backpressure/ sensor error/ sensor located too far from heater bands/ hangup in nozzle tip/ nozzle adapter and end-cap. “Brown streaks/burning”: wet feed/ [melt too hot]*/ [shear heating in the nozzle]*/ [degradation, mechanical]*/ loose nozzle/ wrong nozzle/ dead spots in hot manifold/ mold should be cold runner system/ gate or runner too small/ [contamination]*/ injection speed too fast/ booster time too long/ injection pressure too high/ mold design lacks vents at burn location/ gate size too small or at wrong location/ plunger has insufficient tolerance to allow air to escape back around the plunger/ poor part design/[venting of mold insufficient]*/ [residence time too long]*. “Brown streaks at the weld lines or at the end of flow paths, black or charred marks”: [air trapped in mold]*. “Brown streaks at the same location”: nozzle loose, wrong, too hot/ [shear heating (at gate, runner, cavity restrictions)]*. “Brown streaks dispersed throughout”: material fault at the hopper: wet material. “Weld burns”: [melt too hot]*/ injection speed too fast/ [mold too cold]*/ injection hold time too long/ injection pressure too high/ faulty nozzle heating bands/ [air trapped in mold]*. Surface finish: “Sink marks” (difficult to remove by changing processing conditions): [cooling insufficient before removal from mold]*/ [short shot]*/ [melt too hot, causing excessive shrinkage]*/ [solidification at mold wall too slow]*/ wrong location for gate/ holding pressure too low/ injection speed to fast/ backpressure too high/ gates too small/ faulty runners/ booster time too low/ fault with nozzle, sprue or runners/ [mold temperature non-uniform]*/ moist feed/ [thick sections continue to shrink after the melt path is frozen]*/ hold time too short/ holding pressure too low/ [backflow from mold]*/ lubricant insufficient/ volatiles in feed/ [solidification at mold wall delayed]*/ [viscosity too high]*/ cooling-water temperature too low /excessive cushion in front of ram/ size of nozzle, sprue and runner too small/ [air trapped]*. “Fine ridges running perpendicular to the flow front”: [melt too cold]*/ [short shot]*. “Flow lines” see also “jetting”: feed moist/ injection pressure too low/ [melt too cold]*/ screw not rotating during injection/ injection speed too slow/ backpressure too low/ nozzle orifice too small/ [mold too cold]*/ gates too small/ [venting of mold insufficient]*/ feedrate too small/injection rate too low/ injection hold time too low/ booster time too low/ relocate gates/ faults with nozzle, runner, sprue or gate/ clamp pressure too high/ unequal filling rates between cavities/ core position incorrect/ gates too small/ mold design fault with non-uniform thickness of sections or excessive heavy bosses or ribs. “Low gloss, dull or rough surface”: moist feed/ injection pressure too low/ [mold too cold]*/ [melt too cold]*/ injection rate too slow/ relocate gates/ mold cooling time too short/ surface of sprue, runner, or cavity rough/ [contamination]*/ [venting of mold insufficient]*/ screw rpm too low/ injection made without screw rotation/ injection speed too fast/ backpressure too low/ nozzle ori-

113

114

3 Rules of Thumb for Trouble Shooting

fice too small/ increase or decrease mold temperature/ gate size too small/ [melt too hot]*/ diameter or depth of cold slug is too small/ wrong location of gate/ wrong location for water channels/ particle size not uniform/ too many fines in feed/ wrong type of lubricant. “Streaks on part”: stock temperature too high or too low/ screw rpm too high/ nozzle or shutoff valve no tight/ injection speed too fast/ backpressure too low/ cooling and mold-open time too short. “Splay marks: coarse lines or lumps”: [degradation of melt thermally]*/ injection rate too fast/ increase or decrease the mold temperature/ screw decompression too long/ overall cycle too long/ [contamination, fluid]*/ [shot size too large]*/ [drooling]*/ screw decompression is missing from the molding cycle/ gates too small/ fault in the hot-runner system/ nozzle orifice too small/ sprue and runner size too small/ gate not perpendicular to runner. “Splay marks: fine lines”: wet feed/ residual non-aqueous volatiles in feed. “Blush at the gate, dull spot in the part at the gate”: moist feed/ [melt fracture at the gate]*/ [mold too cold]*/ injection pressure too low/ [melt too hot]*/ injection speed too fast/ injection hold time too short/ nozzle diameter too small/ gate land area too large/ diameter of sprue, runner and /or nozzle too small/ depth or diameter of cold slug too small/ wrong location of gate/ [venting of mold insufficient]*. “Silver streaks”: moist feed/ [nozzle or cylinder too hot]*/ plasticizing capacity, in kg/s, of machine is exceeded/ variation in temperature of feed in hopper/ plastic temperature is too high/ injection pressure too high/ air trapped between granules in the cold end of the machine/ [mold too cold]*/ injection speed too fast/ lack of or excessive external lubrication/ feed is mixture of course and fine particles as with reground/ rear cylinder temperature too high/ [venting of mold insufficient]*/ gates not balanced or at wrong location/ insufficient addition of zinc stearate when using reground/ [air trapped in melt]*/ [degradation of melt, thermally]*/ [air trapped in mold]*/ [cold slugs at the nozzle or hot tip]*/ [contamination]*/ faulty mold design with too many sharp corners or edges. “Drag marks”: rough surface of mold/ injection pressure too high/ injection hold time too long. “Worn tracks on part”: [melt too cold]*/ [nozzle too cold]*/ screw rpm too low/ injection speed too fast/ backpressure too low/ nozzle orifice too small/ gates too small/ cold slug well too small. “Jetting” dull spots and disturbances that look like a jet: moist feed/ [melt too cold]* / [mold too cold]*/ injection rate too fast/ nozzle diameter too small/ depth or diameter of cold slug too small/ diameter of sprue, and runner too small/ wrong location of gate, incorrectly at a thick section / [nozzle too cold]*/ gate too small/ gate land length too long. “Wave marks”: feedrate too small/ injection pressure too low/ [melt too hot]*/ [mold too hot]*/ clamp pressure too low/ injection hold time too short/ cycle time too short/ wrong location of water channels/ stock temperature either too hot or too cold/ injection speed too fast or too slow/ nozzle diameter too small/ moist feed. “Flashing”: the flow of material into unwanted areas; if at the end of the flow paths then its cause is usually [shot size too big]; flash in the runner system may indicate continued holding pressure after the gates freeze off: [melt too hot]*/ injection pressure too high/ injection hold time too long/ injection speed too fast/ clamp pressure too low/ [mold too hot]*/ rework the mold design/ vents too deep/ damaged mold/ misaligned platen/ wet feed/ [shot size too big]*/ [ feedrate into mold too high]* / erratic feed/ [design of part faulty]*/ erratic cycle time. “Weld lines, knit lines”: injection rate

3.9 Size Enlargement

too small/ injection pressure too low/ injection hold time too short/ [mold too cold]*/ [melt too cold]*/ vent missing in location of weld/ overflow well missing next to the weld area/ wrong gate location/ too much filler. Transparency: “Cloudiness or haze for clear plastics”: [contamination]*/ moist feed / [melt too cold]*/ faulty adjustment of barrel temperature profile/ injection pressure too low/ backpressure too low/ [mold too cold]*. “Bubbles in clear plastics”: moist feed/ [melt too hot]*/ injection pressure too low/ injection rate too fast/ injection hold time too short/ booster time too low/ [mold too cold]*/ mold cooling time too short/ [cooled too fast]*. .

Strength or Shape Defects

“Voids”: [short shot]*/ [mold: external surfaces solidify and shrinkage occurs internally]*/ [thick sections continue to shrink after the melt path is frozen]*/ injection rate too fast/ [melt too hot]*/ booster time too short/ molding cooling time too short/ [cooled too fast]*/ feed moist/ insufficient blowing agent. “Blisters”: feed moist/ injection pressure too low/ backpressure too low. “Lamination, peeling”: moist feed/ [mold too cold]*/ [melt too cold]*/ injection speed too fast/ nozzle diameter too small/ gate land area too large/ depth or diameter of the cold slug is too small/ diameter of sprue, runners and or nozzle too small/ [contamination]*/ backpressure too low/ injection made without screw rotation/ screw rpm too low/ [nozzle too cold]* / injection rate too low. “Warpage, part distortion”(usually caused by non-uniform shrinkage as the molded part cools from ejection temperature to room temperature): incorrect differential in mold temperatures to account for geometry or mold design/ incorrect handling after ejection/ injection hold time too short and stopped before gate freezes/ cooling time too short/ injection pressure too high or too low/ [mold too cold]*/ shrink fixtures and jigs to promote uniform cooling are missing/ wrong gate locations and too few/ gates too small/ faulty part design/ uneven cooling system on molds/ injection pressure too high/ [melt too cold]*/ holding pressure and time too long/ screw not rotating with injection done/ injection speed too fast/ backpressure too low/ mold temperature either too hot or too low/ time for cooling and mold-open are too short. “Weld weak”: [mold too cold]*/ injection speed too slow/ [melt too cold]*/ injection pressure too low/ nozzle opening too small/ gate land area too large/ sprue, runner or gate size too small. “Brittle”: feed moist/ [mold too cold]*/ injection rate too fast/ [melt too cold]*/ injection pressure too high/ gate diameter too small/ nozzle orifice too small/ not enough gates or gates at wrong location/ injection pressure too low/ gate land area too large/ sprue, runner or gate size too small/ gates too small/ cold slug well too small/ holding pressure and time too long/ screw rpm too low/ screw should rotate during injection/ injection speed too fast/ backpressure too low/ mold temperature either too high or too low/ sensor error/ [degradation material]*/ [stress high in part]*/ faulty mold design with notches causing local stress. “Dimensional variation”: faulty feedrate/ [melt too cold]*/ injection pressure too low/ [mold too cold]*/ injection hold time too short/ injection speed too fast/ cycle time too short/ nozzle diameter too small/ faulty gate location/ gate land area too large/ diameter of sprue, runner and or nozzle too small/ incorrect location of water channels/ [control of machine faulty]*/ [mold con-

115

116

3 Rules of Thumb for Trouble Shooting

ditions wrong]*/ poor part design/ moist feed/ irregular particle size/ batch to batch variation in feed. “Cracking”: [mold too cold]*/ wrong mold design/ ejection pins poorly located and give unbalanced push/ [shot size too large]*. “Low heat distortion temperature”: [variation in section thickness]*/ [mold too cold]*/ mismatch between cylinder and mold temperatures/ feedrate too high/ pressure too high/ plunger dwell too long/ excessive temperature variation between front and back of mold/ freezing in the gate because gate orifice too large. .

Operation

“Sticking in cavity”: [mold too hot]*/ [melt too hot]*/ injection pressure too high/ injection hold time too long/ injection hold time too short/ gate land area too large/ diameter of sprue, runner or gate too small/ mold surface is rough/ injection speed too high/ faulty mold design/ incorrect radius of nozzle and sprue bushing/ mold release not used/ air was not provided for ejection/ hold pressure too high/ feed not adjusted to provide a constant cushion/ cooling time too long or too short/ cavity or core temperatures do not have the < 7 C differential between mold halves/ nozzle too hot/ mold has undercuts and insufficient drafts. “Sticking parts”: [mold too hot]*/ injection pressure too high/ rough surface on mold/ holding pressure too high/ wet feed/ cooling time too short/ faulty design of ejector/ highly polished or chrome-plated mold surfaces. “Sticking in sprue bushing”: injection pressure too high/ injection hold time too long/ booster time too long/ cooling time too short/ mold is too hot at the sprue bushing/ nozzle pulled back from mold/ [nozzle too cold]*/ incorrect seat between the sprue and mold/ nozzle orifice is not 7.5 mm smaller in diameter than OD of sprue/ rough surface on sprue/ sprue puller ineffective/ sprue does not have sufficient draft angle for easy release/ screw decompression too low or missing. “Runner breaks”: holding pressure and time too long/ [mold too hot]*/ sprue, runners and gates are rough/ incorrect radius in the nozzle and sprue bushing/ time for cooling and open time too short. “Discoloration of sprue”: [melt too hot]*/ nozzle or shutoff valve not tightened/ injection speed too high/ nozzle orifice diameter too small/ [mold too cold]*/ cold slug well too small/ gate diameter too small. “Drooling at nozzle”: shutoff valve dirty or clogged/ injection too soon/ wrong nozzle pressure/ poor radius of nozzle and sprue bushing/ [nozzle too hot]*. “Screw does not return”: screw rpm too low/ backpressure too high/ wet feed/ hopper out of feed/ obstruction/ temperature in the rear zone too high. “Ejection of part poor”: rough mold walls/ [shot size too large]*/ knockout system inadequate/ insufficient taper. “Cycle erratic”: operator/ [pressure erratic]*/ [ feedrate erratic]*/ [cylinder temperature cycles]*. “Cycle too long”: [cooling cycle too long]*/ [heating cycle too long]*/ [operator issues]*/ material should be more heat resistant. .

Symptoms

[Air trapped in melt]*: screw decompression/ backpressure too low. [Air trapped in mold]*: [venting of mold insufficient]*/ gate diameter too small/ [mold too hot]*. [Backflow from the mold]*: suckback/ faulty non-return valve. [Backpressure too high]*: injection rate too fast.

3.9 Size Enlargement

[Barrel too hot]*: melt temperature > 271 C/ sensor error/ faulty barrel heater control system/ worn or incorrectly fitted screw and barrel configuration. [Contamination]*: dirty machine/ dirty hopper/ moist feed/ too many volatiles in feed/ [degradation]*/ lubricant or oil on mold/ incorrect mold lubricant/ feed contaminated during material handling/ faulty raw material from supplier/ poor shutdown procedures. [Contamination, fluid]*: water or oil leaking into mold cavity. [Control of machine faulty]*: incorrect screw stop action/ inconsistent screw speed/ malfunction of non-return valve/ worn non-return valve/ uneven control of backpressure/ faulty temperature sensor/ heater band faulty/ control system fault or poorly tuned/ machine has inadequate plasticizing capacity/ inconsistent control of cycle. [Cooling cycle too long]*: [melt too hot]*/ [mold too hot]*/ inadequate cooling in local heavy sections. [Cooling insufficient before removal from mold]*: faulty mold design especially for rib design/ injection speed too slow/ injection hold time too short/ injection pressure too low/ melt too hot/ mold too hot/ [venting of mold insufficient]*/ sprue and runners too small diameter/ gate too small/ gate land length too long/ gate not close to thicker areas/ core missing from heavy section. [Cooled too fast]*: [mold too cold]*/ mold cooling time too short. [Cylinder overheated]*: nozzle too hot/ cylinder temperature too high. [Cylinder temperature cycles]*: controller fault/ sensor error/ incorrect line voltage/ power factor problems/ heater bands faulty/ variation in feed temperature. [Degradation, mechanical]*: barrel temperature in the feed area is too low combined with high screw speed or high backpressure/ short transition section in screw/ radius between the screw root and the flighter is too small/ small tolerance between the plunger and the wall/ fine material trapped between the plunger and the wall/ excessive reground/ rear cylinder temperature too low/ plunger off-center. [Degradation in the extruder of melt thermally]*: temperature sensor error/ [melt too hot]*/ temperature controller fault/ improperly designed or defective non-return valve. [Design of part faulty]*: incorrect mold dimensions/ unequalized filling rate in cavity/ mold not sealing because of flash between surfaces/ [venting of mold insufficient]*/ vents too large/ gate land area too large/ runner, sprue and gate dimensions incorrect. [Drooling, introduces solid material into part giving defects]*: wet feed/ [melt too hot]*/ suckback pressure too low/ injection pressure too high/ injection forward time too long/ injection boost time too long/ shutoff valve dirty or clogged/ injection too soon/ poor radius of nozzle and sprue bushing/ [nozzle too hot]*. [Feed rate erratic]*: feeding mechanism/ bridging in hopper/ hopper design fault. [Feed rate into mold too high]*: injection feedrate too fast/ feed setting too high/ sensor error. [Flow of polymer into the cavity uneven during high velocity flow into an open area]*: injection rate too fast/ faulty gate location/ gate too small.

117

118

3 Rules of Thumb for Trouble Shooting

[Granules not melted]*: plastic temperature too low/ cycle too short for cylinder capacity/ nozzle diameter too large. [Heating cycle too long]*: insufficient heating capacity. [Injection too slow]*: screw rpm too high/ backpressure too high/ injection speed too slow/ injection pressure too low/ injection forward time too short/ booster time too short/ cycle too short. [Insufficient plastic in mold]*: thick sections, bosses, ribs/ not enough feed/ injection pressure too low/ plunger forward time too short/ unbalanced gates/ piece ejected too hot/ variation in mold open time/ no cushion in front of injection ram with volumetric feed. [Melt not homogeneous]*: backpressure too low. [Melt too cold]*: sensor error/ control system error/ lack temperature confirmation via hand-held pyrometer or laser sensor/ cylinder too cold/ screw rpm too low/ backpressure too low/ insufficient plasticizing capacity of machine/ [nozzle too cold]*/ heating band fault/ excessive flow length in mold. [Melt too cold at the nozzle or hot tip]*: nozzle too cold/ temperature sensor error/ too few heater bands/ heater bands too far from nozzle tip/ hot tip heat source too far from orifice or faulty/ sharp corners near the gate. [Melt too hot]*: sensor error/ control system error/ lack temperature confirmation via hand-held pyrometer or laser sensor/ cylinder too hot/ screw rpm too high/backpressure too high/ [mold too hot]*/ [nozzle too hot]*/ injection rate too slow/ gate too large/ gate land too short/ resin too hot/ holding pressure and time too long/ moist feed/ cooling and mold-open time too long/ [residence time too long]*. [Melt too hot, localized overheating]*: [barrel too hot]*/ faulty barrel heater control system/ [nozzle too hot]*/ sensor error/ faulty or incorrectly designed check valve/ worn or incorrectly fitted screw and barrel configuration. [Melt fracture at the gate]*: [melt too cold]*/ temperature sensor error/ injection rate too fast/ gate too small/ sharp edge in gate area/ cold slug well in the runner too small. [Mold conditions wrong]*: [mold temperature non-uniform]*/ injection pressure low/ injection forward time too short/ injection boost time too short/ [barrel too hot]*/ [nozzle too hot]*/ inconsistent control of cycle. [Mold temperature non-uniform or erratic]*: poorly designed water or coolant lines/ [venting of mold insufficient]*/ coolant supply fault. [Mold: external surfaces solidify and shrinkage occurs internally]*: [mold too cold]*/ mold includes sections that are “too thick”/ [melt too cold]*. [Mold too cold]*: cooling lines in wrong location/ coolant too cold/ coolant flowrate too high/ sensor error. [Mold too hot]*: cooling lines in wrong location/ coolant too hot/ coolant flowrate too low/ sensor error. [Non-uniform shrinkage as the molded part cools from ejection temperature to room temperature]*: wrong packing times/ wrong packing pressures/ wrong gate location/ cooling system fault/ temperature sensor fault/ need separate temperature adjustment for mold halves.

3.9 Size Enlargement

[Nozzle too hot]*: sensor error/ control error/ temperature setpoint at nozzle too hot/ localized heater bands on the nozzle instead of being spread along the nozzle. [Operator issues]*: slow setup of mold/ need to trim “flashing”/ poor monitoring of cycle times/ excessive machine dead time. [Premature gate freeze-off ]*: gate size too small. [Pressure too low]*: injection pressure too low/ loss of injection pressure during the cycle/ feed control set too high causing lower injection pressure. [Pressure too high]*: injection pressure too high/ injection time too long/ boost time too long. [Pressure erratic]*: sensor error/ control system tuning fault/ leaks in the hydraulics. [Residence time too long]*: machine provides shot size that is too large/ dead spots in hot manifold/ temperature too high/ poorly designed manifold system/ [contamination]*. [Resin feedrate too low]*: no material in the hopper/ hopper throat partially blocked/ feed control set too low/ faulty control of feed system/ bridging in the hopper/ faulty hopper design. [Shear heating of melt]*: injection rate too fast/ injection pressure too high/ gates too small/ nozzle orifice too small < 0.8 of sprue bushing/ nozzle dirty/ sharp corners/ injection rate too fast/ shutoff nozzle used instead of a general purpose nozzle/ improperly designed or defective non-return valve. [Short shot]*: [resin feedrate too low]*/ injection pressure too low/ [mold too cold]*/ injection speed too low/ [melt too cold]*/ injection hold time too short/ cycle time too short/ diameter of gate, sprue, and runner too small/ nozzle orifice too small/ gate land length too long/ incorrect gate location/ [venting of mold insufficient]*/ [nozzle too cold]*/ nozzle dirty/ shutoff valve dirty/ inject with screw notrotating/ machine undersized for the shot required/ cycling from wet to dry resin/ excessive flow length in mold/ excessive feed buildup in cylinder/ [mold temperature non-uniform]*/ [air trapped in mold]*/ not enough external lubricant/ poor balance of plastic flow into multiple cavity mold/ holding pressure too low. [Shot size too large]*: resin feedrate too high/ injection pressure too high/ machine shot size much larger than mold requirement. [Shrinkage excessive]*: [melt too hot]*. [Solidification at the mold wall delayed]*: [mold too hot]*. [Stress high in part]*: [mold too cold]*/ [melt too cold]*/ injection pressure too high/ faulty post-mold conditioning/ faulty mold design. [Thick sections continue to shrink after the melt path is frozen]*: faulty mold design, with too much variation in part cross section/ [premature gate freeze-off ]*. [Venting of mold insufficient]*: injection rate too fast/ booster time too long/ injection pressure too high/ vents plugged/ not enough vents/ clamp pressure too high/ wrong location of gates relative to vents/ [melt too hot]*/ [mold too hot]*. [Viscosity of melt too high]*: [melt too cold]*/ wrong resin.

119

120

3 Rules of Thumb for Trouble Shooting

Extruders for Polymers Good practice: for startup, the barrel heaters are critical because screw is not rotating. Major concerns about cold start. Rear-barrel temperature usually remains important because it affects the “bite” or rate of solids conveyed in the feed. Barrel temp. must be set appropriately for polymer. Head and die temperatures = desired melt temp (except where want gloss, flow distribution or pressure control). Screw speed is changed by reducing the motor speed by one of three options: 1) 10 to 20 in two stages: either pair of gears or pulley but second stage is always gears with the screw set in the middle of the last big “bull” gear. For very slow-moving extruders (e.g. twins for rigid PVC) there are usually three stages of reduction to get to < 30 rpm. Most extruder drives are constant torque with max power only available at top screw speed with the reduction ratio sometimes mismatched to the job. For maximum solids conveying “Stick to the barrel and slip on the screw”. Most plastics normally slip on the root of the screw as long as the feed temperature < melt temperature with those that are most likely to stick being highly plasticized PVC, amorphous PET and certain polyolefin copolymers. For amorphous PET covert the feed to less-sticky semicrystalline form by heating to high temperature for at least an hour in an agitated hopper. Particles must stick to the barrel; trouble occurs with a “slippery feed” such as HDPE and fluoroplastics. Material is the biggest cost (usually 80%) so reuse as much trim and scrap as possible and keep close thickness tolerances so as not to have excessive thickness. Shear rate is important because this affects viscosity. All common plastics are “shear thinning” eg. PVC flow is 10 faster if double the push but LLDPE flow increases 3 to 4 times for double the push. Single screw: typically operated 100% filled. Usually flood feed. Twin screw: typically operate 20–100% filled. Cannot be flood fed if running at high speeds. Twin screw with vent: melt seal is about 1 L/D upstream of the vent; feed screw section under the vent operate < 0.5 full. During startup increase the vacuum gradually. Use low degree of fill. Coating wire and cable: preheat the wire to about 120 (for HDPE) to 175 C (cellular PE) to minimize shrinkage. Trouble shooting: Extruders: the approach usually is to 1) adjust the temperature profile, 2) check the hardware such as the thermocouples, controllers, speed, 3) alter the processing conditions or 4) change the resin or the screw and barrel design. The symptom-cause information is presented as issues related to production, off-spec thickness or shape, off-spec strength, off-spec surface features, and usual symptoms. 3.9.6.2

.

Production: “Throughput < design”: [low bulk density of feed]*/ wrong screw design/ worn screw or barrel elements/ screw rpm too low/ wrong temperature set points/ caking on the feed screw/ caking on the feed port. “Slow and steady reduction in throughput”: buildup of contaminants on screen pack. “Torque > design”: feedrate too high/ [degree of fill too high]*/ screw speed too

3.9 Size Enlargement

.

low/ heat zone set points too low/faulty screw design.”Machine stalls above a certain speed or with certain materials”: constant-torque drive (magnetic clutch)/ AC-DC drive system with constant-drive and constant torque combination. “Feed from hopper not feeding smoothly”: material too light and fluffy for gravity feed/ material damp/ bridging/ screw channels in the feed zone are not deep enough/ too much external lubricant. “Drive amps > design”: polymer viscosity too high/ screw pumping too high/ screw speed too high/ barrel temperatures set too low. “Amps high for melt pump drive or pump won’t rotate”: shear pin/ degraded polymer caught in gears. “Variation in drive amps”: [solids conveying instabilities]*/% regrind too high/ feed bulk density wrong. “Cycling motor amps”: [surging]*. “Extruder noisy”: loss of feed/ foreign or metal contaminant in feed/ bent screw/ bent barrel/ 12 heater burned out. “Local temperature fluctuations with cycles < 5 min”: instrument circuitry fault/ inconsistent melt/ poor heater contact/ thermocouples poorly seated/ sensor error/ poor sensor location/heating element fault/ controller fault/ [solid conveying instabilities]*. “Barrel temperatures differ from the set temperatures”: controller fault/ burnt-out heater/ blue screw syndrome where the rear end bites off more than the front end can pump. “Real wall temperature > the set point”: the rear end bites off more than the front end can pump. “Variations in melt pressure”: drift or cycle time variation > 1 min: [low feeding efficiency]*/ low friction characteristics/ [low bulk density of feed]*/[melting too soon]*/ adequate early barrel pressure but [melting unstable]*/first barrel heating too high/screw tip pressure too low. “Screw tip pressure too low”: no resin in feed hopper/ bridging in feed hopper/ temperature too high in extruder entrance zone/ polymer wrapped around screw. Related topic [Screw tip pressure too high]*. “Unstable melt pressure”: screw speed too high/ degree of fill too low/ screw design gives mixing of melt inadequate or low shear/ cycling control on heat zones/ feeder problems. “Unstable pumping”: for vented extruders: first stage [surge]*/ poor screw balance between stages. “Material flow out vent”: for vented extruders: poor vent-diverter design/ first stage pumping rate too fast/ screw tip pressure too high/ flowrate > design/ temperature set points for last barrels too low or heaters faulty/ screw design gives localized pressure under the vent/vacuum too high/ if adding liquids, then poor mixing. Product thickness or shape does not meet specifications. “Small size variation”: variation in drive speed/ wrong screw design/ variation in puller. “Large size variation”: [surging]*. Here are more specific details: “Variation in thickness in transverse direction and always in the same place”: see “Variation in local temperature”. “Variation in thickness in transverse direction and floating across or around the product”: [mixing of melt inadequate]*/ die temperature setting wrong/ dirty die/[screw tip pressure too low]*/backpressure on extruder < 20 MPa/ wrong design or die or screw/ [degradation of melt in extruder]*/ for blown film: air ring not centered or level/ thermocouple error/ bubble subjected to hot or cold air/ polymer feed has > 6 C variation in melt temperature. “Waviness or ridges around the circumference”: [surging]*/ nonuniform

121

122

3 Rules of Thumb for Trouble Shooting

water cascade/ uneven take-off speed/ vibration in the take-off equipment. “Variation in thickness in the direction of extrusion”: [surging]*/ puller slip or incorrect control of tension/ drawdown too much or too fast/ poor alignment/variation in take-up reels/ erratic variation in feed materials/ hot-lips controller cycling/ untuned controller/ faulty controller/ temperature variation in die/ variation in motor load/ variation in melt pressure/ damaged orifices in die or feedblock/ holes in die too large/ incorrect barrel temperature profile/ faulty adjustment of die/ faulty screw design/ plugged screen pack/ temperature sensor fault in barrel/ hopper bridging/ throughput too high/ gels/ for blown film: inconsistent nip roll speed control/ frost line too low/ polymer melt temperatures too low/ bubble-cooling control fault/ variation in air flow from blower/ cooling air nonuniform/ gap opening too large/ cooling air flowrate too low. “Periodic variation in thickness in direction of extrusion”: spinneret temperature too low/ orifice wrong diameter/ wrong drawdown ratio. “Cyclical variation in thickness in the direction of extrusion”: [draw resistance instability]*. “Filament breaks”: [surging]*/ some die holes blocked/ temperature variation in head or die/ melt temperature too low or too hot/ drawdown too great/ holes in die too large/ moisture/ [contamination]*/ [melt too hot]*/ gap between bath and die too large. “Wrong filament shape: correct cross section but too large”: too little pull/ draw distance from die to take-off is too short/ take-off speed too slow/ die-land length too short/ melt temperature too low. “Wrong filament shape: distorted cross section”: unequal die temperatures/ die incorrect shape. “Wrong filament shape: size OK but warped”: cooling too intense / linear take-off speed too fast. “Filament oval cross section”: filaments too hot while passing over rolls/ rolls too hot/ die holes oval/ temperature gradients in die/ tension too high in take-up rolls. “Holes in blown film or coating”: moisture in resin/ die lip gap too large/ air gap too small/ vacuum too high; for coating: moisture/ substrate too rough/ coating thickness too thin/ contamination/ decomposition/ compound temperature too high/ see also “Gels”. .

Product does not meet strength specifications. “Product strength < specs for all samples”: faults with the feed/ [degradation of melt in extruder]*. “Product strength < specs for some samples”: [contamination]*/faults with the feed/ [degradation of melt in extruder]*”. Pipe strength < specs”: melt temperature too low/ throughput too fast/land length too short/ air gap too short/ excessive drawdown at cold temperatures/ too much scrap in feed/ moisture in resin/ dirty metal surfaces/ material sticking on extruder parts/ short die-land length/ high internal angular discontinuities into the die-land section/ linear extrusion speeds excessive/ uneven water coolant cascade/ misaligned sleeve/[mixing of melt inadequate]*. “Stiffness < design”: for cast or sheet: chill roll temperature too low/ low resin density. For coatings: “Poor adhesion”: a variety of apparently contrary causes related to polymer viscosity [polymer viscosity too high]*, degradation especially [oxidation]*, tackiness, temperature:

3.9 Size Enlargement

.

melt temperature too low or high/ air gap too small/ chill roll temperature too low or hot/ line speed too fast/ poor match between coating and substrate/ substrate problems. “Low tenacity”: ratio of roll speeds too small/ [degradation of melt in extruder]*/ wrong resin/ nicks in die. “For wire and cable: covering separates from wire/adhesion”: wire not preheated hot enough/ melt temperature too low/ dirty or moist wire/ [degradation of melt in extruder]*/ cooled too fast/ air-cooling gap too short/ air trapped between wire and coating [trapped air]*. “Low modulus of elasticity”: melt temperature too low/ air gap distance too short. “Interfacial instability for coextruded film”: excessive shear stress at the die gap > 0.06 MPa/ throughput too high/ die gap too narrow, melt temperature too low/ polymer viscosity too high/ the relative velocities where polymer flows combine differ by > 4: 1. Appearance: gloss, fisheyes, shark skin, pits, holes, clarity. Some are surface effects, such as shark skin, regular, ridged, surface deformity with ridges perpendicular to extrusion direction. Others are defects of the whole body of extrudate, caused by [melt fracture]*. These include spiral, bamboo, regular ripple. “Rough surface or dullness”: [contamination]*/ moisture/ linear speed too fast or screw speed too fast/ die holes too small/ die temperature too low/ die land too short/ [mixing of melt inadequate]*/ [melt fracture]*/ no vents used/ hopper vacuum inadequate/ [screw tip pressure too low]*/ discontinuity in the melt flowlines / low melt temperature/ dirty metal surfaces/ material sticking on extruder parts/ uneven water coolant cascade/ misaligned sleeve/faulty screw design/ screw too hot/ extrudate too hot in the coolant bath causing boiling. “Pits on surface”: [contamination]*/ moisture/ water sprays onto extrudate just after exiting the die/ water bath too hot. “Fisheyes in film”: moisture/ damp polymer/ too many volatiles in polymer. “Gels”: [contamination]*/[degradation of melt in extruder]*/ [shear intensity too low]*/ [screw tip pressure too low]*/ number and density of the screen pack too low/ moisture too high/ screw speed too low/ incompatible blend/ [residence time too long]*/ lack of streamlines in extruder/ incorrect startup procedures/ [melting inadequate]*/ [melt too hot]*/ for reactive: localized initiator concentration too high. “Shark skin”: [melt fracture]*/ die temperature too low at the land end/ linear extrusion speed too high/ throughput too high/ viscosity of polymer too high/ MWD of polymer too narrow/ lubricant additive missing/ [shear intensity too high]*/ die gap too small.

“Polymer buildup on die”: melt temperature too low/ throughput too high/ die gap too small/ wrong screw design/ low level of antioxidants. “Porous or bubbles in product”: poor melt quality at vent/ plugged vent opening/ insufficient vent vacuum/ excessive volatiles in feed/ screw speed too high/ vacuum vent needed. “Spotted, warped or pocked surface”: [mixing of melt inadequate]*/ moisture/ roll too cold/ contamination/ screen size too large/ dirty die/ [trapped air]*/ dirt on rolls/ drafty air/ wrong tension/ boiling on extrudate in cooling bath. “Lines on the product”: surface scratches on tip or die/ local buildup/ [die swell too high]*/ throughput too high/ polymer adhesion on channels, tip or die/ incorrect contact in

123

124

3 Rules of Thumb for Trouble Shooting

the quench tank/ melt temperature too low/ throughput too fast/land length too short. “Indented pock marks on pipe after water cooling”: coolant water spray velocity too high. “Raised pock marks on pipe”: water drops on surface in the air-drying zone. “Discolored material”: temperatures too high/ wrong formulation/ discontinuities inside extruder. .

Symptoms

[Contamination]*: contaminated feed/ contaminated additives/ dirty die/ polymer on die lips. [Degradation of melt in extruder]*: [RTD too wide]*/ barrel temperature too high/ screw speed too high (causing overheating and shear damage)/ oxygen present/[oxidation]*/ nitrogen purge ineffective/ wrong stabilizer/ wrong screw/ flows not streamlined/ stagnation areas present/ extruder stopped when temperatures > 200 C/ copolymer not purged with homopolymer before shutdown/[residence time too long]*. [Degree of fill too high]*: feedrate too high/ screw speed too slow. [Die swell too high]*: tip too short/abrupt change in flow near tip or die/ melt temperatures too low in die assembly. [Draw resistance instability]*: for blown film, fiber spinning, blow molding: draw ratio too high. [Extrusion instabilities]*: screw speed too high/ screw temperature too high/ barrel temperature at delivery end too high/ channel depth too high in the metering section/ the length of the compression section too short/ read barrel end temperatures too low/ diehead pressure is too low. [Feedrate too high]*: screw speed too fast/ feed from hopper too fast. [Gels]*: [contamination]*/[degradation of melt in extruder]*/ [shear intensity too low]*/[screw tip pressure too low]*/ number and density of the screen pack too low/ moisture too high/ screw speed too low/ incompatible blend/ [residence time too long]*/ lack of streamlines in extruder/ incorrect startup procedures/ [melting inadequate]*/ [melt too hot]*/ for reactive: localized initiator concentration too high. [Low bulk density of feed]*:% regrind too high/ grind too coarse. [Low feeding efficiency]*: low friction characteristics. [“Melt fracture” where the critical shear stress of polymer (about 0.1 to 0.4 MPa) exceeded in the die; excessive shear stress at the wall > 0.1 MPa]*: exit speed at the die is too fast/ melt too cold/ throughput excessive/ die land too short/ die opening too small/ entrance to die not sufficiently streamlined/ screw speed too high/ MM and melt viscosity too high/ cross-sectional area in exit flow channel too small/ external lubricant additive missing. [Melt too hot]*: screw speed too high/exit barrel zone temperatures too high/ degree of fill too low/ [shear intensity too high]*/ heat-zone temperatures set too high/ [screw tip pressure too high]*. [Melting inadequate]*: barrel or die temperature too low/ screw speed too fast/ screw design gives insufficient mixing/ [shear intensity too low]*/ feedrate too high/ material too “slippery”/[degree of fill too high]*/ additional component has too low a melting point/[residence time too short]*.

3.9 Size Enlargement

[Melting too soon]*: wrong bulk density of feed. [Melting unstable]*: especially for screws with high compression ratio and short compression length: insufficient melt capacity/ too large a channel depth in the metering section/ temperature in the metering end of the screw too high/ wrong screw design. [Mixing of melt inadequate]*: [screw tip pressure too low]*/ [ feedrate too high]*/ screw speed too high and [residence time too short]*/ screw speed to low and [shear intensity too low]*/ [degree of fill too high]*/ faulty screw design for mixing/ temperature set points incorrect/ instrument error in temperature sensors/ temperatures too high/ loading excessive for one component/ no static mixer included/ for reactive extrusion: liquid flowrate too high/ screw channel under injection not full of polymer. [Oxidation]*: temperature too high/ screw speed too low/ [residence time too long]*/ oxygen present/ nitrogen purge ineffective/ antioxident stabilizer ineffective/ [trapped.air]*/ hopper vacuum inadequate. [Polymer viscosity too high]*: temperature too low/ wrong blend/ [shear intensity too low]*. [Residence time too short]*: screw speed too high/ too much feed/ [degree of fill too high]*/ poor screw design. [Residence time distribution, RTD, too wide]*: [degree of fill too low]*/ feedrate too small/ screws speed too fast. [Screw tip pressure too high]*: screens plugged/ die or adapter or breaker plates too restrictive and give too much Dp/ [polymer viscosity too high]*/ temperatures in die assembly too low/ barrel temperature too low/ screw speed too high/ [shear intensity too low]*/ lubricant needed/ flow restriction/ throughput too high/ die land too short/ cold start/ [degradation of melt in extruder]*. [Shear intensity too low]*: screw speed too low/ faulty screw design. [Solids conveying instability]*: feed hopper fault/ internal deformation of the solid bed in the screw channel/ insufficient friction against the barrel surface. [Surging]*: 30–90 s: feed particles are not sufficiently softened (usually at the beginning of the second compression zone) / too rapid compression screws/ [low feeding efficiency]*/ low friction characteristics/ low bulk density of feed/[melting too soon]*/ adequate early barrel pressure but [melting unstable]*/ first barrel heating too high/ screw speed too fast/ faulty screw design/ additional compound slippery/ bridging in resin feed hopper/ feed-zone temperature too high/ [screw tip pressure too low]*/ compound temperature too high/screw too high/ nonuniform take-off speed/ take-off speed too high/ throughput too fast/ controller fault/ feed resin not mixed well/ melt temperature too low. [Trapped air in extruder]*: unvented extruder/ wrong screw design/ pressure too low/ rear-barrel temperature too high/ screw speed too high/ vacuum too low in feed hopper/ powder feed instead of pellets.

125

126

3 Rules of Thumb for Trouble Shooting

3.9.7

Coating

Trouble shooting: “Poor adhesion”: a variety of apparently contrary causes related to polymer viscosity, degradation, oxidation, tackiness, temperature: melt temperature too low or high/ air gap too small/ chill roll temperature too low or hot/ line speed too fast/ poor match between coating and substrate. “Rough wavy surface (applesauce)”: wrong resin/ temperature too low or high. “Edge tear”: draw ratio too high/ die end temperature too low/ temperature too low or high. “Oxidation”: temperature too high/ screw speed too low/ flows not streamlined/ extruder stopped when temperatures > 200 C/ copolymer not purged with homopolymer before shutdown. “Pinholes in coating”: substrate too rough/ coating thickness too thin. “Surging”: bridging in resin feed hopper/ feed-zone temperature too high/ wrong screw design. “Voids”: moisture/ leaks in resin handling system/ inadequate drying and storage/ [thermal degradation]*/ [gels]*. “Die lines”: nicks in die/ dirty lips/ particles in die. “Pin holes and breaks”: coating too thin/ contamination/ decomposition/ compound temperature too high/ moisture. “Web tears”: compound temperature too low/ too much drawdown/ die lip opening too large. “Poor adhesion”: compound temperature too low/ substrate problems. “Excessive neck-in”: die-to-roll gap too large/ material temperature too high/ die-lip opening too large/ throughput too low/ use resin with lower Melt Index/die-land length too long. See Symptoms for Section 3.9.6 for the following [Contamination]*; [Degradation of melt in extruder]*; [Screw tip pressure too low]*; and [Shear intensity too low]*. [Gels]*: [contamination]*/[degradation of melt in extruder]*/ [shear intensity too low]*/[screw tip pressure too low]*/ number and density of the screen pack too low/ moisture too high/ screw speed too low/ incompatible blend/ [residence time too long]*/ lack of streamlines in extruder/ incorrect startup procedures/ [melting inadequate]*/ [melt too hot]*/ for reactive: localized initiator concentration too high. [Surging]*: screw speed too high/ take-off speed too high/ backpressure too low/ compound temperature too high. [Thermal degradation/crosslinking]*: polymer temperature too high/ screw speed too low.

3.10

Vessels, Bins, Hoppers and Storage Tanks

Bins and hoppers: In general cohesive strength of powders increases with consolidation pressure. Trouble shooting: “No flow”: [arching]*/ [rat holing]*. “Erratic flow”: obstructions alternating between arching and ratholing. “Flooding or flushing” when a rathole collapses it entrains air, becomes fluidized and the material floods through the outlet uncontrollably: “ fine powders such as pigments, additives and precipitates that tend to rathole/ insufficient residence time in hopper for deaeration. “Flow-rate limitation”: fine particles where movement of the interstitial air causes an adverse Dp.

3.11 “Systems” Thinking

“Limited live capacity”: [ratholing]*. “Product degradation”: [ratholing]*. “Incomplete or non-uniform processing”: [ratholing]*. [Arching]*: particle diameter large compared to outlet/ cohesive particles probably caused by moisture or compaction. [Ratholing]*: cohesive particles probably caused by increased moisture or by compaction (fine powders < 100 mm such as pigments, additives and precipitates).

3.11

“Systems” Thinking

Processes are complex systems in which the performance of one piece of equipment interacts with and influences another. Typically when we trouble shoot we tend to focus on individual pieces of equipment and forget that perhaps something away up stream or even downstream could be causing the local difficulty. To some extent we are already familiar with systems and system interactions when we consider a distillation column. This is not one piece of equipment. Rather it is a complex collection of a column with a series of trays linked by vapor and liquid flow with condenser, reboiler, pumps, controllers all interacting. Thus, troubles on the condenser are reflected in the performance of the reboiler and everything in between. We are skilled at being able to make those connections for a distillation “system”. However, when we have a crystallizer, screen, centrifuge, dryer combination we may not be able to easily visualize how the performance of one can dramatically affect the others. In “systems thinking” the focus can be on how performance from one unit is transmitted to units elsewhere in the system. Since equipment is linked by pipes and conveyors carrying fluids and solids, the interaction is through the temperature, flowrates, pressures, compositions and cycling in the streams. Another perspective to “systems thinking” is startup and how the system can move from cold, air-filled conditions to the on-line temperatures, pressures and compositions. A third perspective is how the environment system affects the process system. A useful approach it to reflect on the principle that anything that is created inside the system must go somewhere; where does it go and how does it get out of the system? For example, mass is conserved. 1. 2. 3. 4.

When you start a plant, all the equipment is filled with air; where does the air go? If a dissolved trace metal gets into the system, where does it go? If a particle breaks down where do the particles go? If recycle is used then any trapped species will build up unless there is a bleed or purge.

In trouble shooting “systems”, some common issues include: solvent losses somewhere in the system; fouling; foaming or stable emulsion formation that cause equipment malfunction and carryover; corrosion; recycle causing a buildup of species that may not be removed from the system without adequate bleeds or blow-

127

128

3 Rules of Thumb for Trouble Shooting

down. Although many of these are considered for specific pieces of equipment, we include a generic consideration of some of these here. In this listing, the concept or symptom is shown in parentheses and italics, for example, “Foaming”, followed by possible causes separated by /. If the cause is not a root cause, then it is represented in square bracket plus an *, [ foam-promoting systems]*. These intermediate causes are then listed alphabetically . .

“Fouling”: velocity too slow/ [particulate fouling]* for example, rust, corrosion products from upstream, scale from upstream units, oil, grease, mud or silt/ [precipitation fouling]* for example sodium sulfate, calcium sulfate, lignin / [biological fouling]* species present such as algae and fungi/ [chemical reaction fouling]*, example coke formation and polymerization fouling/ [ flocculation fouling]* or destabilization of colloids, for example asphaltenes or waxes from hydrocarbons/ corrosion products for this unit, see Section 3.1.2/ [solidification fouling]* or incrustation such as the freezing on a solid layer on the surface or crystallization/ [condensation fouling]* such as vaporization of sulfur.

[Biological fouling]*: temperature, pH and nutrients promote growth of algae and fungi/ biomaterials present. [Chemical reaction fouling]*: polymerizable species in the feed/ high temperature causing cracking/ high wall temperatures/ stagnant regions near the wall or velocity too slow < 1 m/s/ reactant droplets preferentially wet the solid surface/ addition of “fouling suppressant” insufficient, for PVC polymerization oxalic acid or its salt or ammonium or alkali metal borate/ pH change. [Condensation fouling]*: wall temperature too low/ contamination in the vapor. [Flocculation fouling]*: pH at the zpc/ low concentration of electrolyte/ increase in humic acid concentration in water in the fall and spring/ colloids present. [Particulate fouling]*: filter not working or not present/ contaminant in feed/ upset upstream/ erosion/ increase in silt and clays in water in the spring. [Precipitation fouling giving scale or sludge]*: soluble species present in feed/ temperature high for invertly soluble/ temperature too low for incrustation or crystal formation. [Solidification fouling]*: wall temperature too low/ missing insulation/ cold spots on wall/ sublimation. .

“Foaming”: [ foam-promoting systems]*/ [ foam-promoting contaminants]*/ [gas velocity too high]*/ [liquid residence time too low in GL separator]*/ antifoam addition faulty/ faulty mechanical foam breaker/ [liquid environment wrong]*.

[Foam promoting contaminants: soluble]*: naturally occurring or synthetic polymers/ naturally occurring or synthetic organics >C10; example lube oils, asphaltenes/ naturally occurring or synthetic surfactants; for amine systems: the surface active contaminants include condensed hydrocarbons, organic acids, water contaminants, amine degradation products/ faulty cleaning before startup; surfactants left in vessels.

3.11 “Systems” Thinking

[Foam promoting contaminants: solid]*: [corrosion products, see Section 3.1.2]*; for amine systems: iron sulfides/ faulty cleanup before startup; rust left in vessel/ dust/ particulates. [Foam promoting systems]*: those that foam naturally: methyl ethyl ketone, aerobic fermentation, textile dyeing foam more readily than amine and glycol absorption systems and latex strippping > amine, glycol and Sulfolane strippers > slightly foam promoting: fluorine systems such as freon, BF3/ systems operating close to the critical temperature and pressure/ surface tension positive system/ [Marangoni effects]*. [Gas velocity too high]*: temperature too high/design error/ [ foaming]*/ vessel diameter too small for gas flow/ column pressure < design/ trays or packing damaged or plugged giving excessive vapor velocity/ temperature too high/ upstream flash separator passing liquids: feed contaminated with excessive volatile species/ stripping gas fed to column too high/ input stripping gas flowmeter error/ design error. [Inaccurate sensing of the interface]*: instrument fault/ plugged sight glass. [Liquid environment wrong]*: pH far from the zpc/ electrolyte concentration too low. [Liquid residence time too low in gas liquid separator]*: interface height decreases/ [inaccurate sensing of interface]*/ turbulence in the liquid phase/ flowrate > expected/ sludge settles and reduces effective height of phase/ inlet conditions faulty. . .

“Corrosion”: see Section 3.1.2. “Stable emulsion formation”: contamination by naturally occurring or synthetic surfactants: example, lubricating oils/ contamination by particulates: example, products of [corrosion, see Section 3.1.2]*, amphoteric precipitates of aluminum or iron/ pH far from the zpc/ contamination by polymers/ temperature change/ decrease in electrolyte concentration/ the dispersed phase does not preferentially wet the materials of construction/ coalescence–promoter malfunctioning/ improper cleaning during shutdown/ [rag buildup]*.

[Marangoni effects]*: non-equilibrated phases/ local mass transfer leads to local changes in surface tension and hence stable interfacial movement [Rag buildup]*: collection of material at the interface: naturally occurring or synthetic surfactants: example, lubricating oils/ particulates: example, products of [corrosion, see Section 3.1.2]*, amphoteric precipitates of aluminum/ naturally occurring or synthetic polymers. .

Other possible causes of trouble for “systems” include:

“Bumping resulting in plate damage”: trace amounts of water. “Catalyst contamination”: trace amounts of water. “Conversion less than expected”: temperature spike causing catalyst decomposition/ temperature sensor reading low/ [catalyst contamination]*. [Cycling]*: two vessels in series on level control/batch processes in sequence but out of synchronization/ no intermediate storage. “Distillation overhead off spec”: excessive inerts from upstream/buildup of trace/ purge not sufficient from recycle.

129

130

3 Rules of Thumb for Trouble Shooting

[High Dp across bed of catalyst/particles/resin]*: temperature spike/ temperature sensor reading low. “Leaks”: temperature spike/ pressure spike/ temperature sensor reading low/ glands not tight enough around rotating shafts/ valve stem leaks/ thermal expansion of the different metals and parts not correctly accounted for, example, flanges and gaskets on high-temperature heat exchangers. “Particle agglomeration in pneumatic conveying”: increased humidity in air/ trace contaminants in air. “Separation performance of column decreases”: trace amounts of water/ trace amounts of water trapped in column/ [bumping resulting in plate damage]*. “Temperature runaway in reactor”: operator error: overcharge reactor/ trace water/ [poor temperature control]*.

3.12

Health, Fire and Stability

In general, we should all be aware of the hazards before we encounter trouble on the plant. Our response should be instantaneous in identifying the degree to which a trouble on the plant or process poses a hazard. 3.12.1

Individual Species

Many indicators are available to guide us about the hazard posed by individual species. The hazards are usually considered to be health hazard (toxic or causing death; causing genetic defects; causing long term disability; causing specific illnesses such as asthma, cancer; impacting on the environment); flammable hazard and explosive hazard. A simple guideline is the NFPA ratings. These are a scale 0 to 4 with the higher number indicating the extreme in hazard (see Woods, 1994, Data for Process Design and Engineering Practice, Prentice Hall). MSDS documentation is available for most chemicals. Some sources include http://www.msdssearch.net or http://www.msdsxchange.com although these direct you to manufacturer’s MSDS information. The quality of the documentation varies from company to company. Also consult http://toxnet.nlm.nih.gov. Health

Vinyl species, benzene ring compounds. Carcinogenic to humans: asbestos, benzene, benzidine, b-naphthylamine, vinyl chloride. Proven carcinogenic in animal trials: acrylonitrile, butadience, N-nitrosamines Flammability

Species with low Ignition temperatures, say in the range 85–100 C. Recall that the ignition temperature for combustion in air could be lower if in the presence of pure

3.12 Health, Fire and Stability

oxygen or chlorine. For example, for toluene the ignition temperature in air of 535 C is reduced in chlorine to 175 C. Stability/explosiveness

Guidelines from structure: azide, perchlorates, nitro compounds, peroxides, vinyl species. For reactivity or potential for explosion or thermal runaway: if heat of decomposition > 0.2–0.3 MJ/kg; those > 0.5 MJ/kg may be explosive and those > 0.8 MJ/kg may be detonable. Hazardous reactants and types of reactions

Rosenmund reduction, oxidation of nitrous acids; oxidation of low molar mass peracids; alkylation of alkali acetylides, alkylation via Arndt–Eistert; alkylation of diazoalkane and aldehyde; alkylation of diazoalkane; condensation of carbon disulfide with aminoacetamide; esterification of carboxylic acid and diazomethane; esterification of acetylene and carboxylic acid-vinyl ester; reactions involving concentrated peroxides and peracids. Heats of reaction

Extremely exothermic: direct oxidation of hydrocarbons with air; chlorinations, polymerizations of polyethylene without diluent. > 3 MJ/kg (or > 150 MJ/kmol: examples: combustion –900 MJ/kmol; hydrogenation of nitroaromatics, –560 MJ/kmol; nitrodecomposition, –400 MJ/kmol; diazo-decomposition, –140 MJ/kmol). Strongly exothermic: nitrations, polymerization of propylene, styrene butadience. 1.2–3 MJ/kg (nitration, –130 MJ/lmol; amination –120 MJ/kmol; sulfonation or neutralization with sulfuric acid, –105 MJ/kmol; epoxidation, –96 MJ/kmol; diazotization). Moderately exothermic: condensations or polymerization reactions of species with molar mass 60–200, 0.6–1.2 MJ/kg. Exothermic heat of reaction gives an adiabatic temperature rise > 100–200 C. Impact sensitivity of < 60 J for solids and < 10 J for liquids. For powders: explosive usually if diameter is < 200 mm with highest probability of dust explosion if diameter < 65 mm. Note upper and lower concentration for explosive mixture. 3.12.2

Combinations

With water, with air with other chemicals: Chemicals that react violently with water include: acetyl chloride, aluminum alkyds, aluminum alkyl halides, aluminum chloride, calcium hydride, calcium oxide, ethyl aluminum dichloride, fluorine, lithium hydride, phosphorous pentoxide, phosphorous trichloride, potassium, silicon tetrachloride, sodium, sulfur trioxide, sulfuric acid, thionyl chloride, titanium tetrachloride, zinc alkyls. Moderately: acetic anhydride, activated alumina, aluminum phosphide, calcium, calcium carbide, cal-

131

132

3 Rules of Thumb for Trouble Shooting

cium phosphide, lithium, activated molecular sieves, potassium hydroxide, activated silica, sodium hydroxide, sodium peroxide.

133

4

Trouble Shooting in Action: Examples In this chapter we meet five engineers as they trouble shoot the five cases introduced in Chapter 1. While you are reading these cases focus on the trouble-shooting process used. Compare the approaches taken here with the approaches you took in addressing the problems posed at the end of Chapter 1. The engineers have a range of practical experience. In Case ’3, Michelle has three years experience. In Case ’4, Pierre has 12 years experience. For Case ’5, Dave has relatively no practical experience. For Case ’6, Saadia has 10 years experience. For Case ’7, Frank has 25 years of experience. These cases have been carefully selected to provide a range of approaches, degrees of difficulty and to illustrate the variety of trouble shooting approaches taken. Each of the five scripts consists of about three parts with each part concluding with a few questions for you to consider. This reflective break was introduced to give you a chance to reflect on how you would have handled the case, and to decide what you should do next. I recommend that as you read each script, that you play the game. At the end of each case an assessment is given of the problem-solving processes used by each of the trouble shooters. Their scores range from F to A+; the richness is in the detail of the feedback. Not everyone will follow Lieberman’s (1985), Gans’s (1983), Kister’s (1979) or my style in trouble shooting. The key is to identify your style and develop confidence in using it. The cases do not have to be addressed in any order. Select what you would prefer. Enjoy!

4.1

Case ’3: The Case of the Cycling Column

Michelle1) graduated three years ago. She is on her journey to become an excellent trouble shooter but she lacks much practical experience. “Cycling column? Called in as chief trouble shooter!” exclaimed Michelle. Well, I’ll do my best, she thought. Michelle recalled the Trouble-Shooter’s Worksheet. She carefully noted the key infor1) In all these cases, the names of the engineers

have been changed.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

134

4 Trouble Shooting in Action: Examples

mation: startup after maintenance, iC4, and cycling level in the bottom of a distillation column. From the diagram, it looks like a thermosyphon reboiler and an inverted bucket steam trap. Well, I want to and I can! I have an organized strategy to apply. What are the strengths and weaknesses of Michelle’s approach so far? What would you have done differently? Would you head out to the plant directly or are some quick checks to be done from your files first? Let’s check, thought Michelle. This isn’t an emergency. Until I learn more I cannot think of a “safe-park” condition I should impose. The key information is that these symptoms are occurring just after shutdown so I really need to find out what was changed and what was worked on. She cautioned that she should change her thinking from “cycling level in the bottom of the column” “to apparent cycling in the level”. Michelle quickly scanned the P&IDs for the iC4 unit and confirmed that the diagrams showed a thermosyphon reboiler and an inverted bucket trap with a bypass. She checked in Chapter 3 for the cycling and reboilers and then checked steam traps: Reboilers in general:

“Cycling (30 s – several minutes duration) steam flow, cycling pressure on the process side and, for columns, cycling Dp and cycling level in bottoms”: instrument fault/ controller fault/ condensate in instrument sensing lines/ surging/ foaming in kettle and thermosyphon/ liquid maldistribution/ steam-trap problems, see Section 3.5, with orifice Dp across trap < design/ temperature sensor at the feed zone in a distillation column/ collapsed tray in a distillation column. For thermosyphon: “Cycling (30 s – several minutes duration) steam flow, cycling pressure on the process side and, for columns, cycling Dp and cycling level in bottoms”: in addition to general, all natural circulation systems are prone to surging/ feed contains high w/w% of high boilers/ vaporization-induced fouling/ constriction in the vapor line to the distillation column. For horizontal thermosyphon: maldistribution of fluid temperature and liquid. Section 3.5 steam traps: .

Good practice: Install a demister.

Steam traps: install trap below condensate exit (or with a water seal if the trap is elevated), use a strainer before most traps, use a check valve for bucket traps. Slant pipes to the trap. Use a downstream check valve for each trap discharging to a common header. Pipe diameter 3 trap inlet pipe diameter. Prefer to install auxiliary trap in parallel instead of a bypass. Do not group trap thermodynamic traps because of their sensitivity to downstream conditions. Float and thermostatic: usually discharges continuously, low pitched bubbling noise. High-pitch noise suggests live steam is blowing. Balanced thermostatic: leave about 0.6 m of uninsulated pipe upstream of trap.

4.1 Case ’3: The Case of the Cycling Column

Inverted bucket: use initial prime to prevent steam blowing. Thermodynamic: about 6 cycles/minute Trouble shooting: the major faults are wrong trap, dirt, steam locking in the trap, group trapping, air binding and water hammer. Too large a trap gives sluggish response and wastes steam. Too small a trap gives poor drainage, backup of condensate. “Cold trap + no condensate discharge”; steam pressure too high/ no water or steam to the trap/ plugged line or strainer/ orifice enlarged by erosion (bucket trap)/ incorrect Dp across the orifice (inverted bucket)/ bucket vent clogged (inverted bucket)/ current operating pressure > design/ trap clogged. “Hot trap + no condensate discharge”: bypass open or leaking/ trap installed at high elevation/ broken syphon/ vacuum in heater coils. “Live steam blowing”: bypass open or leaking/ worn trap components/ scale in orifice/ valve fails to seat/ trap lost prime (inverted bucket)/ sudden drops in pressure/ backpressure too high (thermodynamic). “Continuous discharge when it should be discontinuous”: trap too small/ dirt in trap/ high-pressure trap installed incorrectly for low pressure service (bucket trap)/ valve seat clogged with dirt/ excessive water in the steam/ bellow overstressed (thermostatic). “OK when discharging to the atmosphere but not when to a backpressure condensate header”: condensate line diameter too small/ wrong orifice/ interaction with other traps connected to a common header/ condensate line partially plugged/ [backpressure too high]*. “Slow and uneven heating of upstream equipment”: trap too small/ insufficient air handling capacity/ short-circuiting when units are group trapped. [Backpressure too high]*: return line too small/ other traps blowing steam/ obstruction in return line/ excess vacuum in return line. Michelle grabbed her notebook and headed out for the unit. She mentally went over some hypotheses. What are the strengths and weaknesses of Michelle’s approach so far? What would you have done differently? What is the problem? What are your hypotheses? Wait a minute. Here I am thinking of hypotheses when I should slow down. First let’s do an IS and IS NOT, define the problem and focus on what was done during the shutdown. She joined a group of operators at the base of the column. You could hear the cycling in the steam control valve. Her temptation was to immediately put the control system on manual and see what happens but she methodically gathered the following information: During the shutdown the condensate from this inverted bucket trap, that previously discharged to atmosphere, was repiped to discharge into a 200 kPa g condensate header. The condensate discharged into the top of the header. Steam to the reboiler was “saturated” at 1.7 MPa g. For this unit the only other maintenance was instrument checks and visual inspection of the trays. The visual inspection involved opening the access holes. No faults were found in the instruments and no changes were made to the settings of the controllers. Michelle then wrote the following in her notebook:

135

136

4 Trouble Shooting in Action: Examples

What IS: Where IS: When IS: Who IS:

cycling in the level in the sight glass. IS NOT: changes in feedrate or composition; no other apparent cycling upstream or in the feed. on this unit. used to work OK before the shutdown. this is the first shift.

What is the problem? Problem SMARTS$ Specific: stop the level in the sight glass from cycling. Measure via level in glass Attainable? Should be. Depends on the root cause. Reliable? Depends on the root cause. Timely? Several simple tests should be able to resolve this. Safe? Doesn’t seem to be an issue here. $ yes, we are losing money every minute. She confirmed that the steam entered the top of the reboiler. However, she thought “all natural circulation systems are subject to surging”. She quickly listed some hypotheses: . . . . . . .

collapsed tray. change in downstream pressure affecting bucket trap. instruments wrong; liquid level is not cycling. control system. restriction in the vapor line. change in concentration of high boilers in the feed. temperature sensor incorrectly located in the feed zone.

She discarded “instrument wrong” because she could see the level rising and falling with her own eyes. It may lag or lead the actual level in the column but “something is cycling in there.” Collapsed tray is an unlikely possibility because no one went down through the column checking each plate. They “visually inspected” by looking in the access holes. She reflected on the type of tests she could do and prioritized them. 1. Checking the control system was relatively inexpensive and fast. 2. Opening the bypass on the trap and/or changing the condensate to discharge to atmosphere temporarily are other options to check out while she rechecks the specs and sizing used on the bucket trap. 1) When the control system was put on manual the level gradually increased. “Let’s think about this”, she said, “steam enters the thermosyphon reboiler, condenses and boils a given amount of bottoms. This causes a pressure differential and fresh bottoms cycles into the tubes. However, if the trap does not remove the condensate then the condensate builds up, decreases the area, decreases the heat transferred and hence the boilup. The level increases because feed continues to the column but the boil up rate is decreased.” When the set point was increased manually, the valve stem on the steam moves up, the liquid level appears in the sight glass and continues to drop

4.1 Case ’3: The Case of the Cycling Column

2)

but shortly thereafter the level appears in the glass and is rising. I conclude that the control system is not at fault. Open the bypass and manually try to change the bypass setting to “level out” the level. Frustrating as this is to try to adjust, this seems to provide some level of control of the cycling.

Michelle concludes that the plant could operate by having the bypass partly open or by unhooking the condensate from the main and discharging to atmosphere as it had operated before. She elects to do the former while she returns to her office to check on the sizing and selection of the trap. What are the strengths and weaknesses of Michelle’s approach so far? What would you have done differently? Did she address the symptoms or the root cause? What tests and questions would you have posed? What corrective action would you have taken? When Michelle consulted the design file, she found that the trap had been selected to handle a design condensate flowrate of 0.3 kg/s for an inlet pressure of 1.7 MPa and atmospheric discharge. Over the years the production rate had increased so that, from her energy balance calculations, the actual condensate flowrate was about 0.6 kg/s. However, throughout these changes the trap had not been changed and was now operating at its maximum condensate capacity. When the differential pressure across the trap was reduced, the condensate handling capacity was reduced by about 10% so that the trap was now undersized. To correct this Michelle selected a new trap, or a different orifice in the existing trap, to be installed to handle 1.2 kg/s condensate under the steam pressure of 1.7 MPa and the downstream pressure of 200 kPa. This followed the general guideline of selecting inverted bucket traps based on double the normal capacity. She also recommended that a check valve and strainer be installed upstream of the trap. Discussion Overall degree of difficulty of this problem is 4/10, relatively easy. Michelle took half a day to solve this problem. Let’s reflect on Michelle’s approach following the guidelines in the feedback form. Overall, Michelle recognized her inexperience and tried to follow the Worksheet and to draw on the suggestions from Chapter 3. Michelle treated this as a “problem” and not an exercise right from the beginning. Problem Solving

She was systematic, organized, used the IS and IS NOT approach and wrote things down. Not much verbal monitoring was apparent and limited checking and double checking. She became a little too hasty in the hypothesis checking and prioritization stages. She discarded some hypotheses without explicitly noting this. This worked OK for her in this case but she should be developing good habits. Overall a rating of B.

137

138

4 Trouble Shooting in Action: Examples

Data Handling

Actions were based on fundamentals and she astutely enriched her experience by checking files and trouble shooting symptoms/cases (as given, for example, in Chapter 3). Her reasoning is OK for what she described. However, it is difficult to assess because she mentally discarded hypotheses and didn’t systematically check the evidence with the hypotheses. This worked for her this time because there were only two pieces of evidence: changes were made to the condensate system in the turnaround and the level in the sight glass cycles. Overall rating B. Synthesis

She listed a variety of hypotheses. Considered many viewpoints initially. Narrowed into the two hypotheses that were easy to check. Overall rating B+. Decision Making

She seemed to continually try to prioritize. No obvious bias was apparent. She should have discussed her decision to “operate on bypass” with the operators because they have to buy in to this change. The criteria she used were not apparent. Overall rating C. Strengths: systematic, based on fundamentals, wrote things down, variety of hypotheses, used resources. Areas to work on: more explicitly aware of the process used, think about the operators.

4.2

Case ’4: Platformer Fires

Pierre is frustrated. Another fire. “That hydrogen just leaks out of every crevice. And the high temperatures we are talking about, 500 C. Heh, how can we hope to ratchet the flange tight at room temperature and hope that it will remain tight when we heat it up!” He pulled out the book on flange design and starting doing calculations of the relative thermal expansion of metals for 1 1/4 Cr–12 Mo. And we hydrostatically tested the tube bundle based on the pressure differential of 2–2.7 MPa to test for leaks between the tube sheet and the shell. We were careful to use the differential, instead of the absolute pressure of 4.8 MPa. How would you characterize Pierre’s approach so far? What are the strengths and weaknesses of his approach so far? What would you have done differently? What is the problem? What are your hypotheses? Pierre tossed the pencil down on his desk. Let’s go back to basics. I have a gas at high pressure and temperature that is leaking through a crevice, igniting and flaming. We have used all our best mechanical engineering brains to design the flange to prevent leaking. Different gaskets, different tightening approaches. These were de-

4.2 Case ’4: Platformer Fires

signed so that the temperature differential on the bolts was negligible even though the tubesheets were very thick. They’ve done all the thermal expansion stuff already. All the bolts were supposedly torqued the same. I’ve got to think out of the box! Let’s pull out the Trouble-Shooter’s Worksheet I used to use. Is it a safety hazard? Yes siree! It is a fire hazard. Hydrogen + oxygen + spark. What’s going on is pretty straightforward, except that it’s not clear whether the naphtha feed is on the tube or shell side. OK, that completes Engage. Define the stated problem is next. What IS and IS NOT: When IS and IS NOT: Where IS and IS NOT:

fires are on the effluent exchanger. IS NOT elsewhere nor on other units. ever since we started up IS around the flanges on the shell and tube. IS NOT on the exchanger

This is a fundamentals problem. OK. I think I have written down the facts so far. Explore. This is a problem! But, I need to bring out the files and clarify information. From the files, Pierre found that: a) b)

c)

the naphtha is on the tube side and at the higher pressure; the hydrogen-rich gas is on the shell side. the platformer temperature is kept < 540 C to prevent thermal degradation of the catalyst. The hydrogen-rich gas is 500 C and used to preheat the naphtha. from the diagrams of the effluent exchanger, the most likely hydrogen leak seems to be between the flange from the shell, the head and the bonnet.

OK, Let’s put this into perspective via the Why? Why? The problem as I see it is to “prevent hydrogen-rich gas from leaking out the flange” so I’ll put that at the start. Perspectives. Why? Why? Why?

Why? Why? Why? Why? Why?

6. so that I can be paid › 5. so we can sell platformate and make profit › 4. so the whole process can operate › 3. make it safe; prevent flames to other parts of the process › 2. prevent fires on the effluent exchanger ›

Start ﬁ 1. prevent hydrogen-rich gas from leaking out the flange That was pretty useful. Maybe I have been working on the wrong problem! Maybe I should just let the hydrogen leak out and focus on how to prevent fires on the effluent exchanger.

139

140

4 Trouble Shooting in Action: Examples

What are the strengths and weaknesses of Pierre’s approach so far? What would you have done differently? What would you do next? How to prevent fires? A fire is hydrogen + oxygen + spark; I said that right at the start. OK, if I have the hydrogen, then I can remove the oxygen or the spark or both. Let’s brainstorm to remove the oxygen: nitrogen blanket, steam blanket,... I’m going to stop there. I’ll just install a circular sparge ring whose circumference is about 5 cm beyond the flange, drill holes on the inside and sparge low-pressure steam at the flange. The heat from the steam will tend to minimize the temperature differential between the inside and the outside of the flange and the steam should displace the oxygen. Viola, no fire. Perhaps in the future they will develop Platforming processes that operate at lower temperatures and pressures, said Pierre wishfully. Discussion The overall degree of difficulty for this problem is about 3/10, this is relatively easy. However, it wasn’t easy at the start for our frustrated Pierre. His frustration shows through in his approach. Fortunately, he got out of the box by using the Why? Why? Why? technique. Consider now some feedback for Pierre. Problem Solving

Pierre started with more frustration and lack of confidence. He forgot that he had an organized strategy that had been helped him in the past. He used very little monitoring, checking and double checking and even when he started to use the Worksheet he used it rather superficially. He did use the IS and IS NOT effectively and tried the Why? Why? Why? Fortunately, he used the latter effectively. Overall rating C–. Data Handling

He gathered information from the files, and tried to reformulate the problem based on fundamentals. However, he really didn’t get to hypothesis generation. His reasoning was OK based on what he told us explicitly. He didn’t formally draw on past experience. Overall rating D. Synthesis

He failed to explicitly list hypotheses. Considered very few viewpoints. Overall rating D to F. Decision Making

He seemed to place the blame on the mechanical engineers and the inability to design a flange that would keep the gas from escaping. He focused initially on preventing the gas from leaking. No apparent criteria were given for the decisions. Overall rating D. Strengths: got an answer, could think outside the box, was somewhat systematic once he started to use the Worksheet, used resources, based on fundamentals. Areas to work on: self-confidence, be more systematic.

4.3 Case ’5: The Sulfuric Acid Pump

4.3

Case ’5: The Sulfuric Acid Pump

Dave is a new graduate engineer. Dave notes that the evidence points to cavitation. However, being relatively new on the trouble shooting scene he starts to write down a description of the process. Sulfuric acid is stored below the level of the pump. The pump has to “lift” the acid up to the intake. The storage tank is open to the atmosphere. Ok, let me check. Have I understood this correctly? Dave runs his pen around the diagram and checks that he has included all that he thinks is pertinent about the situation. He underlines the evidence and inserts “?” when he has something he is concerned about: the vertical dimensions, two hour cycle of operation, receives “acid”? via return lines from all over the plant; dilute acid (?density and corrosiveness); when the site gauge reads 0.7 m left in the tank, the pump makes a “crackling” noise that the operator says “sounds like cavitation”. Concern about erosion of the impeller (? Why, because of acid? Because of cavitation?). OK, “I want to and I can!” says Dave as he takes out his trusty Trouble-Shooter’s Worksheet. “Ok, what’s my next step? I think I am finished with Engage. But wait a minute. What about Safety? Hazard? Safe-hold operation? Other than the fact that this is acid, I don’t see this as an emergency priority so I can proceed with Define the stated problem. He quickly wrote down the IS and IS NOT information. Dave noted that he probably had enough information here to “solve the problem” without going out on the site. He recalled the cardinal rule of trouble shooting “Go out on the site and see it!” Dave acknowledged this but felt he should do some detailed calculations before going out. His focus would be on the “Basics” around NPSH and suction lift since this trouble occurred from startup. What were the strengths and weaknesses of Dave’s approach so far? What would you have done? Would you go directly to the plant or would you do calculations first? What calculations would Dave make? Is this really an NPSH problem? Is this a suction-lift problem? Before Dave started into the calculations he looked out and saw the rain beating down. It was a good time to stay in the office! He thought. Then, he remembered that when it’s raining the atmospheric pressure is usually low. Since this process is open to the atmosphere, he noted that he should include a possible lowering of the atmospheric pressure as a potential cause of the lack of NPSH. Oh, Oh! I did it again, I said lack of NPSH instead of saying apparent lack of NPSH. By this time Dave realized that he had skipped into the Explore stage without doing his usual checking and double checking. He perused what he had done, said a few positive “yeps” and then started focusing on the SMARTS$. His goal is to stop the “crackling” noise when the level drops to 0.7 m. In his mind Dave realized that he had changed the problem from a “crackling” problem to a “cavitation” problem to a lack of NPSH problem. Hmmmm.

141

142

4 Trouble Shooting in Action: Examples

Dave decided to continue his analysis of the NPSH and suction lift. He hauled out some texts and posed a number of What if? questions: what if the acid return liquid is more dense? Less dense? What if the atmospheric pressure really drops? What if the vent to atmosphere becomes plugged? What if the site gauge is reading wrong? What if gas or air is dissolved in the acid? Dave stopped his ruminations and consulted his files on cavitation: [cavitation]*: design fault/ liquid too hot/ non-condensibles in liquid/ vortex entraining gas/ decrease in density of the liquid/ blockage or excessive Dp on suction/ suction velocity too high/ increase in rpm without increase in NPSH/ increase flowrate increases NPSH demand. For suction-lift situations: suction lift too high/ low atmospheric pressure (for open systems)/ air leakage into suction line. Dave reflected on this information and decided that the key hypothesis to check was “design error”. He realized he should list a variety of hypotheses about the root cause of “cavitation” but he wanted to do the simple checks first. If the design estimations looked OK, then he would list a range of hypotheses. However, the design calculations seemed too fundamental to ignore. He would: 1) estimate the NSPH supplied; 2) check the files and see the NPSH required from the vendor’s information and 3) compare. He systematically did each in turn. 1)

2)

3)

Estimate the NPSH supplied. The design files showed that the typical “dilute acid” had a density of 1.2 Mg/m3 (whereas 98% acid had a density of 1.84 Mg/m3) and at the temperature of operation the vapor pressure of water is about 5 kPa; that of dilute sulfuric acid is about 3 kPa. It was raining and the atmospheric pressure was low, he estimated it to be 90 kPa abs. He plugged values into the equation for NPSH supplied for an atmospheric pressure of 90 kPa, and an estimated friction loss of 0.7 m= (Atmospheric pressure– vapor pressure of fluid at operating temperature) converted to head–head of lift–head loss in friction. 7.4 m–3.6 m–0.7 m= 3.1 m of acid. He rechecked his calculations. OK. From the vendor files he found that the NPSH required was 1.4 m “of water at 21 C” for this pump at 1800 rpm and the design flowrate of 15 L/s. Since the NPSH required is for a pressure drop in the horizontal plane, NPSH required is relatively independent of the density and temperature of the fluid (although he remembered that some corrections have been published for the NPSH required for hydrocarbons that show the NPSH required as being about 0.6 times the NPSH required for water. However, for conservative purposes use NPSH required for cold water.) OK, so NPSH required = 1.4 m. Compare. Recommendations about the relationship between the NPSH supplied and required vary. Some suggest that NPSH supplied should be 20% higher than that required; or 0.5 m of water higher than that required or the ratio of NPSH supplied/NPSH required > 1.4. For this problem NPSH supplied is 160% higher; larger by 3.1–1.4 = 1.7 m and the ratio is 2.6. Let’s check.

4.3 Case ’5: The Sulfuric Acid Pump

Dave went over his calculations. I’d better return to my earlier What ifs? What if the acid density is 1.4 and not 1.2? Dave quickly rechecked his values and came to about 2.1 m acid. This still meets the requirements. What if the pump flowrate is 20 L/s? From the vendor’s data Dave noted that the NPSH required increased to about 2.5 m water. OK, thought Dave; it doesn’t look as though it’s a design error. I’ll now go back and formally list some hypotheses consistent with a “crackling noise” that an operator interprets as being cavitation. Let’s see, and Dave checked over the list for “cavitation” and selected – liquid too hot – non-condensibles in liquid – air leakage into suction line – vortex entraining gas – decrease in density of the liquid – blockage or excessive Dp on suction – suction velocity too high – increase in rpm Wait a minute! I’m depending solely on a list from some book. Let me think from basics about this. What were the strengths and weaknesses of Dave’s approach so far? How well is Dave following the Trouble-Shooter’s Worksheet? What would you have done? Do the calculations prove that there should be sufficient NPSH supplied? What are your hypotheses? What would you do next? Dave put on his raincoat and headed out to the plant. By now he had zeroed in on the hypothesis that a vortex was forming when the level reached 0.7 m. After he talked with the operator and looked around he decided to test his hypothesis by reducing the flowrate from the pump. This should lower the velocity in the suction line, and lower the friction loss in the suction line. Under these conditions, the liquid level should drop lower in the tank before it starts cavitating. His discussions with the operator and his view of the layout confirmed what he had thought. He realized, however, that if he had come out to the plant earlier he could have hooked up the utility air line to the vent, increased the pressure in the vessel and checked to see if it was an NPSH required being insufficient quite easily. When the operator reduced the flowrate the liquid level dropped to 0.58 m before cavitation. Ahah, cavitation is caused by vortex formation. Dave shifted gears now to think about how to correct this temporarily and in the long run. The key goal is to prevent cavitation (and the resulting erosion of the working parts of the pump) and to lengthen the pump cycle. To prevent the vortex formation he could float, temporarily, a wooden egg-carton construction on the surface to serve as a vortex breaker. Although Dave thought of floating a double layer of ping pong balls on the surface he realized that if a vortex did form and the balls were sucked into the pump there would be real trouble! This wooden floating vortex

143

144

4 Trouble Shooting in Action: Examples

breaker should allow pumpage to lower levels in the tank. Another option would be to reduce the flowrate as the level drops. This would mean a signal would have to be transmitted to a location near where the pump exit flowrate could be controlled. At the next shutdown, a well-designed vortex breaker could be attached over the exit pipe; or a control valve could be installed on the pump discharge that gradually closed as the level in the tank dropped. Dave reflected on how he handled this, one of his first real trouble-shooting problems. He made notes about what he would do the same and what he would do differently on his next trouble-shooting assignment. Discussion Overall degree of difficulty of this problem is 5/10, relatively easy. Dave took a day to solve this problem. Overall, Dave tried to follow the Worksheet and to draw on the suggestions from Chapter 3. This was a “problem” and not an exercise for Dave, and he handled it that way. Problem Solving

Dave tried to be systematic, organized, used the IS and IS NOT approach and wrote things down. He included much verbal monitoring and checking and double checking. Overall a rating of A. Data Handling

Actions were based on fundamentals and he used books, files and trouble shooting symptoms/causes (as given, for example, in Chapter 3). He should have visited the site earlier. He systematically checked the evidence with the single hypothesis. Overall rating B–. Synthesis

He focused early on the process operator’s judgment about the symptom. He used the resources in Chapter 3 to create a list of hypotheses but then prioritized these and checked first the design error. Many hypotheses should be kept active concurrently. Overall rating C. Decision Making

He prioritized. No obvious bias was apparent. Overall rating B. Strengths: use of resources, monitoring, checking and double checking, being systematic, based on fundamentals. Areas to work on: visit the site, keep more hypotheses active.

4.4

Case ’6: The Case of the Utility Dryer

Saadia perused the trouble-shooting case that appeared on her desk. Over the past ten years she had very successfully solved a range of problems with dryers, but

4.4 Case ’6: The Case of the Utility Dryer

usually the adsorbers were hooked in parallel with one unit on-line and the other being regenerated. For this case, the units are in series with the regeneration occurring before the adsorption. “Very interesting,” she murmured, “but I’m going to treat this case as a problem and not an exercise.” She retrieved the Trouble-Shooter’s Worksheet and checked off the items. This is not an emergency; except that we may need dry instrument air from a skid mount on these cold days. Let’s be sure that I understand this particular process. She traced the path of the feed air on the diagram and noted that all traps discharged below ground and were inaccessible for sampling. “Fortunately, there seems to be enough sampling ports,” she noted as she checked off S1 to S4. “Often the problems in the past have been leakage of steam from the regenerator into the air, leaky valves, inadequate regeneration, wrong adsorbent or adsorbent damaged by excessive regeneration temperature, or adsorption cycle too long or change in concentration in feed.” “Hold on a minute,” Saadia said to herself, “you said you were going to treat this as a problem and not relive your past achievements! Make sure you understand this flow diagram and then move to define the stated problem.” Saadia looked again at the flow diagram and traced the flows for Bed B being regenerated, then cooled. She wrote down succinctly: P1 reads 550 kPa ? Cycle set on V1 for 2 h? Then 1 h ? And valves V2 and V3 for 3 h? TRC 1 = 175 C? Flowrate about 12 design? Proportioning valve actually shut? Is the air really wet leaving our unit? Is this the first time this is reported? Did the vendor give us a performance check on this unit? Current weather? OK, I think I have finished with the Engage part. Now to Define the stated problem. What IS and IS NOT: When IS and IS NOT: Where IS and IS NOT: Who IS and IS NOT:

instrument air lines freezing; claims of “wet” air leaving the drying bed. on colder nights; not reported at other times on instrument lines. No reports for air supplied elsewhere in the plant. plant operators claim following vendor’s instructions.

“I’ll focus on the basics. I think I’m ready to really get into this problem. Let’s Explore” said Saadia. What were the strengths and weaknesses of Saadia’s approach so far? What would you have done? What would you do next? Before I define the real problem, I want to gather three forms of baseline data: 1) I’ll pull the vendor’s specs on this unit; 2) I’ll do a simple spot-check of performance data, with some samples and 3) do some simple rule-of thumb checks. .

Vendor’s specs. The design conditions guaranteed by vendor’s file said: – Adsorbent: 5000 kg of activated alumina H-151 per bed: – Drying cycle: 3 hours – Regeneration cycle: 2 hours – Cooling cycle: 1 hour

145

146

4 Trouble Shooting in Action: Examples

– – – – – – – – –

Dry air product, standard cubic decimetres per second, Ndm3/s, FRI 1: 7300 Minimum air flow for regeneration, Ndm3/s: 2360 Proportional valve: part closed but closed enough to ensure minimum air flow for regeneration. Inlet air temperature, C, T1: 32.2 Inlet air pressure, kPa g, P1: 550 Moisture content of inlet air, saturation temperature at the design pressure, C: 32.2 Final moisture content in “dry air”, C, S4:–42.7 at pressure Pressure drop, kPa, P1–P4 = 13.8 Steam pressure to the heater, MPa g, P2: 1.0 min–5.5 max.

.

Simple spot-Test I. Readings should be made of all the instruments 1 hour into the cycle. Use an open cup dew point apparatus attached to the sample lines S1 (inlet dew point), S3 (after the separator) and S4 (dry air dew point). I realize this is the dew point at atmospheric pressure and I have to convert it to pressure conditions if I want to compare. The results were: – Pressure, Inlet air, P1, kPa: 689 – Temperature, Inlet air, T1, C: 22.2 – Dew point, Inlet (atmos) S1, C:–6.7 – Dew point, Inlet (press), S1, need to calculate. – Air flow to regeneration, all because pressure P3 is full on to close the valve. – Pressure, Steam P2: 5.2 MPa – Temperature, Steam T2: 277 C – Temperature, TRC 3 exit gas from heater: set 177 C – Temperature after regenerator, T4, C: 104 – Temperature after cooler, T5, C: 24 – Dew point, (atmos) exit of Separator, S3, C: unable to measure because it fogged up immediately at ambient temperature and pressure. – Pressure, Effluent air, P4, kPa g: 662 – Air flow, FR 1, Ndm3/s: 4000 – Exit (atmos) Dew point, S4, C:–48.3 – Dp, kPa, calculated P1–P4, 27

.

Simple checks: “Lots of numbers, but I want to select the key ones,” said Saadia, astutely. – Anything unexpected with the feed, the steam, the regenerator, the separator? No, temperatures and conditions pretty consistent with design conditions from the vendor: except flowrate = about 12 design. – Dew-point conversion between atmospheric pressure and pressure conditions of about 762 kPa abs is a) determine the partial pressure, pp, of the water vapor at the dew point, atmospheric; multiply by the ratio of the process pressure absolute to atmospheric to obtain the partial pressure of water vapor under the process conditions; from humidity tables/charts

4.4 Case ’6: The Case of the Utility Dryer

–

–

–

find the dew point comparable to this partial pressure of water. Alternatively, I could just express everything in ppm water. Since Saadia realized she would be doing this conversion often she created a simple program to speed her calculations. For the test run, the Exit (press) Dew Point, corresponding to sample location S4 = –29.5 C or a moisture content of about 50 ppm or more than 3 greater than expected of 15 ppm. According to vendor’s specs of 32.2 C saturation at 650 kPa abs, the incoming moisture content is 7375 ppm. For the test conditions, the incoming moisture content was 3430 ppm. Hence, the exhaust specs of 15 ppm should be easy to achieve if everything is working as expected. Interesting: at sample location S3, exit from the separator, since today’s temperature is 21 C the moisture content is > 24,000 ppm or 2.4% v/v moisture. “I’ll call this a surprise because I have no other data as to what I should expect here. This suggests a carryover of mist from the separator.” Surprises: exit amount of moisture in the air is = 3 times greater than the target value of 15 ppm and pressure drop= double expected. Before she proceeded further Saadia took several minutes to recheck her calculations. Let’s see, Dew point (atmospheric) of –48.3 C corresponds to a Dew point (at pressure conditions) of about –30 C (which is consistent with the dew point under pressure conditions not meeting the specs of –42.7 C).

Saadia realized that this simple test showed that the problem was the exit moisture content (from the sample at S4) in the exhaust, “dried” air is about 3 times higher than expected. This is consistent with “freezing” lines when the outside temperatures are cold. The other symptom is the pressure drop is double the expected value even when the flow rate is 12 the design value. “It’s time for hypotheses! ... and tests” declared Saadia, whereupon she added following list to her chart: Symptom a. exit air “wet”: 3 higher than specifications b. pressure drop double expected value even when the flowrate 12 design c. d. e.

147

148

4 Trouble Shooting in Action: Examples Working Hypotheses

1. Steam leak 2. Excessive moisture carryover from the separator 3. Valve S2 leaking 4. Adsorbent lacking adsorption capacity. 5. Absorber on-line too long: breakthrough 6. Not enough regeneration time 7. Condenser not cooling sufficient 8. Instruments wrong, pressure 9. Absorbent broken down 10. Temperature TRC, T3 wrong

Initial Evidence a S S

b N N

S S S S S N ? S

N N

c

Diagnostic Actions d

e

A

B

C

D

S S S

For the Diagnostic actions or tests, Saadia thought that the root cause of a high Dp could be “damaged alumina adsorbent” causing the Dp across the bed to be too high. One factor that could cause such damage would be excessive temperature in the regeneration. This might be caused by an incorrect TRC T3 sensor that reads 177 C but it really is 245 C, for example. Such damaged adsorbent would also have less adsorbent capacity. Test A. Check/calibrate T3. Test ?: Another test would be to open up the adsorbers and sample the adsorbent, checking the adsorption loadings and checking for decomposition. However, this is expensive and takes the unit off-line. She preferred to do other simple on-line tests first. Test II and III: Gather a set of data that can be used for four other tests, called B, C, D and E. Since this is a batch-cyclical process, most of tests require data that are gathered every 15 minutes over the whole cycle and with bed A being both adsorber and regeneration mode. Since the steam trap and the separator traps discharge below grade, data about the steam/condensate flowrate are not easily accessible. However, the trap on the separator can be isolated and condensate collected from S2. In Test II, 6-hour cycle with the proportional valve closed, measure the exit moisture concentration in the air at S4, collect condensate at S2, measure temperature T4, plus all other usual readings. In Test III, a 6-hour test with the proportional valve open and a corresponding small flow of air through the regenerator heater and the condenser-separator, measure the exit moisture concentration in the air at S4, measure temperature T4, plus all other usual readings. Such information could then be used to do four other tests: Test B. Mass balance on moisture. Excessive water coming out might suggest a steam leak. Test C. Compare adsorption specs on alumina with moisture adsorbed in one cycle. Low adsorption/ mass of adsorbent means the cycle time is not long enough to load the adsorbent or the alumina is lacking the adsorption capacity.

4.4 Case ’6: The Case of the Utility Dryer

Test D. Compare time plots of temperature T4, liquid collected S2 and Dew Points from S1, S4. She sketched the time plots she expected. These are shown in Figures 4-1 and 4-2.

Figure 4-1

Prediction of the adsorption cycle.

Figure 4-2

Prediction of the regeneration.

149

150

4 Trouble Shooting in Action: Examples

Figure 4-1 shows a typical loading with the exit concentration below the target 15 ppm. Near the end of the cycle, the concentration may increase as the start of the breakthrough curve appears. That’s what I would expect to see for all the tests. Figure 4-2 shows what I would expect for the regeneration and condensation of the water from the separator. During the hot regeneration period, the temperature on T4 should increase and level off on a plateau around 105 C and then rise to the inlet temperature, 177 C when the bed is almost regenerated. If regeneration was started before the bed was fully loaded, then the alumina near the end of the bed will be dry, then wet and then dry as the desorption band sweeps through the bed. If the bed loading is small (or the gas velocity high) then the length of the temperature plateau is short. The condensate collected should spike before the cool-down period. For the cool down, the effluent temperature T4 should continue to rise but will drop after the high-temperature wave from the heated, regenerated adsorbent passes through the bed. Test E. Compare pressures on P1–P4 for four different conditions: a) the heating cycle for regeneration versus b) the cooling cycle (to see Dp over heater) and d) when the proportional valve is shut versus e) partially open to see the Dp in the adsorbing bed versus including the Dp for regenerating bed and cooler-separator. These tests should be done with both A and B beds being adsorbers. Before I order the tests, let me check. She reflected on how she had handled the program so far, her hypotheses and the proposed tests. Satisfied, she planned the tests. What were the strengths and weaknesses of Saadia’s approach so far? What would you have done? Should Saadia have just replaced valve V2? Should Saadia have called in the vendor? Should she have opened both beds and tested the alumina? What would you do next? Saadia was pleased that she had resisted the urge to jump in and change Valve 2, test the alumina and to call in the vendor. The 2 days of tests and analysis of data should clarify many issues and dispel incorrect hypotheses before such major action is taken. Here are the results. Test A: TRC 1 responds to change and calibrates OK. Conclude: the regeneration air temperature is 177 C. Hypothesis 10 that the regenerated air temperature is hotter than 177 C is disproved. Test B: Mass balance on moisture. Test II, for the 6-hour cycle with the proportional valve “closed”; exit gas flowrate, FR 1 = 2830 Ndm3/s; loading bed B for the first 3 hours and collecting the condensate from the regeneration of B during the second 3 hours: 79.7 kg in =? 66.1 kg out. For bed A: 79.7 kg in =? 74.3 kg out. The balance isn’t within the 10% that I like to see, but this suggests that steam is not a source of water coming into the system. The condensate collected is less than

4.4 Case ’6: The Case of the Utility Dryer

the apparent bed loading. Reject hypothesis 1 that there is a steam leak and perhaps consider poor separation in the separator with resulting carryover of entrained water to the adsorber. Test C: Check adsorption loading of water/ mass of dry adsorbent. From my files for activated alumina: 0.14–0.22 kg water/kg dry adsorbent. Calculated from the vendor specs: 0.089 kg water/kg dry adsorbent for air flow 7300 Ndm3/s Calculated from Test II for the 6-hour cycle with proportional valve “closed”; exit gas flowrate, FR 1 = 2830 Ndm3/s; loading bed B with subsequent regeneration: 0.016 kg/kg. Notes: for the Test II conditions the flowrate is below vendor design specs so the moisture coming to the unit is less. If this incoming moisture is increased in proportion to the flowrates, however, the loading on the adsorbent becomes 0.016 7300/2830 Ndm3/s or 0.0412 kg/kg. This suggests that the adsorbent is not adsorbing the amount of water expected from the vendor design specs (0.09 versus actual 0.04 or about 12) nor that expected based on file data about activated alumina (0.14 versus actual 0.04). Conclude that Hypothesis 4 might be true. Test D: Plots of data are given in Figure 4-3 for Test II, the 6-hour cycle with the proportional valve closed. The data are the exit concentration S4 from the adsorbing bed for beds A and B respectively, in Figures 4-3 a and b. What a surprise! For both beds A and B on the adsorption cycle, there is a spike of water in the middle of the cycle. A comparison with the water evolution from the concurrently occurring regeneration, in Figure 4-4 a and b, shows consistency in the spike of water. This suggests leakage across valve V2. Saadia estimated the amount of gas leaking through the valve based on the times of the peak condensate collection to be about 1.5–3 dm3 / s. I conclude that Hypothesis 3 is correct.

Figure 4-3a

Test II, bed A adsorb with proportional valve closed.

151

152

4 Trouble Shooting in Action: Examples

Figure 4-3b

Test II, bed B adsorb with proportional valve closed.

Figure 4-3c

Test III, bed A adsorb with proportional valve open.

4.4 Case ’6: The Case of the Utility Dryer

Figure 4-3d

Test III, bed B adsorb with proportional valve open.

Figure 4-4a

Test II, bed B regenerate (while bed A is adsorbing) with proportional valve closed.

153

154

4 Trouble Shooting in Action: Examples

Figure 4-4b

Test II, bed A regenerate (while bed B is adsorbing) with proportional valve closed.

Figure 4-4c

Test III, bed B regenerate (while bed A is adsorbing) with proportional valve open.

4.4 Case ’6: The Case of the Utility Dryer

Figure 4-4d

Test III, bed A regenerate (while bed B is adsorbing) with proportional valve open.

Test III, a 6-hour test with the proportional valve open and a corresponding small flow of air through the regenerator heater and the condenser-separator. The adsorption results are given in Figures 4-3c and d. Bed A seems to show the appropriate adsorption behavior. Bed B exhibits breakthrough of the water front. No condensate from S2 was gathered from this run so a mass balance could not be done. Since the flow through the cooler is relatively small, I am beginning to think that the separator may not be operating well, based on high concentrations of water found at S3 of > 24 000 ppm. The temperature profiles during regeneration are given in Figures 4-4 c and d. These profiles are not what I expected even though Bed B adsorbed only a small amount of water. Test E. Pressure drop profile: P1–P4. Vendor: = 14 kPa. Simple spot test I with proportional valve closed = 27 kPa. Test II: proportional valve closed: regeneration of bed B and loading of bed A: Dp= 62 kPa and remained the same during the cooling cycle. I conclude that the pressure drop across the heater is negligible because there was negligible change in Dp when the heater was bypassed via the 3-way valve V1. For the regeneration of bed A and the loading of bed B: Dp= 125 kPa with negligible change when the heater was bypassed. Test III: proportional valve is open with most of the air going directly to the adsorbent bed: for bed A adsorbent Dp= 34 kPa; with bed B adsorbent Dp= 62 kPa. These results suggest that there is a significant Dp across the cooler-separator and that the adsorbent in bed B, with the larger Dp and early breakthrough behavior, should be tested.

155

156

4 Trouble Shooting in Action: Examples

Saadia used the data to perform an energy balance over the regeneration operation and found, to her surprise that the energy would balance only if the flow through the “closed proportional valve” was 1270–1360 Ndm3/s. The balance could also be closed if 60% of the water entering the separator was entrained. Saadia arranged for the 4-way valve, V2, to be changed, and she contacted the vendor. For the new valve she recommended the replacement of the iron core with a stainless steel core. She also was concerned that the temperature of the gas leaving regeneration exceeded the plateau temperature of about 110 C. Discussion Overall rating of the difficulty of this problem 9/10, relatively difficult because many faults are present. Saadia took several days to identify the fault in the valve and to define a series of concerns about this unit. Overall, Saadia drew on her experience with adsorption, especially in designing the tests and predicting what she expected to see before the test results were obtained. She astutely chose to follow the Worksheet because this setup was different from her experience. This was a “problem” and not an exercise for Saadia and she handled it that way. Problem Solving

Monitored, especially by predicting the shape of the adsorption and regeneration curves before doing the tests. She was organized and systematic. She checked and double checked. Her dominant P style showed through in that she delayed changing the valve and calling in the vendor until she had more data. Overall a rating of A. Data Handling

Actions were based on fundamentals. She related the test to the hypothesis. Good reasoning displayed. Overall rating B+. Synthesis

She listed a variety of hypotheses. Considered many viewpoints. She had large masses of data to consider; she managed this well. Overall rating A. Decision Making

She seemed to continually try to prioritize. No obvious bias was apparent. Criteria were present for some of the decisions, but for others, the criteria were not noted explicitly. Overall rating B+. Strengths: Recognized the need to problem solve, used fundamentals, systematic, managed large volumes of data well, rich set of hypotheses, did not get discouraged. Areas to work on: provide more explicit details for decisions.

4.5 Case ’7: The Case of the Reluctant Crystallizer

4.5

Case ’7: The Case of the Reluctant Crystallizer

Frank has 25 years of experience. He is a process engineer responsible for several processes on site including the crystallization unit. The phone call comes in from Phil, the operator in the control room. “Something wierd is happening on the vaccum crystallizer, Frank. We’ve been operating the VC for about two hours now. The first hour with only two ejectors turned on, the operation was OK. We turned on the booster ejector and it “held” for the first 1/2 h or so and then it “kicked out”. While the booster ejector was “holding” the liquid level in the crystallizer drops at a fantastic rate. What’s going on here? We just can’t produce quality crystals under these conditions.” “I agree, it sounds wierd. I’ll be right out,” responds Frank. Frank has developed a good working relationship with all the operators so it doesn’t surprise him to receive the call and for Phil’s succinct summary. He took out the diagram for the process and mentally reviewed his thoughts. The vacuum crystallizer, VC, is the core of this operation. Pregnant liquor feed enters the vacuum crystallizer at 55 C where it is concentrated and cooled to precipitate the product. This is a batch process. To start, we pull a vacuum on the VC (using the two-stage steam ejectors with city water to the interstage condenser), open the valves from the feed tank and syphon feed into the VC until the unit is 2/3 to 3/4 full, as seen on the sight glass. The feed valves are then shut and the eight-hour batch process begins. While operating with the two-stage ejector we estimate the absolute pressure to be about 6.5 kPa abs. After about an hour, the temperature of the liquor has decreased to 40 C and the liquid level has dropped by about 10 cm. At that time, the booster ejector and the second barometric condenser are turned on to decrease the pressure further to about 2.5 kPa abs. At the end of the 8 hours the liquid level has dropped about 40 to 50 cm. The temperatures of the water in both barometric condensers are monitored to ensure that the water temperature does not exceed 26 C. The operators have some control over this temperature by judiciously mixing cold city water with warmer water from the bay. Experience has shown that if the booster ejector is turned on too soon, it will not “hold” but rather “kicks out”. “Kicking out” produces a very distinctive sound. This jargon has been used ever since Frank started on the unit. “Kicking out” has been interpreted as being when the steam controlled by valve S3 goes directly into the VC instead of through the ejector nozzle and thus does not pull the required vacuum. Frank headed out to the control room. Here are some of his thoughts. Steam ejectors ... . they are so sensitive to upstream steam pressure. Frank mulled over his experience: “unstable operation or loss of vacuum”: steam pressure < 95% or > 20% of design/ steam superheated > 25 C/ wet steam/ inlet cooling-water temperature hot/ cooling-water flowrate low/ condenser flooded/ heat-exchange surface fouled/ 20– 30% higher flow of non-condensibles (light end gases, air leaks or leaks from fired furnaces)/ seal lost on barometric condenser/ entrained air in condenser water/ required discharge pressure requirement high/ fluctuating water pressure. Hmm.. It was a hot summer day.. so the bay water is likely hotter than usual; the maintenance

157

158

4 Trouble Shooting in Action: Examples

turnaround was only a month ago and the barometric condenser was inspected and looked OK. Everything had operated fine since the turnaround. We usually operate the steam to the booster ejector with the valve half open. Hmm, lots of possibilities. By this time Frank had zeroed in on five hypotheses: . . . .

.

steam to the booster ejector was at too low a pressure, steam superheated or wet, cooling-water inlet temperature too high, loss of vacuum caused maybe by loss of seal in the barometric leg or leaks in VC, pressure in the barometric condenser higher than usual.

In a moment he would be in the control room. He formulated some questions he wanted to ask Phil before he went out on the plant. What were the strengths and weaknesses of Frank’s approach so far? How would you have handled the call from Phil? Would you go directly to the plant before seeing Phil so that you are better prepared to ask questions? Would you have phoned the boiler house first to see if they had upsets? What questions would you ask Phil? What are your hypotheses? What is the evidence so far? Frank realized that he worked better when he wrote things down but he wanted to do that with Phil’s help. Furthermore, one of the reasons Frank got along so well with plant operators is that he always went to the control room first, before ever venturing out on the plant, and he always respected their experience and ideas. After greeting a worried Phil, Frank and Phil sat down to put their thoughts on paper, the way they usually did in situations like this. Phil knew that Frank liked to use the IS and IS Not approach to summarizing the evidence. “Check me if I’m wrong,” said Frank, “but this is what we have so far.” He wrote: IS: what and when: operating OK during startup, feed intake and first hour of operation before the booster ejector started. The booster ejector operates on “hold” without “kicking out” for an hour. IS NOT: what and when: is not operating as usual when booster operating as “hold” with evidence that the liquid level (on the sight glass) drops “at a fantastic rate”. Phil looked at the list and said “Looks OK.” and added “When the liquid level started to drop I went up and listened to the booster ejector and it sounded fine. The only thing I noticed was the pressure gauge on the bay water line to the booster condenser was fluctuating wildly. I also knew that you attribute most vacuum-system malfunctioning to steam pressure to the ejector. I checked the steam pressure gauges on P1, P2, P3 and P8. They are all 550 kPa g and rock steady. Gauge P4 was 725 kPa g, as usual, and steady.” Frank said, “Good thinking. I need some clarification. What did you observe when the liquid level drops at a fantastic rate?”

4.5 Case ’7: The Case of the Reluctant Crystallizer

“The level in the sight glass dropped about 1 cm every two seconds. You could visibly watch it go down.” “Thanks. When you checked the pressure gauge P8, did you tap the gauge?” “Yes, I tapped it and it seems to be working OK. The pressure on all steam gauges was 80 psi.” Frank mentally converted this to 650 kPa abs, that was normal steam pressure. “The pressure on the bay water line usually reads 205 kPa g; today it fluctuated wildly. How extensive was the variation, Phil ?” “We don’t normally check that pressure. It isn’t shown in the control room. Out on the plant that gauge needle was jumping back and forth between 140 and 550 kPa g. Wild!” Thinking about today’s hot August temperature Frank asked “What about the temperatures on the barometric legs?” “The temperatures on both legs were usual, certainly less than 27 C; T2 was about 24 C and T1, about 22 C.” “Just for the record, what were the readings on the instruments you checked on the plant and how do these compare with the usual valves?” “The steam pressures I talked about earlier; pressure on the city water P6 was the usual: 310 kPa g (45 psi) before the valve and P5 was 0 to 35 kPa g (5 psi) after the valve. I didn’t note the temperature of the liquor in the VC. The level had disappeared below the sight glass. That’s when I called you.” Frank added that evidence to his charts and thought about his hypotheses now in the light of the new evidence: .

.

.

.

.

Steam to the booster ejector was at too low a pressure. But all the pressure gauges read the same usual values, and they are steady! Perhaps the pressure gauge P8 is faulty. Steam superheated or wet. Can’t tell, but if it was then the other ejectors should malfunction as well. They don’t; so this is unlikely. Cooling water inlet temperature too high. Still possible but the exit temperatures in the barometric legs suggest this is not a cause. Loss of vacuum caused maybe by loss of seal in the barometric leg or leaks in VC. Perhaps; relatively easy to check, so why not? Pressure in the barometric condenser higher than usual. No gauge is available on the condenser or downstream of valve W5. Not easy to tell but perhaps this is connected to the wild variation in pressure in the bay water coming in. Perhaps valve W5 has a worn valve stem that is vibrating and causing the oscillations in pressure. Slugging flow of water? Is the bay water pump behaving itself? Should it be checked?

So to sum up, Frank is down to three hypotheses: 1) steam pressure to the booster is low or fluctuating but the pressure gauge is broken. Check the pressure gauge. 2) changes in the vacuum with leaks; questionable but let’s check. Frank can’t see how the fluctuating pressure gauge on the bay water is connected to rapid evaporation in the VC. However, he decides to change his focus to be “Why is gauge P7 fluctuating?” He hypothesizes that P7 is fluctuating because of 3) fluctuations in the pump

159

160

4 Trouble Shooting in Action: Examples

exit pressure from the bay water pump or oscillations of the valve stem in valve W5. What’s going on here? What were the strengths and weaknesses of Frank’s approach so far? What would you have asked Phil? What are your hypotheses now? What has Frank missed? Is Frank exhibiting pseudodiagnosticity or fixation when he keeps looking for a steam fault to the booster ejector? Should Frank have considered the evidence of “kicking out” of the booster ejector? What would you do next? Frank makes a list of his next steps. 1.

2.

3.

While the plant is still operating, visually inspect the VC for a leak. Attach a gauge to the VC via the valve and nozzle at the top and read the vacuum/ pressure noting it every 10 minutes for two hours. Hypothesis, if there is a leak, the pressure will gradually increase. If the pressure remains constant, then there is no leak. Shut down the process, and remove and visually check valve W5; while we are at it check valve W4. Frank realizes that this is not going to confirm any of his hypotheses, except perhaps his wild idea of a vibrating valve stem in hypothesis 3. However, he still can’t see any connection between the oscillating pressure and the rapid evaporation. However, since the change is occuring near the valve, let’s check out the valve. To test hypothesis ’1, either replace the pressure gauge at P8 or recalibrate the existing pressure gauge P8.

Frank hoped that the results of these tests would resolve the mystery. The results were: .

.

.

Hypothesis ’1) steam pressure to the booster is low or fluctuating. Evidence is that gauge P8 is working and calibration is correct. Hypothesis denied. Hypothesis ’2) changes in the vacuum with leaks. Evidence is that a visual inspection and the vacuum test could be interpreted as being no leaks. Hypothesis denied. Hypothesis ’3) fluctuations in the pump exit pressure from the bay water pump or oscillations of the valve stem in valve W5. Frank received telephone confirmation from Utilities that the exit pressure on the bay water pump is constant, steady and the usual value. Valve W5 (and valve W4) were removed, visually checked as OK. A frustrated Frank decided to replace W5 with a new valve anyway. The process is started up. A number of other engineers have gathered in the control room to see if Frank has solved the mystery. The system behaves the same as it did before the changes: the booster ejector “held” for the first 1/2 h or so and then it “kicked out”. While the booster ejector was “holding” the liquid level in the crystallizer drops at a fantastic rate.

4.5 Case ’7: The Case of the Reluctant Crystallizer

Frank revisits his starting hypotheses and notes that perhaps the water temperature is too high in the barometric condenser. After all it’s a really hot day. Frank follows his dominant J behavior and decides to take action and “correct the fault” rather than gather more data. Frank repiped the system with a temporary hose that sent city water into the bay water feed line so that only cold city water went to the condenser. The use of the hotter bay water was eliminated. Unfortunately this repiping makes no improvement! Frank returns to his office. “It’s back to basics! I’ve never seen anything like this before. I can do this!” What can cause the liquid level in the crystallizer to drop at such a rate? Brainstorm: liquid is leaking out, the change is in the sight glass and not in the vessel; something is sequestering the liquid, the vacuum is much higher than expected and the liquid is flashing off, the temperature is higher than expected, the absolute pressure is much lower than expected; steam flow into the booster is higher than expected; the steam flow is oscillating but its oscillations are so fast that the changes aren’t picked up on the steam gauge P8 but are picked up on the water load and reflected on P7; the fluctuating water flow into the condenser is causing a vacuum that is pulling off more vapor than expected; liquid is going back into the feed tank; the discharge line is open, liquid leaks around the mixer shafts; the temperature gauge on the VC is broken so the temperature is much higher than it reads; liquor from the VC is entrained and not evaporated; liquor goes out through the booster ejector. Now let’s refocus and brainstorm on the causes of wildly fluctuating gauge P7 when the bay water pump exit pressure is stated as being steady and the usual value. Frank wrote down the following: change in water pressure, upstream oscillations, oscillations in the steam, oscillations in the liquid level, oscillations in condenser, city water hitting the bay water in condenser, oscillating liquid level in barometric leg, vibrating baffles in condenser, vibrating valve stem on water, water bypassing periodically directly into the barometric leg, corrosion particle in the pressure tap, unstable bourdon tube to pressure gauge, bay water pressure is not steady. “I recall the case of the filter press operation was affected by the upstream oscillations of a centrifuge”, murmured Frank to himself. “An idea that links the rapid evaporation and the fluctuating pressure gauge could be oscillations in the steam flowrate going into the booster ejector.” Frank hasn’t quite given up on the idea that the steam into the ejector is the key. Frank headed back out to the control room. Frank and Phil went to scrutinize the steam line into the booster ejector. The valve was a globe valve that was operated partly open. When Phil carefully adjusted the valve he felt some vibration. “Maybe that’s it,” exclaimed Frank. They shut down the troubled plant, removed the steam valve and found that the valve stem had been severely eroded. They replaced the steam valve S3 with a tapered plug valve. The replacement of the valve solved the problem. Case dismissed. Discussion The overall degree of difficulty of this problem is 10/10, very difficult. Frank took a week to solve this problem. The costs in loss of product and loss of pregnant liquor carryover into the hot well was extremely high. Let’s reflect on Frank’s approach

161

162

4 Trouble Shooting in Action: Examples

following the guidelines in the feedback form. Frank started this problem using an exercise approach although even here he showed excellent problem-solving characteristics: IS and IS NOT, systematic, patient. Problem Solving

Systematic and organized even though he was very frustrated. Patient and takes time to understand the system (even though he had worked many years on this plant). Some monitoring. Some checking. He’s active and wrote things down. Overall rating B+. Data Handling

Based on fundamentals–especially after he recognized it as a problem and not an exercise. Very systematic in planning and gathering evidence. Used simple tests. Overall rating A–. Synthesis

Used a variety of hypotheses. Flexible. Kept an open mind even when nothing seemed to be working out. Overall rating A–. Decision Making

Prioritized. His bias about the low steam pressure actually paid off. Decisions tended to be made intuitively with few explicit criteria. Overall rating B. Strengths: excellent interaction with the operators, very systematic, didn’t panic, flexible, time well spent exploring and understanding the problem, patient. Areas to work on: more monitoring and more explicit thought process especially with criteria for decisions.

4.6

Reflections about these Examples

In each of the cases, the engineer “solved the problem.” Each used a different style. The less- experienced engineers, Michelle and Dave, used the Trouble-Shooter’s Worksheet with varying degrees of comfort. The more-experienced engineers, Pierre, Saadia and Frank, tended to resort to the Trouble-Shooter’s Worksheet as needed. Pierre used it to remind himself to try the Why? Why? Why? approach. He might have tried this without resorting to the Worksheet. Indeed, the problem he was addressing seemed to be one of the few where that technique was useful. Saadia recognized the uniqueness of her problem early on and, although she had a lot of experience with adsorption, she wisely disciplined herself to work slowly and systematically through the Worksheet. Of the five engineers she did the best job of using the “hypothesis-symptoms-actions” chart. Frank treated the problem like an exercise until he discovered it really had more unusual components than he had usually encountered. Before it was too late he started to use the Worksheet.

4.6 Reflections about these Examples

What components of the Trouble-Shooter’s Worksheet seemed to be used more effectively? All tried to slow down the process and work carefully in the Engage process, although Pierre’s efforts were not the greatest. This was astute of the trouble shooters because most mistakes are made right at the start. Dave almost got caught here. Pierre was stymied here for a frustrating few moments. The Define the stated problem stage was handled better when all five engineers used the IS and IS NOT tactic effectively. For the Explore stage, many used a variety of approaches and, indeed, were selective in the elements they included. Dave used the What if? elements effectively. The less experienced engineers drew on the symptom-cause information in Chapter 3 to help them overcome their limited practical experience. Fortunately, for the problems they were working on some data were available. For some equipment, the published information is lacking. The “hypotheses-symptoms-action” chart was only used well by Saadia. The others, especially Frank, tended to use the ideas intuitively, but such an intuitive approach has pitfalls. Saadia also excelled at using her fundamentals to predict, ahead of time, what she expected to see from the results of the tasks. Frank seemed to be the only one who was comfortable working with and drawing on the experience of the operators. This didn’t come without patience in developing trust and a good working relationship with Phil. Michelle, and Dave, should polish their skills in this area.

163

165

5

Polishing Your Skills: Problem-Solving Process Five components are useful in developing your skills: becoming comfortable talking aloud about your thought processes, identifying a strategy and monitoring the stages you use, defining the problem, creativity and self-assessment. Activities to develop your awareness, skill and confidence in these skills are given1). To gain the most, select the areas you want to work on and complete the activities. Developing skill is not a spectator sport. Participate.

5.1

Developing Awareness of the Problem-Solving Process

Awareness is the ability to describe what goes on in our mind as we solve problems and make decisions. Such awareness is a prerequisite for the triad improvement activities in Chapter 8. Such awareness also helps us improve our problem solving – and trouble shooting – skill because: . . .

. . . .

we can explicitly identify “where we are in a problem”, we can compare “how we do it” with how “others do it”, we can identify our strengths and areas to work on to improve our problem solving, we can get ourselves “unblocked” if we cannot seem to solve a problem, we need to be able to describe our thoughts for team problem solving, we can describe our thought processes to another so that we can improve, we build on this awareness to develop skills with strategies, in Section 5.2.

One of the more successful techniques to quickly acquire confidence and skill in “awareness” is the Whimbey pairs or Talk Aloud Pairs Problem Solving, TAPPS, approach. Two people are needed. One plays the role of a “talker” or problem solver; the other plays the role of the “listener”. 1) The materials in this chapter are from the MPS program, copyright

Donald R. Woods 1982 ff and used with permission.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

166

5 Polishing Your Skills: Problem-Solving Process

In this section we list the target skills that are developed by doing the activities in this section, describe the roles, give the activities and the feedback forms. 5.1.1

Some Target Skills

Research in problem solving has shown that successful problem solvers exhibit the following characteristics (related to awareness and the thought processes and the activities in this section): They .

. . .

.

. . .

.

are aware of their thought processes and use that awareness to identify where they are in the process and get themselves unstuck, are skilled in describing aloud their thoughts as they solve problems, pause and reflect about the process and about what they are doing, accept that their particular style works for them; others may have a different preferred style, are active as opposed to passively trying to remember stuff. They are active by writing things down, making charts and diagrams. Being active helps overcome the space limitations of Short-Term Memory (the portion of the brain where problem-solving tasks are done. This portion can only “hold” five to sevens bits of information at a time.) focus on accuracy and not on speed, accept that problem solving is a social process; we interact with others, know that self-assessment is about performance and not about them as a person, know that self-assessment is based on evidence and not on gut feelings or wishful thinking, and so to assess progress in developing skills they will use feedback forms and written evidence and reflections.

In addition, this activity provides an opportunity to 1. 2. 3. 4.

Acquire some skill at listening, Acquire some skill in self-assessment, Acquire some skill in giving and receiving feedback, Through self-awareness, begin to improve self-confidence.

5.1.2

The TAPPS Roles: Talker and Listener

Talker/ problem solver: In pairs, with a partner as a listener, the talker reads the problem statement aloud, and talks aloud as he/she works on/solves the problem. The goal is not to get the “right” answer to the problem. The goal is to talk aloud continuously about the process. The goal is to have, in a ten-minute period of talking, fewer than two silent periods of more than 10 s duration. The goal is to focus on accuracy. The goal is to be active and write things down.

5.1 Developing Awareness of the Problem-Solving Process

Your listener will not give you any hints about how to solve the problem. Your listener will help you to talk about what you are thinking so that the listener can follow and understand what you are saying. More specifically, here is what you do in this role: 1. 2. 3.

4.

5.

6.

Sit side by side; have paper and pencils available. The talker starts by reading the problem statement aloud. Then start to solve the problem on your own. You are solving the problem. Your partner is only listening to you. He or she is not solving the problem with you or for you. Talking and thinking at the same time is not easy. At first you might find it hard to think of the right words to use. Do not worry. Say whatever comes to your mind. No one is testing you or marking you. You are playing the role. Go back and repeat any part of the problem you wish. Use such words as “I am stuck! I do not know what to do! Maybe I should read the problem statement again.” Try to solve the problem no matter how easy it is. You are learning to talk about your thinking methods. We use simpler problems to help make this activity as easy as possible.

Listener: You have an important and difficult role to play. You are to help the talker talk. You are not to solve the problem nor to give hints to the talker about how you would solve the problem. This is the problem talker’s turn. You will get a turn later to be talker. Encourage them to talk aloud. At the same time, you are to monitor their thinking. Can you understand what they are saying? Can you follow the path that their mind is following? Could you describe what they are thinking to others? You are to help them to talk about the mental processes they are using–no matter how silly or incorrect they might be. You must not laugh at them. You must not criticize them and tell them that they are wrong. If you think they made a mistake, then say “Can you check that?” or “Are you sure?” Do not tell them what they should be doing. Do not tell them what you think is the “correct” answer. 1.

2.

Help the talker to see that you are not a “critic”. Instead, you are a helper-fortalking. You might say “ Please keep talking.” or “I was not able to understand or follow what you just said; would you please explain.” “Can you tell me what you are thinking now?” “Do not worry about how it sounds – just say an idea about what you are thinking.” “Can you check?” “Are you sure?” “OK” “Ahmmm”. (For more suggestions about listening, see Section 7.1.2.) You might tell the talker that your role is to: – Remind them to keep talking, – Help them improve the accuracy of their talking about their thinking by saying “Can you check?” “Are you sure?” – Be able to understand and follow each step of the talker’s thinking.

167

168

5 Polishing Your Skills: Problem-Solving Process

3.

4.

Do not turn with your back to the talker and try to solve the problem on your own. Do not solve the problem on your own and then tell the talker what they should do. Do not let the talker continue if: – you do not understand what they have done. Say “I didn’t quite understand that, could you please elaborate on your thoughts, Thanks.” – you think that a mistake has been made. Do not say, “You have made a mistake.” You might say “Are you sure?” or “Do you want to check that?”

Your goal, when you receive feedback on how well you played the listener role, is that you should be within two scale ratings of “about what I wanted” on both the “degree of interaction” and the “tone of the interaction.” 5.1.3

Activity 5.1: (35 minutes)

This activity will take 35 minutes. 3 minutes getting set up; 10 minutes talker A, 5 minutes reflection, complete feedback forms; switch roles, 2 minutes getting set up, 10 minutes talking, 5 minutes reflection, complete feedback forms. .

.

.

.

. .

Getting set up. (3 min) Find a partner, flat table, two chairs side by side, pencil and paper. One person starts as talker, the other is the listener. Read over your role (Section 5.1.2). Talker selects an “exercise” (or perhaps a “problem”) from among Tasks 5.1A to E. First 10-minute talk (10 min): talker talks for 10 minutes. If the talker completes the selected task early, then the talker selects another task. Do not change roles! The talker talks for 10 minutes. Reflections, discuss and feedback forms (5 min). As individuals, write down reflections about the activity, use Worksheet 5-1. (1.5 min). In pairs, discuss what you experienced. Don’t continue working on a Task. (1.5 min). Listener completes feedback form in Worksheet 5-2 and signs it; gives it to the talker as evidence. Talker completes feedback form in Worksheet 5-3 and signs it; gives it to the listener as evidence. Worksheets 5-1 to 5-3 are given in Section 5.1.4. Switch roles and get ready (2 min) Read the instructions about your new role (Section 5.1.2). Talker select a Task from among Tasks 5.1.A to E. Second 10-minute talk (10 min). See instructions above. Reflections, discuss and feedback forms (5 min) as above.

Content-general A.1 (based on Lochhead and Clement, Univ. of Massachusetts) At Mindy’s restaurant, for every four people who ordered cheesecake, there are five who ordered strudel. If C represents the number of cheesecakes ordered and S represents the num-

Task 5.1.A:

5.1 Developing Awareness of the Problem-Solving Process

ber of strudels ordered and M represents Mindy’s restaurant, write an equation using these variables to represent this situation: a) b) c) d) e) f) g) h) i) j)

M: 4C = 5S 4C M = 5S M 4C = 5S Total= 4C +5S (5/9) S = C M: 5C = 4S 5C M = 4S M 5C = 4S S = (4/9) C other

A.2 (reprinted courtesy of Art Whimbey) There are two clocks, A and B. Clock A keeps perfect time but clock B runs fast. When clock A says 4 minutes have passed, clock B says that 6 minutes have passed. Both clocks are set correctly at 5 a.m. What is the correct time when clock B shows 9 p.m.? Tasks related to science and engineering B.1 (based on Lochhead and Clement, University of Massachusetts) A pillar supports a floor. Is the pillar doing any work?:

Task 5.1.B

a) Yes, of course it is doing work; otherwise the floor would come crashing down. b) No. c) Yes, a definition of work is a force acting over a distance; the force is the force pushing up on the floor above and the distance is the height of the pillar d) Yes, we know this because of energy considerations. The potential energy is the mass of the slab or floor times the height of the pillar that is supporting it. If the pillar wasn’t there, the potential energy would turn into kinetic energy and would eventually be released in the form of heat and sound as it crashed down. e) Other General tasks more related to trouble shooting. C.1 A researcher predicted that if part of a leaf is in the shade and the other part of the leaf is in the sun, then equal amounts of starch will be found in both parts of the leaf. Which of the following hypotheses is the researcher most likely assuming is true?

Task 5.1.C

1. 2. 3. 4. 5.

Chlorophyll is present in both the shaded and sunny parts of the leaf. In the shaded part of the leaf photosynthesis is increased. Carbon dioxide and water can enter the leaf cells in both the sunny and the shaded parts of the leaf. By shading part of the leaf, photosynthesis is increased in the sunny part of the leaf. Starch can move to different parts of the leaf.

169

170

5 Polishing Your Skills: Problem-Solving Process

Engineering related D.1 Oversized condenser The overhead condenser on the distillation column is oversized by 38%. That is, it has the correct number of tubes to promote turbulent flow inside the tubes. However, the length of the tubes has been increased. The baffle spacing on the shell side and the baffle window has been sized for the design overhead rate. This is in addition to the usual allowances for fouling. The condenser is horizontal. Which of the following observations are consistent with this situation when the condenser is first put into service:

Task 5.1.D

a) The exit cooling-water temperature will be colder than expected. b) The pressure drop on the shell side will be 1.382 more than we expected. c) The reboiler will act as though it were undersized and the column operation will be unstable with vapor blanketing in the reboiler because film boiling now occurs. d) Nothing unexpected; the controller action will account for the overdesign. e) Increased power required on the cooling-water circulation pumps. f) The feed location should be changed to be closer to the reboiler. h) Other

Cooling water

LIC 1

FRC 1

Feed

TRC

LIC 3

Steam

Figure 5-1

A column.

5.1 Developing Awareness of the Problem-Solving Process

Technical tasks related to methane-steam reforming E.1 In the methane-steam reformer shown in Figure 5-2, the total feedrate to the reformer is 25% higher than the flow instrument reads. The other instruments usually read reformer outlet temperature T1 = 454 C (850 F) with the exit methane concentration of 9.5 mol %. The tubewall temperature is usually 960 C (1765 F). The design Dp= 345 kPa (50 psig). The reformer works on outlet temperature control. Which of the following observations accurately describe the situation: Task 5.1.E

a) b)

c) d) e) f) g) h)

Nothing different; everything works fine because the controllers adjust to yield the same exit temperature. The tube wall temperature will be about 30 C higher because the controllers adjusted to yield the same exit temperature; the exit methane concentration is 7.5 mol %. The tube wall temperature will be about 23.5 C lower than usual because of the cooling effect of the larger mass of gas flowing through. The inlet gas temperature will increase by 38.2 C because of the controller action. With controller action, the methane at the exit will increase to 10.5 mol % and the tube wall temperature will remain 960 C. With controller action, the Dp will be about 500 kPa. With controller action, the exit product gas temperature will be 20 C lower than design. Other.

Figure 5-2

A steam reformer.

171

172

5 Polishing Your Skills: Problem-Solving Process

5.1.4

Feedback, Self-Assessment

Self-assessment is based on written evidence. Throughout this activity there are several times when you are asked to write reflections and complete forms. This is a necessary part of the skill development process. The evidence can include: .

.

the problem statement you used when being the talker and the marks, underlines and notations you made directly on this. the paper you worked on.

Other, more structured evidence includes: . . .

your reflections you made after you did the task, Worksheet 5-1; the feedback from the listener about the task, Worksheet 5-2. the feedback from the talker to the listener about the role playing, Worksheet 5-3.

Worksheet 5-1:

Reflections: A place for you to record your ideas about the TAPPS

Method: Being the talker: What did you enjoy most? What was most difficult about the task? What did you discover by being the talker? What did you discover from interacting with a listener? What were your strengths? Focus on accuracy? Very few silent periods? Being active? Good communication? ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ Being the listener: What did you enjoy most? What was most difficult about the task? What did you discover by being the listener? What did you discover about problem solving by comparing the talker’s approach to yours? What were your strengths as listener? Quality of your prompts and degree of interaction? Tone of interaction? Good communication? Non-intrusiveness? ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________

5.2 Strategies

Worksheet 5-2:

Awareness

Feedback from the listener to the talker about the process used.

____________________________ ___________________________ problem listener

Number of silent periods 0 1 2 3 4 5 >5 Number of checks, double checks >5 5 4 3 2 1 0 Amount of writing/ charting >5 5 4 3 2 1 0 Comments: ___________________________________________________________________ ___________________________________________________________________ Validated by: ________________________________________________________ Worksheet 5-3:

Feedback to the listener:

problem _______________________

listener __________________________

I found the listener: .

The quality of the comments:

–10 –8

–6

–4

–2

J

–2

–4

–6

–8

–10

too passive little too passive about right a little interruptive too interruptive .

The attitude displayed:

–10 –8

–6

–4

–2

J

–2

–4

–6

–8

–10

too passive little too passive about right a little interruptive too interruptive .

The listener’s emphasis was listening to me &; helping me verbalize &; helping me solve the problem &; solving the problem for me &.

validated by talker __________________________________________________

5.2

Strategies

A strategy is an organized approach used in solving a problem. Such an organized approach identifies steps or stages for different parts of the process. Based on a survey of the cognitive literature and a critique of over 150 published strategies, the MPS 6-stage strategy, given in Figure 2-3, is a good working strategy to use. A strategy is important because: . .

we all usually use one, a strategy helps us to be organized and systematic,

173

174

5 Polishing Your Skills: Problem-Solving Process .

.

having a strategy helps calm us down if we become anxious when we are given a very difficult problem to solve, having a strategy helps us to “monitor” our mental processes.

One of the more effective ways to build our skill and confidence with the use of strategies is an extension of the Whimbey TAPPS approach used in Section 5.1. This is an extension of that experience in that the talker is also asked to move a marker to indicate the stage in the strategy in which he/she is working and asked to explicitly say monitoring statements frequently. The listener’s role is expanded to include recording the minutes the talker spends working in each of the stages in the strategy and recording monitoring statements made by the talker. In this section we list the target skills, describe the roles, give the activities and the feedback forms. 5.2.1

Some Target Skills

Research in problem solving has uncovered the following behaviors of skilled problem solvers as they relate to the use of a strategy: .

.

.

.

.

.

.

.

Spend time reading the problem statement (up to three times longer than unsuccessful problem solvers). Define the problem well; do not solve the wrong problem. Are willing to spend up to half the available time defining the problem. Most mistakes made by unsuccessful problem solvers are made in the define stages. The problem that they solve is their mental image of the problem; such a mental image is called the internal representation of the problem. Differentiate between exercise solving and problem solving (that were illustrated in Figures 2-1 and 2-2.) Use an organized strategy that focuses on defining the real problem, whereas unsuccessful problem solvers tend to search for an equation that uses up all of the given variables or given information (regardless of whether it applies to the situation or not). Define the real problem with a focus on key fundamentals whereas unsuccessful problem solvers tend to memorize and try to recall equations and solutions that match the description of the situation. Break “Defining the problem” into three separate activities to avoid errors. These are 1) listen, read, get initial information and manage both distress and panic (called Engage), 2) classify the initial information into the “stated goal or task to do”, “constraints and criteria”, and “description of the situation”, without trying to “define the real problem” (called Define the stated problem), and 3) create a rich, internal image of the problem as seen from many different perspectives and evolve a definition of the real problem (called Explore). Use a strategy so as to be systematic and organized, whereas unsuccessful problem solvers tend to take a trial and error approach.

5.2 Strategies .

.

.

Are aware that a strategy consists of a series of about six stages with each stage using different thinking and feelings. This strategy is not used serially (following rigidly one step after another). Rather it is used flexibly; applied many times while solving a single problem with frequent recycling from one stage to another. Problem-solving skill draws on subject knowledge (needed to solve the problem) and with the sample solutions (from past-solved similar problems) as was illustrated in Figures 2-1 and 2-2. Monitor their thought processes about once per minute while solving problems.

In summary, the goal of this activity is to develop your skill and confidence in 1. 2. 3. 4. 5. 6. 7. 8.

9.

Extending and reinforcing the skills addressed in Section 5.1 on Awareness. Recognizing patterns in the problem-solving process. Realizing that a “strategy” is not applied linearly and sequentially; that it is used flexibly. Recognizing the difference between problems and exercises. Understanding the relationship between subject knowledge, past solutions to problems and problem solving. Acknowledging the importance of defining problems and to recognize this as a three-step process. Acknowledging the importance of reading the problem statement. Realizing that problem solving is not “doing some calculations.” Conversely, to correct the misconception that if you are not “doing some calculations” you are not solving problems. Acquiring skill in explicitly monitoring the process.

5.2.2

The Extended TAPPS Roles: Talker+ and Listener+

Talker/problem solver+: This is an extension of the pairs activity with roles described in Section 5.1.2. Do all of the activities described in Section 5.1.2 and now the talker also 1)

2)

moves a marker on a strategy board, please use Figure 2-3, p. 21, to indicate which of the 6 stages you think is being addressed. By considering the location of the marker, the listener should agree 80% of the time that the activities you describe are consistent with the stage represented by the marker. You should need prompting no more than three times in a ten minute period. tries to say frequently such monitoring statements as “Where am I?” “Have I finished this?” “If I calculate ..., what will that tell me?” “If I ask this question ..., what will that tell me?”” Where do I go next?” “Can I check this?” If a hypothesis is shown to be wrong or if you calculate a “strange answer.” then ask “OK, What did I learn from that?” You should exhibit four verbal management statements during a ten minute period of problem solving.

175

176

5 Polishing Your Skills: Problem-Solving Process

This is difficult to do. Be patient with yourself. You may not completely understand the meanings of the stages yet. You may use different stages than the ones on the Strategy Board. Please, do your best. The listener will not move the marker for you. The listener will not tell you what stage you are in. The listener might ask you “Are you still in the “Explore” stage?” Remember to keep talking, to be active, to use pencil and paper, and to check and check again. Before you start, go over the meanings of the six different stages in the McMaster-6-Step strategy with the listener. Agree on the meanings of the words. 1. 2. 3.

4. 5.

Sit side by side; have paper and pencils available, have the Strategy Board and the marker. The talker moves the marker to the Engage part of the Strategy Board and starts by reading the problem statement aloud. Then move the marker to whatever stage you are going to work on next and start to solve the problem on you own. Keep talking aloud. You are solving the problem. Your partner is only listening to you. He or she is not solving the problem with you or for you. It is not easy to talk, think and move the marker at the same time. You might forget to move the marker. That is OK. Do the best you can. Go back and repeat any stage of the strategy you wish.

Listener+: This is an extension of the listener role described in Section 5.1.2. Here you are to record the amount of time the talker spends in each of the stages on the Strategy Board. Do not correct them; do not argue with them about which stage the talker is working on. Do not move the marker for them. You may have to ask “Are you still in the ... stage?” In all that you do, your interventions will be judged by the talker to be helpful, and not judged to be disruptive. The worksheet to be used to record the evidence for the talker is given in Worksheet 5-4; some example data are given in Figure 5-3. Add a . to the chart whenever a monitoring statement is said. 5.2.3

Activity 5.2: (time 35 minutes)

Allow the same timing as used in Activity 5.1, in Section 5.1.2. The physical arrangement for the ten- minute talk period is illustrated in Figure 5-4. The Tasks may be selected from Tasks 5.1.A to E or from Task 5.2.A including options from Appendix F, Sections 5.1 and 5.2. In pairs, one be a talker, the other be a listener. The talker plays the role for 10 minutes and “solves” problems during all the allotted time. Do not change roles. Use Worksheet 5-1 for reflections and Worksheet 5-5 for feedback about listening. The Worksheet 5-4 will be validated by the listener and given to the talker as evidence.

5.2 Strategies

Figure 5-3

(top) Some example data on the Worksheet.

(bottom) The physical arrangement: the listener is shown on the left; the talker, on the right.

Figure 5-4

177

178

5 Polishing Your Skills: Problem-Solving Process

Worksheet 5-4:

Record of the talker’s strategy with . for monitoring statements.

Talker __________________ Case _________ Listener ____________________ Stage Engage: “I want to and I can!” Define-the-stated problem: Sort the given problem statement Explore the problem to discover what the problem really is Plan Do it Look back: elaborate, check

0

2

4

6

8

10

12

Time, minutes

Task 5.2.A: Terry Sleuth in the Poly Room Terry Sleuth ventured into the polymerizer room on the way to the R&D lab for an appointment with Bill Wright. Seeing Terry, Charlie, the poly-room engineer, called “Hi Terry, please come over here and help us sort out this mess. Look at these reactors. We’re continually losing quality product in reactors two and three but not on reactor one.” “Tell me more,” encouraged Terry. “The problem is that just near the end of the run, the motors driving the mixers overload and cut out. They stop! Then the whole batch is ruined because the heat transfer is insufficient,” explained Charlie. “Does it happen only on reactors two and three and not on reactor one?” asked Terry. “Yes. Reactors two and three have calandria coolers and marine propeller mixers. Reactor one has an internal coil cooler and a turbine mixer. The mixing flow patterns are chosen specifically for the heat-exchanger configuration: the propeller moves the liquid up through the tubes with the coolant on the shell side of the internal calandria; the turbine shoots the liquid out onto the tubes with the coolant inside the tubes. Perhaps the calandria tubes are plugged with polymer,” continues Charlie. Terry thought for a moment and then asked “Are both cooling systems about the same surface area and serviced by the same cooling water?” “Yes.”

5.2 Strategies

“Has this problem ever occurred before?” “This is the first time we have processed this product on any of these reactors; we have processed other products for years on all three reactors and with great success,” beamed Charlie. Terry smiled. Terry had all the information needed to identify the cause. What did Terry say? Terry Sleuth and the Case of the Delinquent Decanter Ring ... The incessant ringing of the telephone disrupted the steady hum of the engineering office. Betty answered, looked perplexed, looked across at Terry Sleuth’s desk and knit her brow as she carried on an animated conversation. Terry could almost guess from Betty’s reactions that the call was about the hexane-water decanter on the soybean-miscella still. The Ministry of Environment had been after us for the last week to get the concentration of hexane in the wastewater from the decanter down to acceptable levels. Unfortunately, the decanter wasn’t working well. Betty hung up, donned her hard hat and headed toward Terry’s desk. “It’s the decanter again, isn’t it!” said Terry. “You’ve said it”, replied Betty disheartenly. “Before we head out to see the beast, let’s review what we know”, suggested Terry encouragingly. “OK”, Betty said and went on “the decanter is a vertical cylinder with a conical bottom; the feed enters about midpoint with exit lines at the top for the hexane and at the bottom for water. The feed is a mixture of about 60% hexane and 40% hot water. The foam or droplet layer fills the center band of the decanter; the drops coalesce to form pure layers of hexane (that rises to the top) and a pure layer of water. There is a cover on the decanter that has a vent. The last time we talked we realized that some hexane dissolvers in the water and some droplets of hexane may go out with the water because the drops haven’t coalesced. How’s that?” Terry replied, warmly, “Very good. However, we were unsure as to whether the feed is hexane drops in water or water drops in hexane. Since there is more hexane, it is probable that the feed is hexane containing droplets of water. So it is the water drops that are coalescing.” “Yes, but I don’t see that that makes any difference whether it is hexane drops coalescing or water drops coalescing. What I have done, since we last talked, is a lot of research about decanters. Mizrahi and Barnea report in a refereed article that for every 10 C increase in temperature, the coalescence time will be faster by a factor of two. I think I’ll rig up a heater on the feed line so that the incoming temperature of the feed is 70 C instead of 60 C. That should fix this baby!” gloated Betty. Terry paused, checked the Handbook of Chemistry and Physics and said, “Don’t be too sure.” What did Terry look up and why did Terry cast doubt on Betty’s idea? Task 5.2-B:

5.2.4

Feedback, Self-assessment

Self-assessment is based on written evidence, not on intuitive feelings. Throughout this activity evidence has been gathered. The evidence includes:

179

180

5 Polishing Your Skills: Problem-Solving Process . . . .

.

the problem statement, with all your underlining and notations, your paper you worked on, the time-stage-monitoring evidence of Figure 5-4, the reflections, similar to Worksheet 5-1 but with prompts related to this activity, feedback about your role as listener, Worksheet 5-5. You might elect to provide additional feedback to the talker from Worksheet 5-2 about the process used.

Evidence for listening: Feedback to listener & the listener will encourage verbalization, an emphasis on accuracy, active thinking and encourage the problem solver to move the marker correctly on the strategy board. Your interventions will be judged by the problem solver to be helpful, and not judged to be disruptive.

Worksheet 5-5:

Activity 1: Talker ____________

Case ___________

encourage verbalization: encourage emphasis on accuracy: encourage active thinking interventions:

not needed not needed not needed not needed

Listener _____________

interruptive interruptive interruptive interruptive

OK OK OK OK

really helped really helped really helped really helped

Comments: ___________________________________________________________________ ___________________________________________________________________

5.3

Exploring the “Context”: what is the Real Problem?

During the Explore stage, we may wish to place the problem in a larger context. A very useful technique is Basadur’s Why? Why? Why? technique (Basadur, 1995). In this technique, we start with the initial “problem”, ask Why? and redefine the problem in a broader context. The process is repeated. After we have completed several levels of context, we then can address “which level of generality is best, given the current constraints and contexts.” This approach was used very successfully by Pierre in the Case ’4, the case of the Platformer Fires, Section 4.3. Consider first an example and then an activity. 5.3.1

Example

Here we illustrate the application of the approach for Case ’7, the Case of the reluctant crystallizer, problem 1.5 in Chapter 1 and considered in Frank’s approach in Section 4-5. Beginning symptom: “the liquid level in the crystallizer is dropping at a fantastic rate.”

5.3 Exploring the “Context”: what is the Real Problem?

Rephrase as Why do I want to stop the liquid level in the crystallizer from dropping at a fantastic rate. OK. Perhaps your answer is “so that I can produce quality crystals from the crystallizer.” OK. Now ask Why? Now your answer might be “So that I have quality crystals to sell to my customers.” This process continues. It is convenient to summarize this are follows: Why? so that

have Happiness and Bliss › have me a rich and productive life; Why? so that › Why? so that have challenging and exciting employment; › Why? so that keep the company profitable; › keep sales healthy, Why? so that › have quality crystals to sell to my customers; Why? so that › produce quality crystals from the crystallizer Why? so that › Start: ﬁ stop the liquid level in the crystallizer from dropping at a fantastic rate. Why?

In this case, perhaps the most important problem to address is “to have quality crystals to sell to my customers.” That being the case, Frank should have addressed his efforts toward obtaining the quality and quantify of crystals, even if it means buying them from a competitor. Thus, this technique helps us to see the problem in a bigger context and prompts us to ask “what is the real problem?”. 5.3.2

Activity 5-3

For Case ’9 The bleaching plant, do a Why? Why? Why? analysis. The bleaching plant. Our process makes margarine. One step in the process is to remove the odors and color bodies from the edible oils used to make margarine. This removal is done in the bleacher illustrated in Figure 5-5. The light volatiles are removed by subjecting the oil to high vacuum. The color is removed by adsorbing the color species on powdered adsorbent that is subsequently filtered from the oil. The bleacher is shown in the diagram. The procedure, as set out in the startup manual, is to charge the vessel with edible oil, isolate the bleaching vessel (by turning the appropriate valves), draw a vacuum by means of the steam ejector system until the absolute pressure is 20 kPa abs on the vessel, open the valve connecting the bleach vessel to the hopper filled with the adsorbing powder, Fuller’s earth, and suck (or pneumatically convey) the powder into the bleach vessel. The physical arrangement, including the approximate elevation, is shown in the diagram. This is a batch process. Case ’9:

181

182

5 Polishing Your Skills: Problem-Solving Process

Steam

Light

View PI 1

Oil feed

Hopper of Fuller's earth

Figure 5-5

The bleacher for Case ’9.

Cooling coils

To filter press to remove Fuller's earth

The gauge on the vessel read 13 kPa abs, but when we opened the valve to convey the powder into the bleacher, nothing happened. That is, we expected to see, through the view port, the powder dumping into the liquid in the bleacher. This is the first time this plant has ever started up. The product has already been sold and because of previous delays in startup we are now losing $15 000 per day. ________________________________________________________________ Why? › ________________________________________________________________ Why? › ________________________________________________________________ Why? › ________________________________________________________________ Why? › ________________________________________________________________ Why? › Start ﬁ _____________________________________________________________ Reflect on this approach. Was it easy to do? What insights did you gather? What might you address as the real problem? When might you use this approach? ______________________________________________________________________ ______________________________________________________________________ ______________________________________________________________________ ______________________________________________________________________

5.4 Creativity

5.4

Creativity

Whether solving exercises or problems we need creativity to generate different points of view (for the creation of a rich internal representation), to create hypotheses about change and to create hypotheses based on the basics. 5.4.1

Some Target Skills

Skilled creative thinkers: 1. 2. 3. 4. 5. 6. 7. 8.

Defer judgment; Are succinct; Can list 50 ideas in 5 minutes; Create a risk-free environment; Encourage free and forced association of ideas; Can piggy back on previous ideas; Use triggers, such as those listed in Table 5-1, to maintain the flow of ideas; Aren’t discouraged. In the last two minutes of a five-minute brainstorming session, over 85% of the ideas are not practical. But, they spend time to identify the treasures among the 15%; 9. Are positive; 10. Manage stress well. Manage any negative self-talk; 11. Use impractical and ridiculous ideas as “stepping stones” to innovative, practical options. After they have generated a large list of ideas, skilled creative thinkers then list measurable criteria and select at least five technically feasible hypotheses. Table 5-1

Checklist of triggers for brainstorming.

Name of trigger to change point of view

Elaboration of how it is to be used

To be used for objects

Function

Objects for engineers are What is the function of this object? How else can “products” and hardware. we achieve this function? What function cannot it do?

Physical uses

What are its physical properties and characteristics? What are they not? How else might we obtain or use these physical properties?

“How to improve ion-exchange resin”

Comments

Probably the most useful perspective; use first. The negative view is often extremely illuminating.

183

184

5 Polishing Your Skills: Problem-Solving Process Table 5-1

Continued.

Name of trigger to change point of view

Situations “I need to design an ion-exchange unit” “The reformer is not functioning; fix it.”

Elaboration of how it is to be used

Comments

Chemical uses

What are its chemical Easy for us to use; use properties? How can these second. be exploited? What are they not?

Personal uses

What personal uses can we make of this object?

Interpersonal uses

What interpersonal uses can be made?

Aesthetic uses

How can we create pictures? music? artistic creations with the object? sculpt? weave?

Mathematical or symbolic properties

What are the mathematical Only applicable for certain objects or ideas. or symbolic uses? What can they not be used for? What are they often confused with?

Checklist

Effective for objects and Of the many checklists some situations. Try each published SCAMPER is viewpoint. Easy to do. probably the most effective: S: substitute who? What? Other processes? Other places? C: combine purposes? Ideas? Appeals? Uses? A: adapt what else is like this, new ways? M: modify, maximize, minimize? P: put to other uses, other locations E: eliminate, R: reverse, rearrange

These three give a unique perspective that is often overlooked.

5.4 Creativity Table 5-1

Continued.

Name of trigger to change point of view

Elaboration of how it is to be used

Comments

Wildest fantasy

Think of the craziest idea or use.

Need to establish self-confidence and group confidence before really outlandish ideas are presented. Build confidence in the merit of this approach by later bridging to unique ideas. Brings laughter and tension relief.

How nature does it

Identify the situation and then think of how nature fulfills this function. Bridge to engineering reality.

Great potential; depends on the situation. Successful in designing new bridges; new bog machines.

What if? In the extremes

Extension of problemExtrapolate from the unfamiliar to a simplified solving skill in defining simple problems. Relatively version easy to apply.

Boundary exploration

Identify the constraints and remove them

Functional analogy

How else is the function achieved?

Appearance analogy

How else might we get the appearance of this situation?

Morphology

Break the problem situation into a series of parts. For each part list 10+ options. Then, systematically combine one option from each part and ask why not?

More mechanical; easy to computerize; most of the work is in setting up the parts and the options. Fun and surprising to see some of the results.

Symbolic replacement

Replace the original problem by an interesting idea generated in the session; refocus.

Occurs naturally in many brainstorming sessions. Add this trigger explicitly if needed.

Juxtaposition

Bring in three completely random words and bridge to current situation. Example “refrigerator, light switch, clock.”

Very effective. Don’t fake it by using words “you think will help”. Use random words.

Much easier for some to do than for others.

185

186

5 Polishing Your Skills: Problem-Solving Process Table 5-1

Continued.

Name of trigger to change point of view

Elaboration of how it is to be used

Comments

Personal analogy Imagine yourself-as part of the situation. Describe your feelings and what you are experiencing. Be imaginative.

Tricky. Works for some people; not for others. Effective with fluid-dynamic problems.

Reversal

Do the reverse.

More challenging than one thinks. Focus systematically on the reverse of different elements of the situation at a time.

Book title

Interesting. Worth a try for Create a title for a “best1 minute. selling” novel. The title should sum up the current situation. “Will exchanging bring happiness to Mary Lou?”

Letter, word, sentence.

Very effective for most of Focus on three different levels of detail correspond- our problems. Consider a 5-minute brainstorm at ing to the big view (a sentence), an intermediate each level. view (a word in the sentence), and a letter (in a word). In ion exchange this might be “the ions, the resin, the packed bed, the separation”.

Visual image

Look at three or four famous paintings. Describe aloud what you see and make connections.

At first glance this sounds too bizarre. However, for some this is one of the more effective ways of seeing the situation from a new perspective. Difficult to do out on the plant.

5.4.2

Example: Case ’10: To dry or not to dry! (based on Krishnaswamy and Parker, 1984)

Our product is dry crystals of calcium nitrate. The crystals are precipitated in a continuous crystallizer. The exit flow from the crystallizer – consisting of a slurry of crystals and mother liquor – goes to a fixed, parabolic screen (DSM) to separate the large, product crystals from the undersized crystals that are recycled. The undersized are pumped con-

5.4 Creativity

tinuously through a hydrocyclone and returned to the continuous crystallizer. As screening progresses the screen gradually blinds because the finer material that gets stuck in the screen cloth causes too many fines to be carried over with the large product crystals. Hence the screen is operated batchwise. The feed to the screen is stopped; the screen is washed, and then the screen is put back in service. The large crystals from the screen go to a scraper-discharge centrifuge. The operation of the centrifuge is continuous but the moist crystals from the centrifuge are discharged batchwise because the centrifuge goes through a filter-wash-peel/discharge cycle. The exit from the centrifuge goes directly to a continuous, steam-tube rotary dryer via a chute and a conveyor screw. In the past, the crystal washing in the centrifuge was not very efficient. They modified the cycle to provide more wash water in the washing cycle of the centrifuge. The system is shown in Figure 5-6. “The crystal product from the rotary dryer has a moisture content of 4.5% whereas the design value is 1.5%. Cake seems to be building up in the dryer feed chute, in the feed screw and on the steam tubes at the feed end of the rotary dryer.” Fix the problem. Get the dryer working again so that the exit crystal moisture content is always below 1.5%.

Feed

Recycle fines Overflow supernatant Parabolic screen

cyclone

Centrifuge

Steam Liquid

Figure 5-6

Continuous steam-tube dryer

dry solids

The system used to dry the crystals; Case ’10.

Example results of brainstorming: Too much water in wash cycle for dryer to handle, cake buildup on steam tubes decreases heat transfer, cake dries up too early causing buildup while the rest of the cake is very wet, not enough steam supplied, wrong centrifuge, wash water carryover from the centrifuge, steam leak into dryer, it’s raining, cycle wrong in centrifuge, cycle from screen not coordinated with cycle from centrifuge, crystals too wet from the screen, washing from the screen contaminates the feed to the centrifuge, feed

187

188

5 Polishing Your Skills: Problem-Solving Process

from the screen too wet, fines carryover to the centrifuge causing blinding in centrifuge, peel cycle too short, peel cycle too long, filter cycle too short, filter cycle too long, cycles now out of synchronization between the screening and filtering, feeder from screen to centrifuge broken, feed screw to dryer rotating too fast, rotating too slow, the rotational speed of the dryer has increased, the residence time in the dryer has decreased, the steam pressure has changed, the quality of steam from the boiler house has changed, the condensate trapping on the dryer is backing condensate up and reducing the heating area, air is blanketing the heat-transfer area in dryer because air was not bled off before restarting dryer, pitch of the dryer has changed, operating procedure has changed, pump from the screen bottoms malfunctioning and water spilling back into feed to centrifuge, cycle on the parabolic screen has been increased, more fines into peeler centrifuge causing poor filter cycle. Try trigger “craziest”: operators not reading instruction correctly, sandwich dropped in line and plugging flow of crystals, wash down water used to clean the floors is splashing onto feed screw, the dryer is turning backwards, the peeler cycle is reversed to peel-wash-filter, the feed to the centrifuge is going directly to the dryer, feed from the crystallizer is going directly to the dryer, for this phase of the moon the vampires are out (is that crazy enough?). Try trigger “reversal”: focus on the product from the dryer is too dry: possible reasons are: long residence time, higher steam pressure/temperature, excess drying area, feed to the dryer from the centrifuge is dryer than expected, centrifuge rpm faster than usual, washing of the cake in the centrifuge is done with new solvent that evaporates faster, feed to the centrifuge from the screen is dryer than usual. OK, for the current problem take the opposite of all these. Try trigger “juxtaposition” using random words elastic bands, stapler and coffee cup. The process I will use is to take characteristics of this idea and find links to possible feasible ideas: I’ll show that as arrows ﬁ and keep going until I get an idea that I think is feasible. Elastic bands: stretches ﬁ returns to original shape ﬁ screen becomes compressed and later pops back Elastic bands ﬁ made of rubber ﬁ gaskets leaking? Elastic bands ﬁ long and narrow ﬁ crystal size changes Stapler: batchwise operation ﬁ dryer should be run batchwise too Stapler: runs out of staples, need to refill periodically ﬁ centrifuge operates but has run out of feed Coffee cup: holds liquid ﬁ liquid condenses near feed end of dryer and drops onto feed; recycle of wet damp gas in dryer. Coffee cup: handle ﬁ holds ﬁ liquid retained in crystals by interstitial forces Coffee cup: handles hot liquids ﬁ wash water is too hot; hot condensate clogs dryer tube Coffee cup: status symbol with logos embossed on it ﬁ words ﬁ instructions wrong. Coffee cup with metal: can’t put in the microwave ﬁ contamination from rust, plugging lines Coffee: brown liquid ﬁ dirty wash water Coffee: tasty ﬁ micro-organisms growing in centrifuge, in wash water, in crystallizer

5.4 Creativity

Coffee: justification for breaks ﬁ operators don’t pay attention when they should; breaks in the cycles between the screen and the centrifuge. Coffee: social event ﬁ water clings to crystal, not water anymore Select the craziest as a “stepping stone” or “piggy back” to feasible idea.”: Crazy idea: “for this phase of the moon the vampires are out” The process I will use is to take characteristics of this idea and find links to possible feasible ideas: I’ll show that as arrows ﬁ and keep going until I get an idea that I think is feasible. Phase ﬁ sinusoidal motion ﬁ cycles ﬁ periodicity ﬁ the cycle of the screen interferes with the cycle of the centrifuge ﬁ periodic pulses of wet stuff to dryer. Moon ﬁ night ﬁ cool at night ﬁ is there a temperature effect on the feed? the feed to the dryer is too cold; is the steam too cold? is the wash water in the centrifuge too cold? Moon ﬁ pale light in the dark ﬁ hard to see ﬁ difficult to see the moisture content in the feed because it changes so fast ﬁ things are happening on this plant that are hard to see ﬁ slow cycles and feedrate down to each section of the plant in turn and try to “see better” Moon ﬁ changes shape from crescent to circle ﬁ the crystals change shape so filtering is different Moon ﬁ changes shape from crescent to circle ﬁ crescent ﬁ curve ﬁ the curve of the parabolic screen is too steep ﬁ decrease in capacity Moon ﬁ changes shape from crescent to circle ﬁ changes ﬁ same ideas as listed above for Phase Moon ﬁ land on the moon ﬁ long distance away ﬁ something a long distance away from the dryer is causing problems ﬁ utilities? steam? wash water? Moon ﬁ land on the moon ﬁ long distance away ﬁ something a long distance away from the dryer is causing problems ﬁ feed to crystallizer? pregnant liquor? crystallizer? screen? hydrocyclone? pump to hydrocyclone? screw feeder to dryer? centrifuge? Moon ﬁ land on the moon ﬁ long distance away ﬁ something a long distance away from the dryer is causing problems ﬁ head office giving new policies and stressing out employees ﬁ spouses stressing out the operators? ﬁ events stressing out the operators ﬁ operators making mistakes Moon ﬁ land on the moon ﬁ long distance away ﬁ something a long distance away from the dryer is causing problems ﬁ vendors supplied faulty equipment? lousy wash system? Moon ﬁ land on the moon ﬁ different environment on the moon, lack of oxygen ﬁ oxygen in wash water affecting crystals? contaminant in wash water causing poor washing? is it the washing cycle? of the drying cycle in the centrifuge? Vampire ﬁ person ﬁ operators error? Vampire ﬁ person ﬁ operators ﬁ operating instruction error? Vampire ﬁ teeth ﬁ sharp objects breaking up crystal size ﬁ slower filtering and slower drainage rate in centrifuge. Vampire ﬁ drinks blood ﬁ blood ﬁ operator injured? water is not water, it’s contaminated with low vapor pressure.

189

190

5 Polishing Your Skills: Problem-Solving Process

Comment:

This session was characterized by:

mix of symptoms and roots causes; lots of duplication; most feasible ideas in the first 32 ideas; a lot of impractical ideas, especially later in the session but there were some new feasible ideas amongst these.

. . . .

Total ideas: 98 with 32 from the initial burst, trigger “craziest” = 8; trigger “reversal” = 8; juxtaposition= 20; stepping stone = 30. The stepping stone seemed to work well for me on this problem. From this are selected the following possible hypotheses: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

not enough steam, wash water carryover from the centrifuge, cycle from screen not coordinated with cycle from centrifuge, feed crystals too wet from the screen, rotational speed to the dryer has increased, more fines into peeler centrifuge causing filter cycle to be too long; fines carryover to the centrifuge causing blinding in centrifuge, centrifuge rpm faster than usual, crystal size change; crystals change shape so filtering is different, centrifuge operates but has run out of feed, dryer feed too cold, vendor supplied faulty equipment. – slow down the cycles and feedrate down to each section of the plant, in turn, and try to see better what is happening. – “slower filtering and slower drainage rate in centrifuge” symptom! use as potential brainstorming idea. – “periodic pulses of wet stuff to dryer” symptom! not a root cause but might be profitable as symptom.

5.4.3

Activity 5-4

For Case ’9, the bleacher (described in Section 5.3) brainstorm 50 possible causes and write these down. Suggested time frame: initial write 6 min. . . . .

.

Say to yourself-“I can do this.” reread the problem statement: 2 min. Try trigger “wildest fantasy” 1 minute Try trigger “what if? in the extremes. 1 minute Try trigger “juxtaposition” using the words pencil, hat, and perfume. 1 minute each word. Try trigger “reversal” 1 minute.

Stretch.

5.5 Self-Assessment

Look over your list and select the “craziest one”. Use this idea as a “stepping stone” to a technically feasible one. By this we mean, take the properties of the crazy idea and use these in a feasible idea. Reflect on what happened. Complete the feedback form given in Figure 5-7. 5.4.4

Feedback, Self-Assessment

Self-assessment is based on written evidence, not on intuitive feelings. Throughout the brainstorming evidence has been gathered. The evidence includes the number of ideas produced and the succinctness of the ideas. In addition we can write reflections, using a form similar to Worksheet 5-1. Figure 5-7 gives a convenient summary form specific for brainstorming.

Figure 5-7

Feedback form for brainstorming.

5.5

Self-Assessment

Your skills and confidence as a trouble shooter will improve the most if you create a formal approach to self-assess your current skill, set goals and gather evidence about your progress to achieving the goals. Skilled self-assessment develops objective awareness of how you perform a task and develops self-confidence. In this section

191

192

5 Polishing Your Skills: Problem-Solving Process

we define the target skills for self-assessment, in Section 5.5.1, and then propose activities for growth in Section 5.5.2. 5.5.1

Some Target Skills

Skilled assessors realize that assessment is: 1. 2. 3. 4. 5.

about performance; it is not about personal worth. based on evidence; it is not based on wishful thinking or gut feelings. essential for growth. not possible without published unambiguous goals and measurable criteria. based on a wide variety of different types and forms of evidence.

5.5.2

Activity for Growth in Self-Assessment

Attitude and skill – these are the key elements to work on. Assessment involves a change in attitude. Many think of exams, performance reviews and evaluations as being stressful and something to be avoided. We need to realize that self-assessment helps develop growth and is one of the most positive activities we could do to improve trouble-shooting skills. Assessment is about performance! Assessment is based on evidence and is a judgment done in the context of published goals and with measurable criteria. The three skills needed relate to creating goals, measurable criteria and designing forms of evidence a)

Ability to identify and create observable and unambiguous goals. Examples have been given in each section. In general unambiguous means that we can observe the performance. Such words as “know”, “create” are unacceptable because we cannot observe someone demonstrating that they “know this”. Words like “list”, “write out” are an improvement. b) Ability to identify and create measurable criteria related to the goals. This task is difficult, boring and tedious – but necessary. For example, in Section 5.4 the goal was to list ideas. To make that measurable we need a number and a time: 50 ideas in 5 minutes. c) Ability to write out, gather and evaluate evidence as it relates to the goals. As demonstrated previously in this book, evidence in the form of reflections, the worksheets and feedback forms are all useful. Usually, once the goals and criteria are created, it is relatively easy to create the forms of evidence. Activity 5-5 Self-assessment Based on the self-assessment in Section 1.3 create goals for growth, measurable criteria and pertinent forms of evidence. Begin gathering a collection of evidence and systematically go over the evidence and self-assess your progress.

5.5 Self-Assessment

5.5.3

Feedback About Assessment

Figure 5-8 provides a feedback form related to the assessment process.

Figure 5-8

Feedback about the assessment process.

193

194

5 Polishing Your Skills: Problem-Solving Process

5.6

Summary and Self-Rating

In this chapter we identified five skills related to problem solving. These were awareness or the ability to describe your problem-solving processes; strategies or the ability to see patterns in the process; exploring the context using the Why? Why? Why? process; being creative and being skilled in self-assessment. For each target skills were listed, some examples were provided, activities to develop the skills were described and forms of evidence were given. Reflect on what you have experienced in this chapter: ______________________________________________________________________ ______________________________________________________________________ ______________________________________________________________________ ______________________________________________________________________ Rate

Awareness: can describe, focus on accuracy, am active Strategies: see patterns, monitor Explore the context using Why? Why? Why? Am creative: 50 in 5 min, use triggers, can step from crazy ideas to feasible ones Self-assess: attitude about performance; needed for growth, evidence-based skill: create observable goals; create measurable criteria; identify pertinent evidence; valid judgment in this context;

I think I am skilled

I have some evidence and have some skill

I am confident. I know the goals the criteria and have a variety of evidence

Not sure this is for me

& & & & &

& & & & &

& & & & &

& & & & &

&

&

&

&

& & & &

& & & &

& & & &

& & & &

195

6

Polishing Your Skills: Gathering Data and the Critical-Thinking Process Critical-thinking skills are needed by trouble shooters. These skills include selecting and designing tests, checking for consistency, classifying sets of ideas or data, recognizing patterns and reasoning and drawing valid conclusions. Here are some examples of where some of these critical-thinking skills are used in the task of trouble shooting. We: .

.

.

.

.

.

.

.

See/receive information that suggests something it wrong: compare conditions with unsafe conditions and decide from among a) emergency shutdown, b) change to “safe-park” or c) continue current conditions. Select the symptoms. Separate fact from opinion. The thinking skills needed are consistency and classification. Gather more information about the current situation and compare and contrast this with the “expected” performance to obtain a rich mental image of the situation. The thinking skill needed is how to compare and identify differences. See if there are any patterns in the data. The thinking skill needed is pattern recognition. Brainstorm possible faults or causes that could cause the symptoms. Alter those ideas that are “symptoms” to be “root causes”. Classify the list and apply criteria to prioritize the list of hypotheses/causes that are consistent with practical experience. The thinking skills needed are classification and consistency. Select hypotheses/faults that are consistent with the symptoms. The thinking skills are using cause–effect information and classification. Create tests; the tests must be pertinent to the hypothesis, and not suffer from confirmation bias. The thinking skill needed is how to select valid diagnostic actions. See if there are any patterns or trends in the evidence collected. The thinking skill needed is pattern recognition. Check that the results of the tests confirm and disprove the hypothesis. The thinking skill needed is reasoning.

In this chapter we consider first, in Section 6.1, how to select valid diagnostic actions. In Sections 6.2 to 6.5, we look at the critical-thinking skills of checking for Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

196

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

consistency, classification, seeing patterns and reasoning. Each is considered in turn and illustrated through examples.

6.1

Thinking Skills: How to Select Valid Diagnostic Actions

First, we offer criteria for selecting diagnostic actions; then we provide a structured list of possible actions. Then we consider how to perform and interpret the actions. In Section 6.1.3 we explore how to become sensitive to personal preferences, style and biases that might interfere with the selection and use of diagnostic actions. Consider each in turn 6.1.1

How to Select a Diagnostic Action

In selecting an action to take, the following criteria are useful: . .

. . . . .

Safety first! Keep it simple. Pertinent, easy-to-gather information should be gathered first. The results should provide the accuracy needed. Safety and time are critical. Select actions that will prove or disprove hypotheses. There is an expense associated with any action or lack of action. Stopping production to “inspect” or “change equipment” is usually very costly.

6.1.2

Select from among a Range of Diagnostic Actions

The actions taken differ widely from immediate emergency response, to gathering information to better understand the problem, to testing hypotheses, to immediate correction of the suspected cause. In general, the actions selected depend on the type of TS problem, the TS strategy you elect to follow and your personal style. A structured list of options includes: a) is it an emergency? b) put the situation in the context of recent events, c) understand what should be happening, d) check what is happening now, e) test hypotheses and, perhaps, f) take immediate corrective action. Each is considered in turn. a. Safety first! Is it an emergency? What’s the general background information about this situation? .

Invoke emergency shutdown? Although trouble shooters should know the hazards related to the process, information about hazard, safety, MSDS

6.1 Thinking Skills: How to Select Valid Diagnostic Actions

. . .

sheets can be obtained from Chapter 3, Section 3.12, the MSDS sheets on file or from two web sites for MSDS data http://www.msdsxchange.com or http://www.msdssearch.net. Is “safe-park” a required option? Tabulate IS and IS NOT data. Note details about the weather and offsite services: air temperature and humidity, rain-snow, ice-fog? bay water temperature, river temperature, city water properties/temperature, tower water, steam and air plant; wastewater treatment plant.

Comment: Start here and put the process into perspective. b. Should I be using a TS strategy best suited for “change”? What has happened recently that might affect the process?

What and when did changes occur: .

. . .

changes made to operating instructions, to feed, to specifications, to flowrates, work done on the plant, what was done during the maintenance turnaround, what was done in routine maintenance.

Comment: Easy to do. You may know this already but check out the details of exactly what was done. This provides the basis for selecting your TS strategy and subsequent diagnostic actions. c. What should be happening: expected, “usual” performance?

This information is usually obtained in the office from files and records, from simple calculations and from simple, astute What if? exploration. Files and records: consult key files and records and predictions of performance: .

.

. . . .

design files and simulation: design basis, fouling factors, assumptions made, specifications and predicted conditions throughout the plant. vendor files for the equipment: performance expected, sometimes troubleshooting information* (* often available in both the engineering office and the control room). operating procedures*. commissioning data. recent P&ID. Often not available, or if available, not up-to-date. recent tests and internal reports.

Data from handbooks and texts in your office. Trouble-shooting files: data relating symptom to root cause with the probabilities or likelihood, similar to that given in Chapter 3.

197

198

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

Calculations or estimations that can be done before special tests are done. .

. .

.

.

calculated/estimated pressure profiles: these can be done from simulations or from order-of-magnitude estimates. Note the pressure difference across barriers to determine the direction of flow if a leak occurs. mass balance: predicted by simulation or estimated. energy balance: heat in= heat out across exchangers, furnaces, condensers, reboilers. Steam usage with 5 kg organic evaporated by1 kg steam, flame temperature,% excess air. thermodynamics: equilibrium conversion, vapor-liquid equilibrium, energy changes for flow across turbines, ejectors and valves. rate: use the temperature difference in the reboiler to estimate whether boiling is in the nucleate, film or transition regime.

Equipment performance calculations: check that the sizing and general operability conditions are met. Will the installed equipment do the job expected? For process control, check the location and type of sensor, type of control used. Exploration from simple “What if?” questions. Sometimes we can quickly eliminate extraneous tests by asking such questions as “What if you could actually look at the flow patterns inside the vessel?” “What if you could look at the catalyst surface?” “What if there was no insulation? “What if there was infinite insulation?” Comment: This should be a valuable, third level of data acquisition. However, the usefulness of these actions depends on the quality of the documentation available (often it is poor), and of your files, on your skill with order-of-magnitude estimates; and on the information available in the case-problem statement. d. What is currently happening?

We start with information in the control room and from the operators. In the control room, here are some options: .

. . .

read the displayed information and check past records of same variables. Look for trends. Look for interconnects (two or more events that seem to be happening together: temperature increase and conversion decrease). check the operating procedures*. scan the vendor files including performance data and trouble shooting*. talk to the operators about changes, about this shift versus previous shifts, about their hypotheses, about the facts.

Comment: This is a vital source. Meet the operators; find out the details; read the instruments displayed. This is a must visit before venturing out onto the plant. e. What is currently happening and gathering information to test hypotheses?

We learn more about what is currently happening i) by putting the process in context; ii) by using our senses to gather information from out on the plant, iii) by gathering data for calculations, iv) by checking out the sensors and controllers, v) from

6.1 Thinking Skills: How to Select Valid Diagnostic Actions

quick checks with specialists, vi) from more ambitious tests and vii) shut-down activities. i)

Put the process in context. Check with the operators of the utilities and of upstream and downstream plants. ii) On the plant visit. Out on the plant we use our senses, do simple tests, check for consistency, and look for trends.

Use your senses. Read the instruments, listen, smell, look. Note the position of the valve stem, look for steam leaks, look at the discharge from steam traps, read the motor amps, check the condition of the insulation, look for rust around a flange, leaks around a gland, look at flows (if they are visible). Hot smells? Listen for noise: indicating flow, cavitation or liquid level in vessels. Use your intuitive feelings. Simple on-site tests: Shift to a more quantitative perspective: estimate/measure: . . . .

. . . . .

the temperature: use the glove test or a laser or surface pyrometer. the humidity or dew point. the response of sensors: check the response to a change in set point. the signals to controllers/valves: inlet pressure/signal; pressure/signal to the valve. diameter of insulation/ of pipes. the liquid flows in accessible drains. the exit pressure for zero flow for centrifugal pumps. if the bypass valve is open or shut. condition of the valves: turn and seal.

Comment: Your senses and simple on-site tests are easy to use and often provide excellent, key information. Check for consistency: do multiple sensors agree? are the lab data consistent with the results from on-stream analyzers? temperature–pressure–composition agreement? phase-rule agreement? mass-balance agreement? make-sense agreement (fluids flow from high to low pressure? thermal energy flows from high temperature to low?) physical-thermal data agree with measurements (pressure enthalpy data for refrigerant)? Check for trends: have the temperatures, pressures or yields been changing gradually over the weeks? Is there a trend every 5 minutes? every 15 minutes? every hour? What is the frequency and amplitude of cycling? Check that the P&ID agrees with the actual process configuration. iii) Gather data for calculations. Fundamentals underpin the process. Check a mass and energy balance. Gather data to estimate performance of equipment. iv) Sensors and controllers. Put the controller on manual. Is there evidence that the sensors are working and are accurate? Use temporary instruments to check measurements. Request specialists to calibrate the sensors or tune the controllers.

199

200

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

v)

Quick checks with specialists. Consultants: call vendors, original designer, consultants, licensee of process and/or suppliers of raw materials, adsorbents, catalysts.

Comment: the operative word is “quick”. Although it’s better to talk by phone than e-mail, the key people may not be available to answer your call. vi) More ambitious tests. Samples of gas, liquids or solids for analyses: obtain the samples: Are the samples representative? Valid? Is there a time dependency? Are all the samples correctly labelled? Can the sampling be done safely? Type of analysis: often chemical or particle-size analysis. Specialists: Should a velocity profile be measured? gamma scan? tracer study? Perhaps a well-designed set of experiments should be run and the results analyzed statistically. Comment: although the process can continue to operate, these tests take time and may be expensive. vii) Shut-down-type activities. Open and inspect: shut the process down, safely isolate the piece of equipment, and clear of the process fluids and vapors. Know what you are looking for, but be prepared for surprises. Comment: this requires that the process be stopped. This is expensive. Try to leave this action as the last resort. Much can be learned from the previous activities. f. Possible corrective action

Some trouble shooters, especially those who prefer action to patiently gathering evidence (dominant J behavior as described in Section 6.1.3c), take “corrective” action prematurely. Others might astutely have identified the cause early; others, are lucky. In general, be cautious in taking corrective action before simple tests of hypotheses can be made. Usually “Take corrective action” requires the expensive shutting down of the process and making alterations. 6.1.3

More on Gathering and Interpreting Data

Here we consider the guidelines for designing experiments to test hypotheses, and list the resources needed for gathering data from simple tests. We also explore the implications of your personal style in planning and selecting tests. 6.1.3.1 Guidelines for Selecting and Designing Experiments to Test Hypotheses The tests should be simple, inexpensive and should provide positive and negative validations of a hypothesis. Usually we try to: .

.

avoid the temptation to “open the equipment and see”. Usually some very simple tests can keep the process going and provide the insight needed to identify the fault. check one hypothesis/variable at a time.

6.1 Thinking Skills: How to Select Valid Diagnostic Actions . .

isolate variables. test the “most likely” fault based on probabilities of past failures. Table 6-1, for example, lists some data related to instrument failure, relative to “equipment failure”.

Failures per annum for instruments and equipment (from Woods, Process Design and Engineering Practice, originally published by Prentice Hall, 1995, Donald R. Woods).

Table 6-1

Instruments

Failures

Analyzer: GLC Analyzer: CO2 Analyzer O2 Analyzer: general Analyzer: pH Flow: Dp transducer Liquid level Dp transducer Liquid level float type transducer Liquid level general Flow: general Purge system Pressure: general Temperature: transducer

20.9 failures/annum 10.5 7 6.2 4.3 1.9 1.8 1.6 1.6 1.1 1 1 0.9 0.9 0.6 0.3 0.15 0.1

Temperature sensor

Equipment

turbine centrifugal pump distillation column, reactor exchanger

The tests should be based on a good understanding of the equipment and the system. Example 6-1: For cavitation of a centrifugal pump, we could a) listen for the crackling noise typical of cavitation, and b) reduce the flow through the pump by partially closing the line on the discharge line. This should cause crackling noise to subside. Fundamentally, at lower flowrates, less NSPH is required and the friction loss on the suction side is reduced by the (velocity)2. Activity 6-1: Selecting tests We hypothesize that there is an obstruction in a 10-m length of pipe, that has numerous bends and fittings, through which liquid is flowing. Andre suggests the following tests:

a. b. c.

stop the operation, open the pipe at the various fittings and look. increase the flowrate and note the difference in pressure drop. decrease the flowrate and note the difference in pressure drop.

201

202

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

d. e. f.

estimate the pressure drop and compare with measured values. stop the operation, open one end and send a plumber’s worm through the line. and maybe there might be some others.

What would you select and why? Activity 6-2: Selecting tests In a refrigeration cycle we hypothesize that impurities got into the refrigerant. Comment on the appropriateness, limitations and assumptions of each of the following tests:

1) 2) 3) 4)

check the pressure difference between the refrigerant side and the process side. sample the refrigerant and analyze for impurities using GLC. read the temperatures and pressures and compare these with the data on pressure–enthalpy charts for the refrigerant. read the pressure gauge on the compressor suction.

Activity 6-3: Case ’6 For Case ’6, Saadia completed the following chart on the Trouble-Shooter’s Worksheet for the case of the utility dryer (details are given in Section 4.4). Complete the chart below by matching the actions to the hypothesis. Comment on the appropriateness of each action.

Symptom a. exit air “wet”: 3 higher than specifications b. pressure drop double expected value Activity 6-3:

Hypothesis chart for worksheet.

Working Hypotheses

Initial Evidence

1. Steam leak 2. Excessive moisture carryover from the separator 3. Valve S2 leaking 4. Adsorbent lacking adsorption capacity. 5. Absorber on-line too long: breakthrough 6. Not enough regeneration time 7. Condenser not cooling sufficient 8. Instruments wrong, pressure 9. Absorbent broken down 10. Temperature TRC, T3 reads low

a S S S S S S S N ? S

b N N N N

S S S

c

d

Diagnostic Actions e

A

B

C

D

6.1 Thinking Skills: How to Select Valid Diagnostic Actions

Diagnostic actions: A. B. C. D. E. F. G. H.

Test calibration of temperature T3 Test calibration of pressure gauges P1 and P4 Calculate a mass balance on moisture Calculate the expected removal of adsorbent from adsorption/kg adsorbent data Sample adsorbent to determine if damaged Read pressure drops across different parts of the system and when different parts are “on-line” Predict water removed via regeneration, temperatures and moisture in exit gas profiles and compare with data taken over several cycles Vary the regeneration time allowed

Resources for Gathering Data Here we consider the types of equipment to take, and the time required to do some tests. 6.1.3.2

a)

Simple equipment and stuff to take with you on-site

Here are some of the things I used to take with me on a trouble-shooting mission: . .

.

.

. . . . . . . . .

notebook and calculator: keep accurate records. leather gloves: especially important to allow me to sense the temperature on either side of a steam trap. stethoscope: very useful to magnify the sounds inside a vessel, steam trap or valve. This helps to identify, for example, the typical flow through a thermodynamic steam trap; to listen for vibrations. string: in trying to estimate the diameter of a pipe or insulated pipe I found it much easier to measure the length of string around its circumference and then calculate the diameter. tape measure. clamp-on ammeter. flashlight. stopwatch. knife. pens and marker, tape, labels, sample bags. tachometer. magnifying glass. brick: to help me estimate a mass or energy balance I often needed to know the flow of water in the drains. On some of the plants these were easily accessible. A brick could be placed in the rectangular drain to create a dam or “weir” and then from a measure of the height of liquid above the “weir” the flow could be estimated.

203

204

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process . . .

.

.

pail and stop watch: measure other types of flows. long needle: allowed me to estimate the thickness of insulation. spray bottle of soap solution and tape: to identify leaks around valve stems, spray soap solution. For flanges, tape around the flange and then spray on a single hole you make in the tape. a surface pyrometer or laser pyrometer may be a useful addition, especially if steam and steam traps are causing problems. camera.

In your journey as a trouble shooter, you too will collect your favorite collection of simple things to make it easy for you to uncover the secrets of the process. b) Time it takes for gathering data Here are some examples of the time it takes to gather certain types of data and to make alterations to the process. . . .

Laboratory analysis: routine analysis: 2 h; special test for impurities 8 h. Instruments: install orifice plate 10 h; rotameter, 5 h, level gauge 8 h. Agitator: hook up, 40 mh; blower, hook up 25 mh; pump, hook up 50 mh.

6.1.3.3 Personal biases and style in collecting evidence and reaching conclusions Trouble shooters create hypotheses and then select cues, evidence and tests to confirm or disprove the hypotheses. They check that the evidence/ data/ cues actually support their hypothesis. Sounds straightforward. However, most mistakes occur here. Each person has a personal style. Furthermore, mistakes and biases affect how data are collected and conclusions reached. Consider each in turn.

a)

Personal style in trading off data gathering versus taking action

Each of us has a preferred style of making decisions. Some prefer to be active, to make choices even though they might be wrong. Others, want to gather data and really understand the situation before action is taken. The P-J dimension of the Jungian typology, or Myers Briggs Type Indicator (MBTI) may provide you with insight as to your preferences. For example, a dominant P style is characterized by wanting to collect detailed information and data before making a decision; a person with a dominant J style wants action and may use an approach of changing a suspected cause before doing simple tests to check whether that really is the cause. Example 6-2: For the Case ’10, to dry and not to dry, and described in Section 5.4, the hypotheses selected by Jason and Heather are 1) wash water carryover from the centrifuge, 2) cycle from screen not coordinated with cycle from centrifuge, 3) feed crystals too wet from the screen, 4) condensate trap on the dryer is malfunctioning causing the condensate to back up in the tubes and reduce the heat transfer area, 5) more fines into the centrifuge causing the filter cycle to be too long; fines carryover to the cen-

6.1 Thinking Skills: How to Select Valid Diagnostic Actions

trifuge causing blinding in centrifuge, and 6) crystal size change; crystals change shape so that filtering is different. Jason, after consulting with the vendor of the dryer, wants to replace the steam trap on the dryer (since the vendor suggests that this is a likely cause and it’s relatively easy to change). “I’m convinced it is hypothesis ’4. Let’s act.” Heather, on the other hand, wants to sample the crystals going into and out of the centrifuge every 30 s for three cycles of operation. “The samples can be analyzed for water content and for particle-size distribution. This is relatively easy to do and this will help us test hypotheses ’1, 3, 5 and 6. On the other hand, Jason, if you really feel strongly that it is the steam trap, let’s go out and check out the trap first.” Comment: Here we have apparent disagreement. Jason is showing predominant J behavior; Heather, predominant P. Provided Jason and Heather see their disagreement as “Hurrah! we balance out our different styles,” then a better result will occur. If Jason and Heather had both been dominant J, then the steam trap would probably have been changed, only to discover that the steam trap was not the fault.

b)

Common biases in collecting evidence

The four major types of error are pseudodiagnosticity, confirmation, availability and representative biases. .

.

. .

c)

pseudodiagnosticity or overinterpretation: actively seek worthless data and change opinion based on irrelevant data; treat noncontributory cues as relevant. This happens about 30% of the time and is the most common bias. confirmation bias: actively support a favored hypothesis even though all the evidence points elsewhere. Seek confirming information and ignore disconfirming cues. availability bias: prefer to use data that are more readily available. representativeness bias: see similarity that doesn’t exist between two events. Even when given the correct underlying knowledge trouble shooters consistently disregard that knowledge in favor of stereotypes about how “representative” these particular data/characteristics are. Common biases in reaching conclusions

These biases tend to relate to personal preference and the amount of training and experience. Personal foibles include: premature closure and anchoring may occur despite experience and training. .

.

premature closure: the conclusion is not justified by existing data. Tend to be a Jungian typology “dominant J” (discussed in part i). anchoring: adhere to a preconceived belief even though the evidence refutes it.

205

206

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

Three other biases that occur usually with inexperienced trouble shooters who lack the training and experience with the process equipment are inadequate synthesis, underinterpretation and misinterpretation. .

.

.

inadequate synthesis: unjustified conclusions are drawn. The trouble shooter fails to use knowledge effectively in interpreting and making inferences from the data. omission of clue or underinterpretation: an important clue is ignored. This occurs about 2% of the time. wrong synthesis or misinterpretation: the available data contradict the conclusion. This occurs about 6% of the time.

Activity 6-4: Confirmation bias Consider the task of testing the hypothesis that “Every card that has a vowel will have an even number on the other side”. The four cards shown below in Figure 6-1 are available to test this hypothesis. (From Johnson-Laird and Watson, 1970). Which card or cards do you need to turn over in order to test the validity of the hypothesis?

U Figure 6-1

F

2

9

Four cards.

Activity 6-5: Feedback about your style To give yourself-some feedback about your style, consider the following case. Assume that the information given in the scenario is factually correct. A code letter is given at the end of separate bits of evidence in the account of the murder, for example “Tom Dayton had many enemies (a).” where the code letter is (a). Worksheet 6-1:

The Tom Dayton murder (adapted from Sherlock Holmes).

Tom Dayton had many enemies. (a) He was a scalawag and a prankster who never passed up an opportunity to embarrass someone through a practical joke (b). It was Tom who invented the joy buzzer and the whoopee cushion, and some even credit him with having originated fake vomit. It is well known that Tom’s favorite target was his old headmaster, Stanley Bosworth, (c) at Bromley School. Stanley Bosworth was the victim of some of Tom’s most elaborate pranks. Tom’s eventual marriage to Bosworth’s daughter, Melissa, was considered by many to be Tom’s ultimate joke (d) on the respected headmaster. Among the more prominent victims of Tom Dayton’s past pranks were Judge Walter Brighton (e), Lord and Lady Morton (f) of Westchester, banker Mortimer Fawcett (g), Doctor Fabian Peerpoint (h), and tobacco merchant Dawes Flescher (i).

6.1 Thinking Skills: How to Select Valid Diagnostic Actions

All of the above were present at the dinner party held on the Bosworth yacht in honor of Stanley Bosworth’s 60th birthday (j). Following an uneventful dinner, most of the guests retired to their staterooms to freshen up. The clock in the dining room struck 10 pm (k) when a shot rang out (l). Most later claimed they heard a second shot (m). All aboard the yacht, including the yacht’s captain, Jonas Fenton, (n) and cook, the curvaceous Mildred Weekson (o), arrived at Tom Dayton’s stateroom to find him dead – shot in the forehead (p). A smoking revolver lay near the doorway (q); Tom’s body lay on the floor across the room (r), just below an open porthole (s). Scrawled in the dust near Dayton’s body were the initials SB (t). In the corner of the room was a suitcase filled with $500,000 (u). No bullets were found in the walls, ceiling or floor of Tom’s room (v). Who killed Dayton? During the investigation the following factual evidence was produced: w. Although just about everyone claim they heard two shots, Tom Dayton had one bullet in his head. x. Lady Morton was being blackmailed. y. The clock in the dining room was 15 minutes slow. z. The $500 000 in the suitcase was counterfeit and was accompanied by a withdrawal slip for $1 000 000 from Mortimer Fawcett’s bank. aa. Bosworth angrily stated at dinner that he felt Tom was mistreating his daughter. bb. The actual murder weapon was found under water below the porthole. cc. Dr Fabian Peerpoint advocates mercy killing. dd. The smoking revolver in the room belonged to Stanley Bosworth. ee. Tobacco merchant Dawes Flescher is a talented mountain climber. ff. Jonas Fenton, the yacht’s captain, went to school with Tom Dayton. gg. Dr Fabian Peerpoint revealed that Tom Dayton was terminally ill with only a few months to live. hh. The bullet in Tom’s head did not come from the smoking revolver on the floor in Tom’s room. ii. Melissa says that she was visiting her father in his stateroom from 9:45 pm until she heard the shot. jj. Mildred Weekson was having a secret affair with Tom Dayton for the past two years; and with Lord Morton. kk. The shot that killed Tom Dayton was fired from outside the porthole. Directly below the porthole is water. Based on the evidence so far would you: 1. 2.

accuse ____________ of murder based on evidence (list the letters of the evidence supporting your conclusion) ____________________________. accuse Dr. Peerpoint of mercy killing based on evidence (list the letters) _____________________________________________________________

207

208

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

3.

conclude that Tom died from __________ based on evidence. (list the letters) ________________________________________________

or... . 4. require that the following information is needed before any conclusion can be drawn: ll) what pranks had Tom played on _________________________ mm) check for poison in Tom’s body nn) other _______________________________________________

Feedback about your style is given in Appendix G. d.

Cautions about interpreting data

All data have errors. We should have an idea of a. b. c.

What is an acceptable error in all the target or specification conditions. What is an acceptable error for all the measurements. When is there a significant enough deviation that we recognize that something is wrong.

The evidence may be relayed to us as fact, opinion or opinionated fact. Actually it is difficult for people to relate just the facts; they want to infer and give their own interpretation to the information. As trouble shooters we need to separate carefully fact from opinion. We need to concentrate on getting the facts and the focus on relating the facts to the cause. Maybe time variation occurs. Perhaps the data have to be collected over a time cycle. Normal instrument error or a fault? All instruments have errors. Judgment is needed to tell whether the following set of measurements are about what we expect from the instrument? Or that there is trouble on the process? Example 6-3: Here are records from the operator’s log book of the digital printout for the temperature at the top of the reactor. 932.56; 938.64, 930.28, 935.67, 932.19, 937.52. If the expected temperature is 936 are the values within acceptable error or is something wrong? Evidence that trouble exists usually comes with error associated with it. We have to distinguish whether the variation in the data represent “expected error” or a fault. Example 6-4: The laboratory reports that the analysis of the recycle gas shows 3.4% methane. It should be 0.3%. All the other instruments on the process read normal. Follow-up action: Another sample was taken; the laboratory reported 3.5%. All other instruments still read normal.

6.2 Thinking Skill: Consistency: Definitions, Cause–Effect and Fundamentals

Next: No action was taken by the process operators. They continued to operate the plant as usual. Each day, the laboratory reported values of methane in the range 3.5 to 3.65%. Later: Two weeks later the head chemist returned from holidays and notes that while she was away, all the analyses had been taken on the “calibration mode”. When the instrument is set on its correct settings, the concentration of methane in the recycle line was 0.25%. Too often we hope that the data are applicable. A colleague, in designing a petrochemical plant was unable to locate the physical properties of the organics. He decided to assume they were the same as water and hope that they would work out. Just a short time spent in a critical assessment of this assumption would have saved six months of wasted work. Too often we accept data from the published literature; yet about 8% of data published are mistakes. “The temperature into the hydrodealkylation reactor is >1150 C” states one reference. This should read >1150 F. Another example is that a major handbook published an incorrect value of the heat of vaporization through several editions. Check the data coming from computer programs and simulations. Check the physical property package estimates. 6.1.4

Summary

Trouble shooters gather information to solve the problem. Information can be gathered for six different purposes: a) safety check plus provide background information about the situation, b) put the situation in the context of recent events, c) understand what should be happening, d) check what is happening now, e) gather data to test hypotheses and, perhaps, f) take immediate corrective action. Criteria are given for how to select which type of information to gather. In general, start simply, isolate variables, test the “most likely” faults first and have a purpose for gathering each particular piece of information. We need to realize that each of us has a preferred style (and perhaps bias) in how we gather and interpret data. Two inventories were used to help identify preferences and biases. These ideas were illustrated by a range of examples and activities including revisiting Cases ’6 (the utility dryer) and ’10 (to dry and not to dry).

6.2

Thinking Skill: Consistency: Definitions, Cause–Effect and Fundamentals

Scriven (1976) emphasizes that the main criterion used in critical thinking is consistency. Indeed, we trouble shoot in a world defined by consistent terms, fundamentals and concepts. We work with facts. A clear distinction needs to be made between facts and opinions. The equipment with which we work has clearly defined symptoms that are generated by each fault. The behavior of fluids and materials follows fundamental principles. We communicate in English, with its defined rules; we work with

209

210

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

mathematics that follows rules. In this section we remind ourselves of those rules and framework in which we check for consistency. 6.2.1

Consistent Use of Definitions

Probably the most crucial concept is Facts. To effectively separate facts from opinion we use definitions based on Obform coding developed by Johnson of the Journalism Department at the University of Wisconsin. a. Facts.

Johnson defines three sources of facts as factual data, conclusions and background information. Factual data may be accepted as fact if we can attribute an observer with being able to hear, feel, smell, taste or see the observations recorded or stated. For example, “the gas evolved from the anode is oxygen and that from the cathode is hydrogen.” The facts we can observe are that colorless and odorless gases are evolved from the anode and the cathode. We can infer something about the nature of the gases from further tests. However, the facts are not that “oxygen is evolved from the anode”, etc. Conclusions from factual data may be taken as facts if the reasoning is correct: 1) the validity of each step in the sequence of reasoning is proven, 2) there are sufficient steps and 3) the nine steps for valid reasoning (given in Section 6.5) are followed. Background information is factual if references are given for the direct quotes and if statements made by people are direct quotations. For the latter, the message that the speaker said might be factually incorrect, but it is a fact that the person said those words. b. Opinion.

This is information other than Facts. c. Opinionated facts.

This is factual information that contains opinion. For example, The temperature is as high as 45 C. The fact is the temperature is 45 C. The opinion is that this temperature is “high”. Put the two together and we have opinionated fact.

6.2 Thinking Skill: Consistency: Definitions, Cause–Effect and Fundamentals

Example 6-5:

Examples of Facts and Opinions

Example statement

Factual statement Factual data about by people operation

John said “The gauge reads 50 kPa.” John said “The pressure is 50 kPa.” John said that the pressure gauge reads 50 kPa. John said the pressure gauge reading is too high.

Yes Yes No No

John said that the pressure is too high. No Yes John said “The pressure gauge reads 50 kPa, the gauge was calibrated yesterday, the sampling line is clear. If I increase the pressure slightly, the gauge reading increases slightly. I infer that the pressure is 50 kPa.” Yes John said “The pressure gauge reads 50 kPa, the gauge was calibrated yesterday, the sampling line is clear. If I increase the pressure slightly, the gauge reading increases slightly. I infer that the pressure is 50 kPa. I conclude that pressure is too high.”

Yes No Yes Opinion via words “too” Opinion Yes

Yes + Opinion

Activity 6-6: Facts and opinions Analyze the following passage and classify it according to facts, opinion and opinionated facts. Here is an accurate account of what happened. The telephone rang! “Trouble out on the ethylbenzene unit,” said Bill. Harry said that he would be right out as he slammed down the phone. As Harry approached the unit Bill came out to meet him and said, “I’m sure that the heat transfer is insufficient in the reboiler to the product column; I’ll show you what I mean.” Harry glanced at the rotameter and saw that the flow to the column was the usual amount of 3000 gpm; the pressure gauge read 150 psig and the bottoms temperature was 140 C. Rounding the column, he saw that the liquid level in the bottoms level gauge was rising at a rate of about 3 cm/min. The liquid level disappeared out of the top of the level gauge. After about two minutes the level reappeared in the sight glass and disappeared out of the bottom of the sight glass within several minutes. “See,” said the operator, “we have lost all the bottoms out of the column just like that!” “It has gone off to the storage tank,” offered John. “No, it has gone through the reboiler and straight up the column. You can see by the instabilities in the pressure gauges that occur just after the level disappears out the bottom of the sight glass,” said Bill.

Based on the account given in Activity 6-6, which of the following statements are True (T), false (F) or can’t tell (?).

Activity 6-7:

1. 2. 3.

Bill said that there was trouble out on the ethylbenzene unit. Harry said “I’ll be out immediately.” The heat transfer is insufficient.

T F ? T F ? T F ?

211

212

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

4. 5. 6. 7. 8.

The trouble is in the reboiler. The flow to the column is 3000 gpm. The pressure was 150 psig. The bottoms temperature was 140 C. The level in the bottom of the column was building up and then suddenly dropping. 9. John said, “ The bottoms have gone off to the storage”. 10. When the level in the bottoms of the column drops the pressure gauges show instabilities.

T T T T

F ? F ? F ? F ?

T F ? T F ? T F

?

Example 6-6: Case ’3: Consider the account of Michelle working on Case ’3. The case of the cycling column (presented in Chapter 4, Section 4.1). At one stage, Michelle is trying to decide if the control system is at fault. Her diagnostic action was to “put the control system on manual.” When the set point was increased manually, the valve stem on the steam moves up, the liquid level appears in the sight glass and continues to drop but shortly thereafter the level appears in the glass and is rising. I conclude that the control system is not at fault. For Michelle’s thinking reproduced above, has Michelle focused on facts? opinion? opinionated facts?

a.

When the set point was increased manually, the valve stem on the steam moves up.

These are both observable, and therefore called “facts”. b.

When the set point was increased manually, the liquid level appears in the sight glass and continues to drop but shortly thereafter the level appears in the glass and is rising. These three are observable and therefore are “facts”.

Comment: Michelle did a good job here. She might have incorrectly said, “When the set point was increased manually, the steam flow increased.” Since Michelle could not have seen the steam flow, saying the steam flow increased would have been an opinion. In summary, both facts and opinions are used to trouble shoot. However, we need to know which are facts and which are opinions. A definition of facts is given. Examples illustrate how to use that definition consistently. We revisited Case ’3, the cycling column. 6.2.2

Consistent with How Equipment Works: Cause ﬁ Effects: Root Cause-Symptoms

Equipment is fabricated and works according the fundamental principles of science and engineering. Therefore, if a fault occurs in the equipment, certain symptoms will appear. We define a symptom as something that can be observed, heard or felt related to the performance of equipment or a system of equipment. The degree to which a symptom can be explicitly observed by the trouble shooter depends on the

6.2 Thinking Skill: Consistency: Definitions, Cause–Effect and Fundamentals

sensors and configuration for the equipment layout. For example, for the usual configuration of a pump the symptoms we might be able to observe include the flowrate (if there is a flowmeter), the exit pressure/head (if there is a gauge), the power on the drive (if the amperage is measured), the “crackling noise” (if we can get close enough to “hear”) and the temperature of the motor/drive and of the suction line (if we can “feel” or use a surface pyrometer to measure the surface temperature). It is important for us to document cause–effects or cause–symptoms for different types of equipment in terms of the usual types of instrumentation that is available. a. Cause ﬁ symptom consistency

An important task is to ensure that we are familiar with cause ﬁ symptom data for different pieces of equipment. Example 6-7: Causes (and symptoms) Three of the common faults with centrifugal pumps (and the symptoms) are: operates at very low capacity (vibration and noise, and pump overheats and/or seizes), rotor not balanced (short bearing life, vibration and noise, short mechanical seal life, pump overheats and/or seizes and stuffing box leaks excessively) and impeller partially clogged with solids (either no liquid delivered or flow lower than expected, power demands higher than expected). Activity 6-8: Listing symptoms for causes or faults. For the depropanizer shown in Case ’8, for the following faults/causes,

a. b.

list the symptoms that would be observed, heard or smelt. estimate the order of magnitudes of the deviations (or the extent of the symptoms).

1.

The vortex breaker in the overhead drum V-30 is welded such that the crosssectional area for flow has been reduced to 15% of the pipe internal crosssectional area. The vortex breaker in the overhead drum V-30 has corroded away. The exit pipe is fully open. In the overhead condenser, E-25, the inerts have been inadequately vented from the shell side (from the tube side) before startup. In the reboiler, E-27, the inerts have been inadequately vented from the shell side (from the tube side) before startup. For the depropanizer, the trays are bent so that the downcomer clearance is 12 what it was supposed to be. For the depropanizer, the trays are bent so that the downcomer clearance is double what it was supposed to be. For the depropanizer, corrosion as increased the diameter of the holes in the sieve trays by 10%. For the depropanizer, tray 5 has collapsed because of inadequate support. For the depropanizer, trays 5, 13, 20 and 25 are not level. They are 30 to the horizontal.

2. 3. 4. 5. 6. 7. 8. 9.

213

214

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

More examples of typical faults/causes are given in Appendix H. Appendix I lists the symptoms for some of these. We can refer to the cause ﬁ symptom statements as If ... then statements and invoke the rules of logic. The characteristics of If ... then statements are (we include the terms antecedent and consequent from the reasoning literature): .

. .

.

If the “cause” (antecedent) is positive, then the “symptom” (consequent) is positive. If the “symptom” (consequent) is negative, the “cause” (antecedent) is negative. If the “cause” (antecedent) is negative, we cannot conclude anything about the “symptoms” (consequent). For example, If the pump impeller is not turning backwards, then we cannot conclude anything about the flowrate and head. If the “symptom” (consequent) is positive, then we cannot conclude anything about the “cause” (antecedent). For example, If the flowrate and head are abnormally low, then we cannot say the impeller is turning backwards. Other reasons can cause this.

This last characteristic causes frustration for the trouble shooter because typically we are not given the cause, rather we are given the symptom and expected to deduce a cause. In other words, symptom ‹ cause. Furthermore, trouble shooting is also confounded because there may be more than one cause that could be contributing the symptoms. “Symptoms” are not necessarily caused by one and only “root cause”. Nevertheless, documenting cause ﬁ symptom information is a good starting point for trouble shooting from which we can develop information about symptom ‹ cause. b. Symptom ‹ cause consistency

The starting point for trouble shooting is a list of symptoms. The challenge is to create hypotheses as to the probable cause that are consistent with the symptoms. This is the reverse of the data discussed in section a. Some example symptom ‹ cause information is given in Chapter 3 for different equipment. The challenge is, that unlike cause-effect data that are “true”, the symptom–cause data are “not necessarily true”. Activity 6-9: Symptoms and causes Consider Case ’11, the symptoms and the hypotheses/causes. Are they consistent? Case ’11: The Lazy Twin (courtesy of W. K. Taylor, B. Eng. McMaster, 1966) The situation: Pump A is usually running. Pump B, identical to pump A, is the spare pump. The operators of the process notice that the flow meter, FRC-100, shows that not enough flow is going to the process. So they switch over from pump A to pump B by shutting off the power to the drive motor for pump A and turning on the power to the drive motor for pump B. Now the flow to the process comes back to where it should be. What is wrong? The pump configuration is shown in Figure 6-2. Data about the flow and setting of the valves are given in Figure 6-3.

6.2 Thinking Skill: Consistency: Definitions, Cause–Effect and Fundamentals

PI 210

V201 V202

FRC 100

A T200

V100 V200

PI 220

V205 V206 B

Figure 6-2

Operator’s sketch of the equipment for Case ’11.

ONLY PUMP A ON

ONLY PUMP B ON

F100

S. P.

F100 CONTROLLER OUTPUT

100%

VALUE = 100%

0%

TIME Figure 6-3

Flow and valve-setting data for Case ’11.

215

216

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

For this Case ’11 we might summarize the symptoms as: a. b.

when pump A is running, the flow meter shows that not enough flow is going to the process when pump B is running, the flow meter shows the correct flow is going to the process

In the following table, column “a” represents symptom “a”, “b”, symptom “b”. Consider each hypothesis and in columns “a” and “b” place a code S supports; D disproves and N neutral or can’t tell this hypothesis or fault. Table 6-2

Worksheet for hypotheses.

Working Hypotheses or cause

Initial Evidence: symptoms a

b

c

d

Diagnostic Actions

e

A

B

C

D

1. Flowmeter is reading wrong. 2. Pump A has a sandwich stuck in the impeller. 3. Pump B is much bigger capacity than we need 4. Full electrical current does not go to pump A. 5. A fuse is blown in the circuit to pump A. 6. Liquid is vaporizing in the line to pump A and so pump A cannot produce the expected flow. 7. The operator cannot read the flowmeter correctly. 8. Some corrosion products are clogging the flowmeter. 9. There is not enough Net Positive Suction Head for the Pump A. 10. There is not enough liquid in the tank to the pumps. 11. The signal is not getting from the controller to the valve. 12. The flow-control valve is stuck partway shut.

c. Symptom or “root cause” dilemma

In brainstorming possible causes (that are consistent with the symptoms) sometimes the idea is another symptom, instead of a “root cause”. Example 6-8: Chapter 3 suggests that for distillation if all temperatures are falling simultaneously then the cause might be low boilup. Is “low boilup” a root cause? No. The root cause is something that causes low boilup, which (for a thermosyphon reboiler) could be such causes as condensate flooding/ inadequate steam supply/ steam valve closed/ superheated steam/ boiling-point elevation of the bottoms/ inert blanketing/ film instead of nucleate boiling/ increase in pressure on the process side/ undersized reboiler/ control system fault/ fouling on the process side/ low liquid level/ high

6.2 Thinking Skill: Consistency: Definitions, Cause–Effect and Fundamentals

liquid level/ heavies in the feed/ pipe lengths< design/ pipe diameter > design/ process fluid level < 30–40% of tube length. To test if we have a symptom or a root cause, test the postulated “cause” by asking “What would cause this?” If there is no new answer, then we probably have the root cause. Root cause For Case ’11, The Lazy Twin, and the possible causes listed, which of these causes are root causes and which are symptoms? In summary, this is perhaps the key section in this chapter. Consistency between cause and symptom; symptom and cause; hypotheses and symptom are key to trouble shooting. It is simplest to start with typical causes/faults and listing symptoms that are produced. Then we move to the trickiest part, namely the reverse: identifying possible causes from a given set of symptoms. A new Case ’11, the lazy twin, was introduced. The challenge of working with “root causes” was emphasized. Activity 6-10:

6.2.3

Consistent with Fundamental Rules of Mathematics and English

Our reasoning and actions need to be consistent with the rules of mathematics and English. For Mathematics consider the following: Example 6-9: Please check the reasoning in the following: rﬃﬃﬃﬃﬃﬃﬃ rﬃﬃﬃﬃﬃﬃﬃ 1 1 = 1 1

rﬃﬃﬃﬃﬃﬃﬃ rﬃﬃﬃﬃﬃﬃﬃ 1 1 = 1 1 pﬃﬃﬃ pﬃﬃﬃﬃﬃﬃﬃ 1 1 pﬃﬃﬃﬃﬃﬃﬃ = pﬃﬃﬃ 1 1 pﬃﬃﬃﬃﬃﬃﬃpﬃﬃﬃﬃﬃﬃﬃ 1 = 1 1 1 = –1 What went wrong? Here we encounter an incorrect answer so we know that we were inconsistent with the rules of mathematics. But what happens when we don’t have a surprise? We need to condition ourselves to check for mathematical consistency in all the calculations we do as we trouble shoot. We do have guidance in Sections 6.2.4 and 6.2.5 because our answers should be mathematical answers that are consistent with fundamental laws and with experience. To some extent Saadia followed these principles in Case ’6. She calcu-

217

218

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

lated the expected results ahead of time. Then when she saw the results she would know if there were surprises. For English, we have the rules of grammar and the meanings of words. Words only have meaning in people. Therefore, we need to use words that are “understood” by all. Consider the following examples: Example 6-10: An open window, a gust of air, glass on the floor, water on the floor, Mary’s dead. Why? Example 6-11: A man is walking along. He tears his sleeve on a sharp corner. Within minutes he is dead. Why? In these two examples, if we realize that Mary is a goldfish and that “walking along” means walking on the bottom of the ocean, then these are not puzzles and we all understand. When we are trouble shooting, some of the words that are ambiguous and that interfere with communication include: “The instrument looks OK!” “I followed the usual practice.” Activity 6-11: Ambiguous words Make a list of ambiguous words that you encounter when you trouble shoot. Then think of ways you can overcome this ambiguity. In summary, critical thinking requires that we are consistent in our use of Mathematics and English. 6.2.4

Consistent with Fundamental Principles Of Science: Conservation of Mass, Energy, High to Low Pressure, Properties of Materials

Mass and energy balances, pressure profiles and an understanding of the unique properties of materials are probably the most useful source of information in trouble shooting. Yet, trouble shooters often neglect these. 6.2.5

Consistent with Experience

Having a wide range of “rules of thumb” and memorized experience factors help us to validate and check answers and ideas for consistency. Sources include Brannan (1998), Walas (1988), Woods (1995) and Woods (2001).

6.3 Thinking Skills: Classification

6.2.6

Summary

Consistency is the key to critical thinking. Here our focus was on consistency with . .

. . .

definitions: to sort facts from opinions; cause ﬁ effect information for equipment: to guide in the creation and testing of hypotheses; the rules of mathematics and English: to help us focus on accuracy; the fundamentals: to remind us of the basics to check; experience: to validate our thinking, calculations and assumptions.

Most of the effort was spent on the first two topics: facts versus opinions and sorting out cause ﬁ effect relationships.

6.3

Thinking Skills: Classification

Classification is dividing the whole into parts such that there is a meaningful relationship among the parts. The classification is done for a purpose, and for each level of classification there must be one and only one criterion. Each level of classification should be complete. There should be no single subclass. The classification should be consistent in the amount of detail given and the classification should have neither faulty coordination nor faulty subordination. These are the general characteristics of a good classification. Skill in classification is needed to: .

.

classify the starting information during the stage of Define the stated problem, discussed in Chapter 2. classify the possible causes generated during the brainstorming session as part of the Explore stage, discussed in Chapter 2 with the brainstorming session being illustrated in Chapter 5.

6.3.1

Classify the Starting Information

For the purpose of understanding the problem, the starting information should be classified using the criterion “what are the key parts to the trouble-shooting problem statement”. Usually the key parts are: . . . . .

the situation or system, the symptoms that suggest a fault, the triggering event, the criteria for success (either stated or inferred), the constraints.

219

220

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

The focus here is to identify the “symptoms”. The symptoms are defined as “evidence stated by the operators, those displayed by sirens or alarms, reports from the laboratory, or the analyzer showing non-specification behavior, or complaints from customers describing unexpected or unusual behavior.” Sometimes the symptom is accompanied by a triggering event. The triggering event is often expressed as “When ...”. Sometimes there is no apparent triggering event. Sometimes the triggering event is not related to the symptom; it is coincidental. However, when we start trouble shooting the problem we may not realize the coincidence. Some examples of triggering events include: When we increased the flowrate to new higher capacities ... When we started up for the first time ... When we started up for the first time after maintenance ... When the operator increased the pressure in the column above the usual value ... The mental TS process is to scan all the given, starting information and apply the definition of “symptom” to identify this portion of information. Example 6-12: Identifying symptoms. For Case ’9, the symptom is stated ambiguously as “nothing happened.” Before we can proceed, we should identify why this is unexpected behavior. From the context, we can rephrase this as, the vacuum inside the deodorizer should have been sufficient to suck the Fuller’s earth into the deodorizer when the connecting valve was opened. However, the Fuller’s earth was not sucked into the deodorizer because we could not see powder dumping in when we looked through the view port. Example 6-13: Identifying pertinent triggering events: Several When’s occur in this situation in Case ’9: “When the valve was opened;” “when the vacuum had stabilized”; “when we started up this plant for the first time”. Are any of these triggering events? The most likely triggering event is when we started up this plant for the first time. 6.3.2

Classifying Ideas from Brainstorming

As illustrated in Chapter 5, usually we obtain 50 to 100 ideas in a brainstorming session. Of these often 80% are non-sensible. That’s OK because it is worth a few minutes to generate such ideas since hidden among the crazy ideas are some real winners. After the brainstorming session, however, we need to classify for the purpose of listing at least six feasible hypotheses. The first level of classification would be “technically feasible” versus “non-sensible” or it could include an additional class of “interesting”. The next level of classification would be to sort the technically feasible ideas according to underlying type of cause.

6.4 Thinking Skills: Recognizing Patterns

6.4

Thinking Skills: Recognizing Patterns

Patterns in the information could include increasing or decreasing trends, cycling data, series or unexpected systematic external changes that affect performance. Here are three examples. Example 6-14: One process operator might be experiencing great personal stress at home. Whenever this operator is on shift the process does not run smoothly. If we were unaware that the operator was under stress it would be hard to identify the cause. Example 6-15: Whenever there is a severe thunderstorm some of the instruments on the plant malfunction. Example 6-16: Everything seemed to be working OK although gradually the performance was moving off specification. The cause was a minor leak in a heat exchanger. In these three examples, the trick was to be able to see a pattern between stress and poor performance; thunderstorms and malfunctioning instruments; and gradual changes in performance and cross contaminant because of a leak. The general guidelines and criteria for identifying patterns are not very helpful. Keep an open mind. Plot (and be sensitive to) time variation in data. Note the possible interaction between different systems and especially the interaction between cycling processes. Consider now more about the patterns in the symptoms and in how we collect data. 6.4.1

Patterns in the Symptoms

The unexpected and outside factors that can affect performance include the weather: .

. .

. .

showers and rain; extreme cold; hot sunny weather affects cooling towers and air-cooled condensers. electrical storms affect sensitive electronic/electrical instrumentation. atmospheric pressure affects pumps that pump liquid from tanks that are open to the atmosphere. extreme cold can freeze vents shut; freeze bucket steam traps. hot weather plus steam tracing can cause vaporization and vapor binding.

Cycling processes, that might interfere with other cycling processes include: . . .

batch distillation and batch reactors, adsorption and ion exchange, centrifuges and filters.

221

222

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

When cyclical processes are used, the data should be reported for different parts of the cycle and not reported just as average values. Averages mask the trends and patterns. When cyclical processes occur, the frequency of the cycle should be determined. For example, cycling steam flow or column pressure for a distillation (30 s to several minutes) coincides with a specific set of probable causes. On the other hand, thermodynamic steam traps should have about six cycles/minute. Example 6-17: The cycling level in the bottom of the column. Engineer Ted Tyler used the cycle to help him predict that the collapsed tray, causing the cycling, was tray 13 to 20 from the bottom. Example 6-18: Another cycling level in the bottom of the column. In another cycling column, the coordination between the cycle of the level and the cycle of the steam trap, as determined by listening with a stethoscope to the trap, helped pinpoint the steam trap as the fault. Example 6-19: The bombastic bagging machine. PVC powder was pneumatically conveyed to a hopper located above the bagging machine. The dust was filtered from the conveying air. The problem was that periodically there would be loud thumping in the hopper that sounded like an avalanche of powder suddenly arrived in the hopper. Yet all the evidence suggested that the powder was being fed to the hopper smoothly and continually. This periodic appearance of slugs of powder caused problems with the continual bagging operation. What was discovered was that the bags in the dust filter were being periodically cleaned. The design fault was that the powder-laden air flowed up the inside of the cylindrical, vertical bags so that the blow-ring cleaners could travel on the outside of the bag. As the blow ring cleaned a bag it blew the powder into the center of the cylindrical vertical bag and the powder was held in the up-flowing conveying air. The powder became fluidized in some of the bags. When the mass of fluidized powder became excessive in the central core of one of the bags, the whole mass from the fluidized bed would avalanche down into the hopper. Which bags became fluidized beds and when the bed became unstable seemed to be completely random. The corrective action was to replace the configuration so that the feed air flowed down the central core of the bag. In this example the cyclic operation was the cleaning of the filters. The gradual buildup of solids, of corrosion products or of trace contaminants can cause trends in performance. The cause might be that the solids are too wet, the amount of purge is insufficient or a leak. Gradual changes that cause trouble are often the most difficult to detect. Good records are needed to note the trends. Unexpected pockets of water or condensate in low-lying sections of pipe can cause patterns in performance.

6.5 Thinking Skill: Reasoning

Activity 6-12: Case ’7: The reluctant vacuum crystallizer The data given in this case are cyclical! Analyze the approach Frank took, as described in Section 4.5 of Chapter 4, and look for patterns in the data that Frank should have spotted. Activity 6-13: Case ’10: To dry and not to dry This case is described in Section 5.4. The operation consists of a continuous crystallizer, a batch screen, a batch centrifuge and a continuous rotary dryer. The symptoms are “The crystal product from the rotary dryer has a moisture content of 4.5% whereas the design value is 1.5%. Cake seems to be building up in the dryer feed chute, in the feed screw and on the steam tubes at the feed end of the rotary dryer.” Critique these symptoms. 6.4.2

Patterns in the Evidence

When we suspect a pattern, then we should collect data that will make the pattern easy to spot. Example 6-20: In Case ’6: The utility dryer Saadia sampled every 15 minutes for the 120-minute regeneration cycle. She also astutely sampled such that she could have data for bed A adsorbing and then being regenerated (and for B adsorbing and being regenerated).

Case ’10: To dry and not to dry Before the change in washing, the centrifuge and screen cycle were 10 minutes online and 3 minutes water wash. What sampling would you recommend to try to identify a pattern?

Activity 6-14:

6.5

Thinking Skill: Reasoning

In evaluating evidence and making decisions about the hypotheses, the reasoning that we use should be sound. In this section we outline an organized nine-step approach to evaluate the reasoning. In general, an overall nine-step approach for critical thinking is: 1. 2. 3.

Classify all the given information into the key parts. Write the conclusion. Identify the context. What are the stated and the inferred contexts? Are there other pertinent contexts? 4. Check the definitions; identify and clarify ambiguous terminology. 4a. Change the argument to show the relationship between the conclusion and the evidence.

223

224

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

5.

6. 7. 8. 9.

Consider the evidence. What is the quality of the evidence? Is it sufficient? Relevant? acceptable? What are the facts? What are the opinions? Is the correct type of data gathered for the purpose? Usually the purpose is to test a hypothesis. Check the evidence. The evidence is the basis of an argument. Formulate the assumptions. Assess the quality of the reasoning. Assess the strengths of the counterarguments. Evaluate the consequences and implications.

To illustrate this process consider some of Michelle’s reasoning as presented in Chapter 4, Section 4.1 as she tries to resolve Case ’3, the case of the cycling column. At one stage, Michelle is trying to decide if the control system is at fault. Her diagnostic action was to “put the control system on manual.” When the set point was increased manually, the valve stem on the steam moves up, the liquid level appears in the sight glass and continues to drop but shortly thereafter the level appears in the glass and is rising. I conclude that the control system is not at fault. 6.5.1

Step 1: Classify the Information

The given information includes the context, the conclusions, the evidence related to the conclusion, the stated counterarguments and the stated assumptions. Later we will add inferred assumptions and counterarguments. The context: the situation. Supply WHO, WHAT, WHERE, WHEN IS and IS NOT details plus the equipment and system. The conclusion: that is usually expressed as either what is the cause or what is not the cause. The evidence: collected as described in Section 6.1. The stated counterarguments: evidence that disproves a statement or hypothesis that you are trying to prove. The assumptions: a statement for which no proof or evidence is offered. The qualifier: a constraint or restriction or limiting condition on the conclusion. Example 6-21: Classify the information given in the short segment of Case ’3, the case of the cycling column: Context: a level in the sight gauge at the bottom of a distillation column is cycling. The frequency is about once per 2 minutes. More details will be given in Step 3. The control system was “put on manual.” Conclusion: I conclude that the control system is not at fault. Evidence: When the set point was increased manually, the valve stem on the steam moves up, the liquid level appears in the sight glass and continues to drop but shortly thereafter the level appears in the glass and is rising. Stated counterarguments: none in this passage. Stated assumptions: none in this passage.

6.5 Thinking Skill: Reasoning

Inferred assumption: the control system actually was on manual; the level in the sight glass is related to the level in the bottom of the column. Stated qualifier: none. 6.5.2

Step 2: Write the Conclusion

It is important that the focus of the reasoning is stated clearly. To locate the stated conclusion look for words like “therefore,” “because”, “I conclude”, “so” and “thus”. Usually the conclusion is near the end of dialogue. There may be several conclusions each building on another. Identify the main conclusion. Write it down so that we can focus on the conclusion. The conclusion can be in many different forms. .

.

.

.

.

We could be identifying the context (as in Section 6.5.3) and our conclusion is “Therefore, we have correctly identified the context.” But have we? We could be monitoring our trouble-shooting process (as was recommended in Chapter 2) and we conclude “Thus, we have completed the hypothesis generation stage, now let’s move to the data collection stage.” But is this correct? We may conclude that “In summary, we have selected a reasonable set of hypotheses from the brainstorming activity in Chapter 5.” But have we? Are the causes root causes or are some of them symptoms? We might have listed the symptoms and the hypotheses and concluded that “Therefore, hypothesis ’1 is consistent with the evidence.” But is this a valid conclusion? We might have gathered some evidence and concluded that a hypothesis has been confirmed, denied and can’t tell.

Example 6-22: Michelle’s main conclusion is “I conclude that the control system is not at fault.” The question is, Is her conclusion valid? Label each conclusion. 6.5.3

Step 3: Identify the Context

What are the stated and the inferred contexts? Are there other pertinent contexts? The context depends on the conclusion. I find it helpful if I sketch a cross section of the equipment to remind me of the internal possibilities. I draw a line around it to define the system showing the ins and outs. Label specific contexts. Example 6-23: in Case ’3 the context is that the level in the sight glass cycles. The distillation column is as shown in the Case and, for this portion of the evidence, the control system was placed on manual. Whether this is included as a qualifier or context depends on how detailed you want to present the argument. A sketch of the context is illustrated in Figure 6-4.

225

226

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

Figure 6-4

Sketch of the system.

This sketch includes more detail around the steam control valve; it includes the bottoms draw-off and uses the dotted line to indicate the system. 6.5.4

Step 4: Clarify the Meaning of the Terminology

Identify and clarify ambiguous terminology. Write out your interpretation of any ambiguous words and any unstated but intended implications. Check that you understand all the words and terminology. Often, on process plants the personnel create jargon words for operations and pieces of equipment. For example, on one plant the condensate receiver was called “the pig”. If you are unfamiliar with the terminology you need to ask for clarification. This takes courage. Use the principles of consistency outlined in Section 6.2, especially Sections 6.2.1 and 6.2.3. Example 6-24: in Case ’3 the words that Michelle should know include “the set point”, “control system” and “on manual”. On manual = signal from the sensor no longer alters the steam-valve setting; manually change the set point and the valve should respond.

6.5 Thinking Skill: Reasoning

6.5.5

Step 5: Consider the Evidence

In trying to decide on the validity of Michelle’s conclusion the factors to consider include: What is the quality of the evidence? Is it sufficient? relevant? acceptable? What are the facts? What are the opinions? Are the correct types of data gathered for the purpose? Usually the purpose is to test a hypothesis. Check the evidence. A diagram usually helps. 6.5.5a

Identify the Evidence

Label each assertion. Sometimes several assertions appear in the same sentence. Separate each one. Don’t give a separate number to the same assertion even though the same assertion might be stated several times. Give numerals to conclusions, counterarguments and assumptions. Example 6-25:

For Case ’3

[With the control system on manual, 2]. [When the set point was increased manually, 3] [the valve stem on the steam moves up, 4], [the liquid level appears in the sight glass and continues to drop but shortly thereafter the level appears in the glass and is rising. 5] [I conclude that the control system is not at fault, 1] 6.5.5b

Check for Consistency

Our basis is consistency. Systematically scan over the different elements for consistency in Section 6.2. 1) Label facts, opinions and remove the opinion from opinionated facts. 2) Check for cause ﬁ effect consistency. 3) Check that definitions did not change in different parts of the argument; that symbols are not defined in one way and used in another; that a variable is not treated as a constant; that a principle is not applied beyond its range of applicability, that an assumption is not made and then violated, that coupled variables are not treated as independent. Adiabaticity An adiabatic change is one in which no heat enters or leaves the system. A primitive Joule experiment in which a lead shot is shaken in a cardboard tube is approximately adiabatic. The lead shot gets hotter. Does it then have more heat in it? If so, how does the heat get there? Is the experiment adiabatic after all? Check for consistency.

Activity 6-15:

227

228

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

Example 6-26:

for Case ’3 and Michelle’s reasoning:

[With the control system on manual, 2]. [When the set point was increased manually, 3] [the valve stem on the steam moves up, 4], is consistent with behavior of a control system. Inference: [if the set point is kept constant, 6], [the valve-stem position is constant, 7], [the steam flow is constant, 8]. 6.5.5c

Which Evidence is Pertinent?

The overall context is to find the cause of the cycling level in the bottoms of the column. Therefore any evidence gathered should relate to what does and does not affect the level of liquid in the bottoms of the column. Here we focus on the hypotheses-symptom-action chart from the Trouble-Shooter’s Worksheet. Alternatively we consider the symptom ‹ cause data given in Chapter 3, subject to the concerns expressed in Section 6.2.2b. Some might prefer to use a diagram to illustrate the symptom with all the related possible causes/hypotheses. Example 6-27:

Figure 6-5

for Case ’3, Figure 6-5 is an example of symptom ‹ cause diagram.

Symptom ‹ cause chart for Case ’3.

6.5 Thinking Skill: Reasoning

6.5.5d

Diagram the Argument

Scriven and Halpern (1996) take different approaches in diagramming an argument. Scriven places the evidence at the top of the page and moves down the page to a conclusion. Halpern places the conclusion at the top and uses different evidence to “support” the conclusion. Here are some suggestions: .

.

.

. .

Use the numbers for the evidence and conclusions that were used in Section 6.5.5a. Connect evidence to conclusions with an arrow with separate arrows for each different form of evidence. Rate the quality of the support that the evidence lends to the conclusion: weak, moderate, strong and write this rating on the arrow connecting the evidence to the conclusion. Include assumptions. I use a dotted line if the assumption is inferred. Include the counter-arguments with a wriggly line.

Example 6-28: Case ’3 Figure 6-6 shows a diagram of the arguments related to Michelle’s thinking during the example scenario.

Figure 6-6

Diagram of the arguments.

229

230

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

6.5.6

Step 6: Formulate the Assumptions

An assumption is a statement for which no proof or evidence is offered. Consider each bit of evidence and the resulting conclusion and identify the inferred assumptions. This should include the assumptions made by a) you, b) the designer, c) the process operator, d) the laboratory analysis team, e) the instrument technicians, f) the maintenance personnel and g) vendor. Show the assumptions on the diagram. Example 6-29: .

.

Case ’3: Michelle has assumed:

valve-stem movement = valve opening; the seat is connected to the valve stem [9] the steam pressure from the utilities plant (and at the Battery Limits) is constant [10]

These are shown in Figure 6-6 as dotted lines 6.5.7

Step 7: Assess the Quality of the Reasoning

Consider the reasoning related to: i) the symptoms; ii) the symptom ‹ cause relationships; iii) the data gathered to test the hypothesis or probable causes; iv) the assumptions and v) the conclusion. Focus on the assumptions and on consistency. Use a chart of the arguments to guide the analysis. Consider each of the components in turn. i)

The symptoms. Were the symptoms correctly identified in Step 1? In the Case ’3, we have no reason now to alter the symptom. There is an assumption related to the symptom, and shown on Figure 6-5 as a dotted line: cycling in the level in the sight glass = cycling in the level in the bottoms =? cycling in the level in the reboiler. This was noted but not addressed. ii) The symptom ‹ cause relationships. Revisit Chapter 3 and see if any pertinent causes have been omitted. Also check that the diagram is consistent, based on the principles of Section 6.2.2. Recall that symptom ‹ cause evidence is where many mistakes are made because, although the cause ﬁ symptom connection is valid, the reverse (symptom ‹ cause) is not always true since multiple causes can display the same symptoms.

Example 6-30: Case ’3: reconsider Section 3.3, reboilers; Section 3.4 distillation and Section 3.5 steam traps. Section 3.3 reboilers: general: “Cycling (30 s–several minutes duration) steam flow, cycling pressure on the process side and, for columns, cycling Dp and cycling level in bottoms”: instrument fault/ condensate in instrument sensing lines/ surging/ [ foaming]* in kettle and thermosy-

6.5 Thinking Skill: Reasoning

phon/ liquid maldistribution/ steam-trap problems, see Section 3.5, with orifice Dp across trap < design/ temperature sensor at the feed zone in a distillation column/ collapsed tray in a distillation column. [ foaming]*: surfactants present/ surface tension positive system/ operating too close to the critical temperature and pressure of the species/ dirt and corrosion solids. Section 3.3 reboiler: thermosyphon: “Cycling (30 s–several minutes duration) steam flow, cycling pressure on the process side and, for columns, cycling Dp and cycling level in bottoms”: in addition to general, all natural circulation systems are prone to surging/ feed contains high w/w% of high boilers/ vaporization-induced fouling/ constriction in the vapor line to the distillation column. For horizontal thermosyphon: maldistribution of fluid temperature and liquid. Section 3.4 distillation: Only lists cycling temperatures: “Cycling of column temperatures: “ controller fault/ collapsed tray/[damaged trays]*/ [ jet flooding]*/ [ foaming]*/ [downcomer flooding]*/ [dry trays]* [ jet flooding]*: excess loading/ fouled trays/ plugged holes in tray/ restricted transfer area/ poor vapor distribution/ wrong introduction of feed fluid/ [ foaming]*/ feed temperature too low/ high boilup/ entrainment of liquid because of excessive vapor velocity through the trays/water in a hydrocarbon column; [downcomer flooding]*: excessive liquid load/ restrictions/ inward leaking of vapor into downcomer/ wrong feed introduction/ poor design of downcomers on bottom trays/ unsealed downcomers/ [ foaming]*; [ foaming]*: surfactants present/ surface tension positive system/ operating too close to the critical temperature and pressure of the species/ dirt and corrosion solids. [Dry trays]*: flooded above/ insufficient reflux/ low feedrate/ high boilup / feed temperature too high. [Damaged trays]*: leak of water into high molar mass process fluid/ large slugs of water from leaking condensers or steam reboilers/ startup with level in bottoms > design/ attempt to overcome flooding by pumping out bottoms at high rate/ too rapid a depressurization of column/ unexpected change in phase. Section 3.5 steam traps. No specific entries for “cycling” “No condensate discharge”: bypass open or leaking/ scale in the orifice/ plugged strainer/ inlet pressure too high/ for inverted bucket trap, bucket vent clogged, incorrect Dp across the orifice. Comment: The focus in Figure 6-5 was on the causes related primarily with the control system. The bold items listed above do not seem to be explicitly included in the chart. Scrutiny of these suggests that no major issues have been left out considering Michelle’s purpose in creating this chart. iii) The data gathered to test the hypothesis or probable causes. Check the evidence from the questions and tests that were gathered. The general options and the biases and interpretation issues were considered in Section 6.1 and analyzed in Step 5, Section 6.5.5.

231

232

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

Example 6-31: Case ’3 From Figure 6-6 we note that Michelle focused on evidence [2], [3] and [4]. She inferred [6] and [7]. She should have explicitly checked this by setting the flow= constant; noting if the valve stem was constant and then checked for cycling [5]. Because Michelle did not do this explicitly, conclusion [8] is not proven. What Michelle should have concluded from evidence [3] and [4] is that the signal to the steam valve works. If she had shown that the steam flow is constant and yet the level cycles [5], then she could conclude that the steam control system is not at fault [1].

iv) The assumptions. Focus on the assumptions and on consistency. Example 6-32: Case ’3 The two assumptions Michelle noted were [9] and [10]. Assumption [9] would have been easy to check.

v)

The conclusion. Strong evidence should support the conclusion; or a large collection of weak to moderate evidence. There should be no strong counterarguments nor strong assumptions.

6.5.8

Step 8: Assess the Strengths of the Counterarguments

Use What if? counterexamples to help clarify the reasoning. These can be the reverse of the assumptions. 6.5.9

Step 9: Evaluate the Consequences and Implications

Once we have accepted a valid conclusion, explore the So Whats? In most troubleshooting situations this means identify another feasible cause or, once the cause has been identified, to suggest remedial or corrective action. 6.5.10

Activity 6-14

Critique the arguments described in one of the Cases presented in Chapter 4.

6.7 Summary

6.6

Feedback and Self-Assessment Reflections:

From this chapter, what have you discovered about trouble shooting? about planning and performing tests? about asking questions? about reasoning and drawing conclusions? ______________________________________________________________________ ______________________________________________________________________ ______________________________________________________________________ ______________________________________________________________________ Rate yourself on the following

already do this

might work not for me for me

Diagnostic actions – systematic and structured pattern to obtain information – use criteria to select tests – unique style and potential biases Consistency – separate fact versus opinion – cause–symptom data – symptom–cause relationships – identify root cause versus “symptom” – rules for English, mathematics, science fundamentals Classification – use single basis/criteria per level – apply classification principles to identify symptoms Patterns – methods to identify patterns Reasoning – systematically use a process similar to the 9-step – draw/construct symptom ‹ cause diagrams – diagram an argument and use this to critique – well-developed methods to identify assumptions

&

&

&

& & &

& & &

& & &

& & & & &

& & & & &

& & & & &

& &

& &

& &

&

&

&

& & &

& & &

& & &

6.7

Summary

A variety of critical-thinking skills are used when we trouble shoot. The first criticalthinking skill relates to how we gather information about the process. As we select diagnostic actions to gather information and evidence about the system we usually start with general background information (although if we are familiar with the process we know this already). Then we build up our mental image of the system and context by determining what happened recently, reminding ourselves of the details of what should be happening and then exploring what really is happening.

233

234

6 Polishing Your Skills: Gathering Data and the Critical-Thinking Process

In testing hypotheses/causes we should start simply and test the most-likely cause early. Above all we should refrain from the expensive tactic of shutting down, opening and inspecting. Much crucial information can be gained from the instruments, simple calculations and sound reasoning. A range of biases and mistakes made in gathering and interpreting data are listed: confirmation bias, over-interpretation, under-interpretation and mis-interpretation, availability bias, premature closure, anchoring. Jungian typology (or MBTI) dimension P-J provides a useful indicator of one characteristic of your personal style for gathering and interpreting data. A second critical-thinking skill relates to the importance of being consistent in our use of words, and our use of knowledge about processes and process equipment. Specifically the focus was on identifying fact versus opinion, gathering accurate cause–symptom data and astutely reversing the connection to link symptom–cause. Our usage should be consistent with the rules of English, mathematics, the fundamentals of science and engineering and practical experience. A third critical-thinking skill is classification; the process of dividing large sets of information into meaningful parts. In making the division we should use a single basis/criteria per level, use no single entries and avoid faulty coordination or subordination. This was illustrated for the task of classifying the starting information into “symptoms” and “triggering” events. A fourth critical-thinking skill is to be able to identify patterns. A final critical-thinking skill discussed here was reasoning. A systematic 9-step process is suggested and illustrated in the context of part of Michelle’s reasoning in Case ’3.

6.8

Exercises

1.

2.

Harry lives on the eleventh floor of an apartment building. Each morning he rides the automatic elevator down to the first floor and goes to his specialized job in the cramped quarters assembling components in the airplane fuselage. Each evening he rides the elevator up to the seventh floor and then walks up to his own floor. Why? Mr Tom Jones goes to the doctor’s office twice a week to pick up his pills. This particular Tuesday his wife decides to go with him so that she can continue on and do the shopping afterwards. Tom parks the car in front of the doctor’s office, leaves his wife in the car and goes in to get the pills. While he is talking with the doctor, he hears a terrific crash, rushes to the window and upon seeing his car completely demolished exclaims “My wife’s been killed!” whereupon the doctor reaches into the desk drawer, pull out a gun and shoots Tom. Why?

6.8 Exercises

3.

Check out the mathematics in the following:

2

þ 1 þ 1 þ 1 þ ...... þ 1Þ n ¼ nð 1|ﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ{zﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ} n 2 þ n þ n þ ﬄn{zﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ þ . . . . . . þ ﬄn} Þ n ¼ ðn |ﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ n dn2 d ¼ ð |ﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ{zﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ} n þ n þ n þ n þ n ...... þ nÞ dn dn n 2n ¼ ð 1|ﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ þ 1 þ 1 þﬄ{zﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄﬄ 1 þ . . . . . . þ ﬄ1} Þ n 2n= n 2=1

235

237

7

Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance In any problem, people are involved. You as trouble shooter are the major person we might focus on. Yet there are others: the operators, the maintenance team, other team members, the designers, your supervisors and your colleagues. Interpersonal skills are needed as you work with others. These include skill in communication, listening, building and maintaining trust and building on personal uniqueness. Furthermore, some factors affect performance: our performance and that of others. These factors include pride and willingness to risk admitting a mistake, stress and distress, and the environment. Consider interpersonal skills and then factors that affect performance.

7.1

Interpersonal Skills

Included here are skills in communication, listening, applying the fundamentals, building trust and accounting for personal uniqueness. 7.1.1

Communication

Communication is speaking and writing that a) correctly identifies multiple audiences, answers their needs and questions; b) has content that includes evidence to support conclusions, c) is well organized with summary and advanced organizers, d) uses a style that is coherent and interesting, defines jargon or unfamiliar words, and e) includes a format that is grammatically correct and follows the expected format and style. These five elements should characterize every communication. In addition, in TS situations requests for changes in operation and for tests should be in writing and should consider the safety of the operator. Activity 7-1: Communication Critique the following request from the point of view of audience, content, organization, style and form. Then, if pertinent, rewrite the communication.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

238

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance

a.

b.

c.

A verbal request from engineer Jose to process operator, Marco. The context is that Jose is trouble shooting the process and believes that the cycling observed in the controller on the reflux relates to the changes in composition in the overhead. Jose is looking for a time variation. “Take about a dozen samples of the overhead and send them to the lab for a full analysis. Tell them it’s urgent.” In March, hourly data have been collected on the day shift over three weeks so that you can do a detailed analysis of the process operation. In April the process will be shut down for routine maintenance. Andre realized it was important to ensure that the meter readings collected over the test period were accurate. The last week of March Andre wrote a memo, not an e-mail, to the Pulp Mill engineer: “Would you please request that the instrument shop determine the calibrations on the six flowmeters during the April shutdown. Thanks.” “Record the temperatures at the top of the distillation column once every shift.” Fact. There is only a temperature indicator at the top of a 30-tray, selfstanding tower. Question: how many operators do you think actually climb the tower in the rain to read the instrument?

7.1.2

Listening

Listening includes focusing attention on the talker, avoiding distracting behaviors, showing respect and frequently acknowledging through appropriate body language and “ahums” and reflecting statements. The process can be modeled as Sensing, Interpreting, Evaluating and Responding or SIER. That is, we sense the message; we internally interpret what is being communicated; we evaluate the message in the context of the situation, our feelings, needs and goals; and we select how to respond. Here’s what we know about listening: .

. .

.

.

sensing the message is complex because about 55% of the message is communicated by body language, 38% by tone and 7% by the words; listening is about four times slower than thinking; about 80% of our waking hours are spent in verbal communication; with about 12 of that spent listening; untrained listeners understand and retain between 25–50% of a conversation; only about 5% self-assess themselves as being highly skilled listeners.

In TS, we usually need to gather key information from others. We ask them questions; we listen to their answers. Trouble shooters need to encourage the person to communicate clearly, to listen carefully and to interpret the answer correctly. Three listening skills to aid this process are: attending, following/tracking and comprehension checking/reflecting. Here are the details.

7.1 Interpersonal Skills

Attending: posture is inclined forward and open, facing squarely approximately 1 m apart; no distracting behavior and eye contact is called “soft focus” (contrasted with looking away or staring). Tracking/following: provides minimal encouragement (for example, “Tell me more”, “sure..” “Oh..”, “ Then ..”) and infrequent questions (for example, prefer “What?” questions to “Why?”) and attentive silence. Reflecting is responding with a concise restatement of the content and feelings expressed in the listener’s own words. That is, include the content and feelings of what was said, express it in the listener’s own words, do not add new ideas, do not leave out ideas. Some example approaches include saying “As I understand it..” or “Are you saying that ..” Reflecting is usually used when someone is very emotional, or when you see differences developing between you and the other person, when there is disagreement, when the talker seems to be confused or when the talker needs encouragement that his/her contribution is valuable. Activity 7-2: Listening Assess the quality of the listening by the engineer in the following situation. Note the five strengths and the two areas to work on. Engineer Ahmed goes out to the control room for Case ’7, the case of the reluctant crystallizer. Upon entering the control room Engineer Ahmed says, “Hi Phil. I’m new here. I understand from Frank that you’re a great plant operator. I hear that we’ve got trouble on this VC. Let’s see, that gauge is vibrating all over the place, so is that one. Hey, this is a zoo. Don’t worry I’ll solve it soon.” Activity 7-3: Reflecting For Case ’7, operator Phil said “When the liquid level started to drop I went up and listened to the booster ejector and it sounded fine. The only thing I noticed was that the pressure gauge on the bay water line to the booster condenser was fluctuating wildly.” Which of the following might you use to show Phil that you are listening to him?

a. b. c.

d. e. f.

“Ahum” “Ok, please continue.” “As I understand it, the booster ejector sounded as you expected when the liquid level started to drop and the pressure gauge on the bay water line really fluctuated. Is that correct?” “Why is that noteworthy, Phil?” “I’m listening” Other

7.1.3

Fundamentals of Interaction

The fundamentals of interaction are summed up in the seven RIGHTS and the four destroyers. Consider each in turn.

239

240

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance .

Claim and honor the seven fundamental rights of individuals, RIGHTS (Woods, 1994) R the right to Respect I the right to Inform or to have an opinion and express it G the right to have Goals and needs and express these. H the right to Have feelings and express them. T the right to have had Trouble and make mistakes and be forgiven. S the right to Select your response to other’s expectations. and the right to Claim these rights and honor these in others.

.

Avoid the four destroyers of relationships (Woods, 1994): Contempt Criticism Withdrawal and stonewalling Defensiveness.

I remember these by recalling the phrase “Did the contemptuous critter sit on de fence or the stone wall?” Activity 7-4: Fundamental rights and destroyers Analyze the conversation and identify claiming of rights, honoring rights in others and evidence of the four destroyers. Two engineers Tonya and Marcos are trouble shooting Case ’8: the depropanizer: the temperatures go crazy. Let’s listen in on their conversation. The parts of the conversation are coded to guide discussion. Tanya: “OK, the six hypotheses are (i) 1. tray collapsed in the stripping section, 2. too much bottoms fed to the debutanizer, 3. too much overheads in the feed, 4. feed valve FV1 stuck, 5. pump F-26 not working, 6. not enough feed to the column (ii). What do you think? (iii) Marcos: “It’s got to be a collapsed tray. (iv) I encountered something like that on the S256 plant last year. Same evidence. It’s got to be a tray. (v)” Tanya: “Hey! (vi) You’ve done it. (vii) You’ve zeroed in on one hypothesis when we need to keep an open mind and do some simple checks first. (viii)” Marcos: “Don’t get huffy about it. (ix) I’m just trying to get this solved fast. (x) My experience tells me it’s the tray! So what great insight are you bringing to this problem? (xi) Ms. Smarty. (xii) Besides your hypothesis ’6 is a symptom, and not a root cause. (xiii)” Tanya: “Now you resort to criticizing my hypotheses. (xiv) That’s not fair (xv)” 7.1.4

Trust

Build and maintain trust. Trust glues relationships together. Trust is based on integrity, competency and benevolence (not doing anything that will hurt the other purpose).

7.1 Interpersonal Skills .

We build trust by such acts of integrity as: – keeping commitments to yourself and others. – clarifying expectations that you have of yourself and of others. – showing personal integrity, honesty and loyalty to others. – promptly and sincerely apologizing when you know you are wrong. – honoring the fundamental RIGHTS listed above and avoiding the destroyers. – taking time to see things from the perspectives of others. – accepting others “warts and all.” and by such benevolent acts as: – not saying ill of the person behind his/her back or when they are not present.

.

We destroy trust by – the reverse of the Builders of trust listed above, and – not meeting commitments. – selectively listening, reading and using material out of context. – not accepting the experience of others as being valid. – asking others to give up their fundamental RIGHTS. and such non-benevolent acts as: – making changes that affect others without consultation. – playing the broken record until youve eventually worn them out. – subtly making changes in the context/issues/wording gradually so that they are unaware of what is happening until it is too late. They were sideswiped.

Activity 7-5: Trust Someone requests “Would you please be chair of the upcoming conference. It won’t take much work and you are the ideal person to do it”. Your situation is that you have been chair of a similar conference before; this would take the equivalent of at least 2 months of concerted effort. You have promised your family to spend more time with them. You are just barely managing to meet your commitments now. The conference will draw many from abroad and being chair would bring you a lot of personal satisfaction as well as increase your visibility and reputation. How do you respond? Activity 7-6: Self-assess trust Complete the inventory about trust given in Worksheet 7-1.

241

242

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance

Worksheet 7-1:

Trust.

Trust is having the confidence that you can mutually reveal aspects of yourselfand your work without fear of reprisals, embarrassment or publicity. Trust works both ways: you trust them and they trust you. Trust is not developed overnight, trust takes time to develop. Trust can be destroyed by one incorrect act. Check your current status Building your trustworthiness getting them to trust you

already do this

needs some work

need lots unsure of work if this is for me

1. Do what you say you will do. 2. Be willing to self-disclose: don’t hide your shortcomings; share yourself-honestly. 3. Listen carefully to others and reflect to validate your interpretation. 4. Understand what really matters to others; do your best to look out for their best interests. 5. Ask for feedback. 6. Don’t push others to trust you more than you trust them. 7. Don’t confuse “Being a buddy” with trustworthiness. 8. Tell the truth. 9. Keep confidences. 10. Honor and claim the 7 RIGHTS. 11. Don’t embarrass them.

& &

& &

& &

& &

&

&

&

&

&

&

&

&

& &

& &

& &

& &

&

&

&

&

& & & &

& & & &

& & & &

& & & &

Checking your trustworthiness do they trust you? always

most times

sometimes

don’t think applies

1.

&

&

&

&

&

&

&

&

& &

& &

& &

& &

& &

& &

& &

& &

2. 3. 4. 5. 6.

Do they disclose confidential information trusting that you will keep it confidential? Do they assign you challenging tasks to do without frequently checking up on you? Do they honor your RIGHTS? Do they seem to look out for your best interests? Honest and forthright. Do not leave you feeling that they haven’t told you everything about the situation; they seem to be holding back. copyright 1999 , Donald R. Woods

7.1 Interpersonal Skills

7.1.5

Building on Another’s Personal Uniqueness .

The Unique You and the Unique Them. Each of us has our biases, prejudices and preferences or style. A variety of questionnaires and inventories can be used to help understand preferred styles for managing conflict, making decisions, applying creativity, differing style of conversing, validating ideas, gathering and using data, accounting for facts versus feelings and considering details versus the big picture. Inventories of Johnson and Johnson (1986), Kirton (KAI: 1976), and Jung (MBTI: 1984) are examples of such inventories. Your style is unique; it will differ from others. Accept, respect and improve the quality of your interaction through these differences. Do not let these differences lead to conflict. Some applications of this are given in Section 6.1.3.3.

Activity 7-7: The unique you You are on a trouble-shooting team with the following people whose personal style are given in the following table. Write in your scores. Johnson style for conflict

You Marie Phil Jean Terry

1.

Jungian

Kirton

Withd

Accom

Force

Comp Negotiate IE

TF

PJ

SN

–1 –6.1 –1.1 –4.1

2 1.3 3.7 3.5

–3.5 –5.8 –3.1 –6.4

3.4 0.6 2.4 1.2

T F T F

J P J P

S S S N

9 7.5 6.7 7.8

I E E E

82 98 87 83

4.

Where are your blind spots? Describe this in actions. For example, if you are a dominant S then your blind spot might be seeing the big picture, focusing too much on the details. Does the team have any blind spots? With whom might you have minimum differences? What are those differences? With whom will you have maximum differences? What are those differences?

5.

How can you build on this to trouble shoot efficiently and well.

2. 3.

We should also recognize the personal tendency or bias to prefer to report interpretation and inferences, instead of “just the facts”. This is discussed in Section 7.2.5.

243

244

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance

7.2

Factors that Affect Personal Performance

If an operator makes a mistake, the type of mistake is likely to be: 90% 5%

5%

no action taken (when some kind of action was needed), took corrective action but moved the correct and appropriate variable in the wrong direction. Thus, he/she knew that the temperature should be changed but increased it instead of decreased it. took corrective action on the wrong variable. Thus, he/she should have changed the temperature but, instead, changed the composition to the reactor.

Why do operators tend to take no action when a clearly defined action is needed? Such factors as inability to admit error, stress, alienation and lack of motivation, tendencies to infer and an “I know best” attitude all affect anyone’s ability to perform a task. These factors can affect the operators and people with whom we must interact, and these factors can affect us. 7.2.1

Pride and Unwillingness to Admit Error

Each person has his/her own pride and self-esteem to keep intact. We do not want to admit: . . .

that we are wrong; that we made a mistake; that we are guilty of wrongdoing.

People want to behave so that they “look good” in the eyes of those who matter... their spouse, their colleagues, their supervisors, themselves. Not all evidence may be presented because some mistakes may have been made or we are embarrassed and do not want to show we have made a mistake. Example 7-1: Marg, the design engineer, is now part of the startup team. She is convinced that she designed the reactor correctly. It may be difficult for her to work with data that suggests the design is wrong. Example 7-2: An operator, while making a preliminary inspection, was to observe and not touch. However, he opens and closes a valve. However, the valve does not go as far shut as he thinks it should. Example 7-3: For a working process, an operator accidentally drops a mercury thermometer into the receiving and blend tank. The mercury will be dispersed throughout the feed

7.2 Factors that Affect Personal Performance

but “the concentration will be very low” and the “glass should be all broken up anyway.” In a couple of weeks the catalyst activity seems to fall off. Example 7-4: The new regulation for energy conservation requires that all bypass valves on steam traps be closed. Bill strongly disagrees with this policy and cracks open the bypass valve. He thinks the process works better that way. Later it is found that the temperature control on the reboiler seems to be inadequate. Example 7-5: Engineers may visit the plant site, but they may not touch or adjust things. Peter, on a recent visit to the site, notices a small leak around a valve stem. Believing that sometimes a leak can be stopped by turning the stem slightly and then returning it to its former position, he tries this. However, the valve jams after he turns it from half open to quarter open. He cannot get it to return to half open. He leaves and returns to his office. A few minutes later he is called out to the plant because there is trouble on the plant. 7.2.2

Stress: Low and High Stress Errors

We encounter stress as an accumulation of events from home, work and play. The work environment also provides daily stress: through the amount and type of interruptions, the noise and cleanliness of the environment and the complexity of the tasks being done. The amount of stress one has experienced affects our performance. If there is not much stress or there is too much stress in our life, we are prone to make mistakes. Powers and Lapp (1983) and Kletz (1986) summarize the probability of different types of operator error to occur. For tasks involving sensing – interpreting – acting, if the person is well trained, and motivated and with no stress then, .

.

he/she will make about 1 error in 1000 trials if feedback is given to the person after they have made the action; he/she will make about 1 error in 100 trials when there is no feedback to the person for the action taken.

if the person is under distress – not because of high stress levels cumulating throughout the year – but because of the situation, then .

.

he/she will make about 1 error in 10 trials. This might happen in a busy operating center where other alarms are sounding, the telephone is ringing and people are asking for information about a part of the process. he/she will make about 1 error in 2 trials if, for example, many complex actions are required and the implications if an error is made are frightening.

245

246

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance

The distress comes from poor training, confusion because of poor training, conflicting data; from the need for fast action; or from a large penalty if a mistake is made, or from extensive confusing and contradictory types of demands. The training should be competency based and oriented to develop skill, and not just to develop knowledge. Training is particularly important when new technology is introduced, for infrequent events and for new people. When we TS, it is wise to monitor our own stress and to be sensitive to the stress experienced by others. Activity 7-8: Stress and control One common cause of stress is that we worry about – and get angry over – things over which we have no control. Research evidence suggests that we have control over only 1 out of 10 things we might worry over. Trouble shooting can be particularly stressful. You are under a lot of pressure to find the root cause and correct it safely and quickly. You are trouble-shooting case ’12 given in the list below. Which of the following do you have control over?

The problem given in Case ’12 arises at 8:30 am. The following background and resources are available. Identify which ones you think are directly under your control. If you think you have control over the item, circle Y; if you do not have control over it, circle N.

Worksheet 7-2:

1. The safety officer, Hack, is overbearing, not liked and gets carried away about simple things. 2. The laboratory can analyze liquid samples with their equipment but gaseous samples cannot be analyzed because their instrument is broken. 3. The lab schedule is busy with top priority analyses. Samples could not be analyzed until after 3:00 pm. 4. The union prevents you from analyzing any samples; if you do, there will be a strike. 5. The upstream styrene plant is operating at 80% capacity. 6. The upstream ethylene cracking furnace is operating at 95% capacity. 7. The upstream propylene plant is shut down. 8. The operator on the ethylene plant is cooperative. 9. Samples can be taken at any of the sewer gates within the Battery Limits and at A, B, and C. 10. No blueprints are available for the sewer system. 11. Your performance review to establish your salary is being done next Thursday at 4:30 pm. 12. You have to prepare your “record of progress” record for the performance review on Thursday.

Y

N

Y

N

Y

N

Y Y

N N

Y Y Y

N N N

Y Y

N N

Y

N

Y

N

7.2 Factors that Affect Personal Performance

Activity 7-9: Stress and self-talk A common contributor to your stress is your self-talk. Too many people say to themselves “You are stupid” “You can’t do this”. Make a list of any negative self-talk comments you say to yourself-during a week. Create an action plan to minimize your negative self-talk and to maximize your positive self-talk.

Safety on the ethylene plant (courtesy of John Gates, B. Eng. 1968, McMaster University) The part of the ethylene plant that relates to this problem concerns the drying section to remove moisture from the feed gas and the distillation train to separate the gas stream into the desired component section. Drying section: Three alumina dryers are installed. One dryer is regenerated while two are hooked in series on stream. For example, dryer V106 is being regenerated with V107 and 108 removing the moisture in the process stream to less than 4 ppm. Then V107 will be regenerated with V108 and 106 in series and so on. The cycle lasts 12 hours. The dried gas goes to a knockout pot (to remove any entrained material) and then is chilled, in exchangers E107 and 108, before entering the separation towers. During regeneration of the dryers, “fuel gas”, heated with 2.8 MPa steam, flows through the dryer in the direction reverse to normal flow. Once it is through the dryer the regeneration effluent gas is cooled and returned to the fuel-gas system. During regeneration the dryer temperature rises to 190 C and is maintained at this temperature for one hour. Then the fuel gas bypasses the heater and is sent directly to the dryer to cool it. For this plant, the “fuel gas” or regeneration gas is the noncondensable overhead from column T 101, the demethanizer. The dryers are all appropriately manifolded and valved so that any dryer can be regenerated, bypassed or used. The regenerating dryer is separated from the line dryers by gate valves. The separation is performed in a train of three distillation columns operating at about 3.2 MPa. These are a demethanizer, de-ethanizer and C2 splitter, T101, T102 and T103, respectively. The process is illustrated in Figure 7-1. The overhead from Tower T101 is condensed with ethylene as refrigerant. The overhead from Tower T102 is condensed with propane as refrigerant. The overhead from Tower T103 is condensed with propylene as the refrigerant. The safety inspector telephones to say that this morning’s gas samples from the sewer drop boxes within our battery limits (for ethylene distillation units) are explosive. These sewers serve other plants: styrene, ethylene furnaces and propylene plants. The plot plan of the unit is shown in Figure 7-2. Clear up this matter immediately; this hazard cannot be permitted! Case ’12

247

248

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance

Figure 7-1

The ethylene plant for Case ’12.

A

B

C

from Styrene, ethylene & propylene

To disposal

T101

T102

T103

Sewer gates

V106

V107

V108

E130 Battery limits E131

Figure 7-2

Plot plan of the ethylene plant for Case ’12.

7.2 Factors that Affect Personal Performance

7.2.3

Alienation and Lack of Motivation

The working environment should motivate employees. Employees should have a clear idea of the expectations, and rewards should be forthcoming for achieving those expectations. People are demoralized if people considered to be “stars” are always given the challenging work, if information sessions about the company are not given and if the promotion and salary adjustments are unclear or seem to be based on who you know and not on performance. People are alienated when a) tasks lack imagination or are to be performed under strained conditions, b) where the working conditions are sloppy maintenance, extremes in temperature, odors and dust and hazards, c) where the emphasis is on machines ... and not on people, and d) when management decisions lack consistency, predictability and transparency. 7.2.4

“I Know Best!” Attitude

In other situations, sometimes, people deliberately fail to follow instructions because “they know better”. An operator opens the bypass valve on the steam trap or overrides a control system because “those silly systems don’t work! I know best.” This attitude is often related to increased alienation and lack of motivation. 7.2.5

Tendency to Interpret

Instead of saying “the gauge reads ...” we tend to say “the temperature is ...” It is vital that a trouble shooter listens carefully to what is said and how he/she interprets the information. Communicating just the facts is boring. We want to describe what we think is wrong. Example 7-6: The designer might say “the rogue data point for the heat capacity of acetic acid vapor, shown in Figure 7-3, is probably caused because the instruments or the researcher made a mistake" and proceed to design the preheater assuming the heat capacity at 200 C is about 0.45 CHU/lb C. Comment: the rogue point is correct.

249

250

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance

Figure 7-3

Heat capacity of acetic acid vapor.

Example 7-7: The operator might say that the valve on the bypass is misbehaving again; see the instabilities in the flow control of the reactor feed. Is it really the valve? Is there really an instability? Example 7-8: The engineer might suggest that the reason the low yield for the reformer, compared with the expected yield, is caused by low catalyst activity. Example 7-9: The engineer says that the pressure in the line is 50 kPa; the temperature is 450 C. Comment: the engineer should have said the pressure gauge reads 50 kPa; the thermocouple reads 450 C. Activity 7-10: Triad talking In groups of three, one is the “talker”, one the “fact-summarizer” and the other is the “inferrer”. The talker describes, for three minutes, his/her greatest frustration, favorite activity or hobby. The fact-summarizer takes one minute to summarize the facts given by the talker. No personal opinions or inferences should be given. The inferrer describes in one minute what inferences and messages came through from the ideas presented, by the body language and by what was not said. Rotate responsibilities so that each has a chance to play all three roles. This activity takes 15 minutes total. At the end, share your experiences.

7.2 Factors that Affect Personal Performance

Pair talking and “just the facts” In pairs, with one being the talker and the other the listener. The talker reads over Case ’13 “The Lousy Control System” to himself/herself-for about five minutes. Then, referring only to the scenario for details the talker has three minutes to describe the facts in the Case to the listener. Do not simply read the scenario to the listener. You may show your listener the sketch of the process but you cannot show the written material. You have to describe the facts. Change roles and repeat with Case ’14, “The condenser that was just too big.” After each role, take two minutes and write out your reflections about the activity. Discuss with your partner. Activity 7-11:

The Lousy Control System (courtesy Esso Chemicals) All the texts on process control say that controlling the overhead condensate exit temperature by varying the fan pitch is a slow and clumsy method of control. Yet, that is the method used on column T6. Today it is raining, the temperature of the condensate is 3 C subcooled. Yesterday it was hot and sunny. The fans were running flat out but still the gas exit valve controlling the pressure was open half-way most of the day. All the uncondensed gas went to flare. The boss storms in “You’ve got trouble on this plant; too much stuff went to flare yesterday. What’s wrong?” You’re sure that the control system is lousy. The system is shown in Figure 7-4. Case ’13

PSV 1

PC 100

TC 200 PI 201

LC 202

Figure 7-4

The control system for Case ’13.

TO FLARE

251

252

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance

Case ’14 The Condenser That was Just too Big Fatty acids are solid at room temperature. Fatty acids are purified in a distillation column operating under a very high vacuum. The “coolant” in the overhead condenser cannot be water because the fatty acids would solidify in the condenser. Instead the coolant is boiling water with the “coolant” temperature controlled by the pressure at which the water boils. This plant is started up for the first time. A common fault in startup situations is that the condensers are over-designed for “clean conditions” because fouling factors have been used in the selection of the overall heattransfer coefficient. This would mean that at startup the condensing fatty acids would be subcooled. When you go out on the plant one hour after startup, the worried operators say “Look at the overhead pressure gauge – it’s much too high. It reads 2.7 kPa (20 mm Hg absolute) when it should be reading 0.4–0.7 kPa (3–5 mm Hg).” Just as you expected! the fatty acids are being subcooled in the condensers. Solid is starting to build up in the tubes. The excessive pressure drop across the tubes means that the vacuum system can’t get rid of the air leaks fast enough and the pressure builds up. You pull out the drawings showing the configuration of the plant (Figure 7-5). This problem needs to be solved fast! PI

Cooling water

Cooling water return

well water

steam

0.4 kPa PI

Booster

Cooling water exit

PI

Boiling Condenser

240 kPa

wet vacuum pump

TI

Backup Condenser barometric condenser

TI

240ºC to 260ºC

TI

120ºC Coil in tank

60ºC TI

Fatty acid Feed to/from Dowtherm system

Figure 7-5

Cooling Water

reciprocating pump

Fatty-acid distillation column for Case ’14.

7.5 Exercises and Activities

7.3

The Environment

The environment affects performance. If making a mistake is considered the worst possible thing to do; then no one will admit to making a mistake. If mistakes are accepted, understood and no serious repercussions occur, then people will be more ready to admit their mistakes. Rate the environment Use Worksheet 7-3 to help reflect on the environment and its impact this might have on the people with whom we interact.

Activity 7-12:

7.4

Summary

We interact with people when we trouble shoot. Our skill in trouble shooting often depends on our skill in communicating, listening, applying the basic fundamental principles of interpersonal relationships, building trust and understanding our own uniqueness and the uniqueness of others. These ideas were illustrated and activities provided to give you a chance to work with these skills. We also need to be aware of the factors that affect our performance and the performance of others. These include unwillingness to admit having made a mistake, and the impact that stress has on our ability to work effectively. Sometimes people are frustrated and unmotivated. Sometimes people follow their own approach even though it may result in incorrect and even unsafe operation of the process. We also prefer to infer and interpret what we see or hear. Reporting the facts is usually boring. So we read a gauge and say “The pressure is 1.4 MPa” when in reality we should have said “The pressure gauge reads 1.4 MPa.” A questionnaire is included to give us a chance to evaluate the environment in which we trouble shoot.

7.5

Exercises and Activities

1.

2.

Consider ten cases in Kletz’s text “What went wrong?” From Kletz’s description of the case, classify the cause of the equipment malfunction, human design mistake, human maintenance mistake, human operator mistake or other. Typical feedback from industrial workshops on trouble shooting.

For Activity 7-10: Fact-summarizing was either easy or difficult depending on the quality of the talker’s presentation. Taking notes helped. Inferring was most enjoyable. Compare your experience with this.

253

254

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance

For Activity 7-11: Giving facts was difficult. Take my time. Repeating was difficult. Write everything down. Compare your experience with this. 3.

Cases to consider. For Cases ’15, 16, 17 and 18 create hypotheses as to the cause.

Worksheet 7-3:

Feedback about your environment.

To what extent do you agree with the following descriptors of your environment where you usually “trouble shoot”. People are willing to admit error: “The people that I work with are very unwilling to admit errors; they blame others, they pass the buck and, if necessary, would purposely mislead me rather than to admit error.” Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________ Encourage risk taking: “Risking is rewarded. We are expected to take risks about 10 times a day. Risks should be wisely, not indiscriminately selected. But, nevertheless, we are not only encouraged but we are rewarded for risk taking.” Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________ General stress at work they are under: “Their environment is very stressful. People have many deadlines and interruptions. The consequences of making mistakes is very high. The issues are complex. The environment changes often and includes a lot of uncertainty.” Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________

7.5 Exercises and Activities

General stress at work you are under: “My environment is not stressful. I do not have many deadlines or interruptions. The consequences of making mistakes are low. The issues are straightforward. The environment is safe, stable and secure.” Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________ People’s listening and responding: “The people are open, communicate well, can clearly identify facts, will offer opinion when asked for, and are very competent but are not aggressive “know-it-alls””. Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________

The flooded boot (problem supplied by W.K. Taylor, B. Eng. McMaster 1966) As part of our energy conservation program we recycle condensate to our boiler house. This condensate comes from large, steam-driven turbines whose exhaust conditions are full vacuum. These turbines drive the feed gas, air, refrigeration and synthesis gas compressors on the reformer section of the ammonia plant. The exhaust steam goes to the shell side of a surface condenser. Chilled water is on the tube side. Vacuum is maintained by a two-stage steam ejector with inter and after condensers. These are located downstream of the surface condenser and pull a vacuum on the condenser itself. Lately we have been increasing the load on the compressors and hence the steam consumption has increased. However, for these new conditions we can’t seem to get rid of the condensate from the boot at the bottom of the surface condenser. We sometimes have to run the spare pump 122.JA (electric drive) because the turbinedriven pump 122.J cannot handle all the flow. “These high levels of condensate and the unusual operation with our spare pump are keeping my fertilizer production down. It’s got to stop! My production rate is barely reaching 120 Mg/d and I think we can easily produce 132 Mg/d. That’s $5000 a day. Fix this bottleneck!”. The system is given in Figure 7-6.

Case ’15

255

7 Polishing Your Skills: Interpersonal Skills and Factors Affecting Personal Performance

256

EXHAUST FROM VARIOUS TURBINE DRIVES

REFRIGERANT COMPRESSOR

GAS COMPRESSOR

TO VACUUM SYSTEM

SURFACE CONDENSER 2" Bypass

CHILLED WATER

LIC

BOOT LC-2A

6"

3/4" 2"

FI

SEALING WATER

PI

6"

ION EXCHNG PI

1/4"

3"

PI

4"

LC-2

2"

6"

3/4"

122JA

4"

6" 122-J

Figure 7-6

The configuration of the flooded boot for Case ’17.

Case ’16 The case of the dirty vacuum gas oil (problem supplied by R.J. Farrell, B. Eng. McMaster University, 1974) Word spread quickly. “There’s trouble on the vacuum tower!” Normally the side product, vacuum gas oil, is water clear. Today the product has been off-specification in color. Indeed “It looks brown” explained a frustrated supervisor. The feed to the vacuum tower is crude oil. The crude oil is heated by a series of three exchangers and the furnace. The first two exchangers extract heat from the vacuum gas oil (API 27.8), and from the mid-side-draw cut that has an API gravity of 25.4. The third exchanger extracts heat from the tower bottoms (of API gravity 11.5). The column has pump-around draw-off for the bottoms, to recycle to crude, to mid-side, and to vacuum oil-overhead condensates. The mid-side and vacuum gas oil are blended and cooled to yield a product of API 26.2. The blend is sampled after the cooler; the blend is off-specification. Fix the problem. Case ’17 Is it hot or is it not? (case supplied by Jonathan Yip, B. Eng. 1998, McMaster) The crude fatty-acid high-vacuum fractionation column is having trouble. The bottoms recirculation and transfer pumps need maintenance immediately. To cope with this the engineer placed the column on safe-park. The feed was stopped and the sump and overheads recycled back to the crude fatty-acid tank. The steam ejectors were turned off and all the valves were shut to keep the column under vacuum.

7.5 Exercises and Activities

The pump was dismantled for repair and an hour into the repair job the plant operator noticed that the temperature gauge at the top of the column was showing an increase in temperature. Indeed, the temperature alarms started screaming. “That’s crazy!” said the operator. “There’s nothing going on in the column. The column is isolated. The column is not running. That temperature should be decreasing instead of increasing. That temperature gauge and those alarms have been temperamental for a long time.” He manually suppressed the nitrogen purge valve and cut off the alarms so that he could hear himself-think. The streptomycin dilemma In our process for the production of streptomycin we evaporate the eluate in tall rising film evaporators. These are fixed tube sheet with a shell expansion bellows to allow for the thermal expansion. Our plant operates on a single, eight-hour shift so that after each shift the evaporator is shut down. However, because we process pharmaceuticals we must clean and sterilize all units after each shift. The procedure is to open the end of the bundle, to brush the inside of all tubes, rinse, fill the tubes with water and then apply steam to the shell to boil and thereby sterilize the tubes in preparation for the next day’s operation. The plant has operated without a hitch for the past four years. However, over the past months the tubes are starting to fail; leaks are developing at the tube/tube sheet and the tubes are buckling. The pattern seems to be random. Get this fixed.

Case ’18

257

259

8

Prescription for Improvement: Put it all Together This is the time to put all your skills together and work on some trouble-shooting cases. The purpose is to polish your skill. In this chapter we suggest several approaches you could take: work in triads or as an individual. Then an outline is given of the different cases included in this book. Then details are given for the cases. Enjoy.

8.1

Approaches to Polish Your Skill

You can work in triads or as an individual. 8.1.1

Triad Activity

One of the most effective ways to develop skill is to use a triad activity in which each person, in turn, plays the role of expert system, trouble shooter and observer. In a 2-hour period each can play each role once. Here are the three roles. a.

The Expert System

The expert system poses a trouble-shooting Case and then provides responses “from the system” to any request made by the trouble shooter. In preparation for this role, 1) select a case; read it over carefully, 2) then track through all of the data requests given at the end of the case; 3) locate the answers and then read Appendix E for a debrief and answer. Understand the process extremely well. You might wish to complete a Trouble-Shooter’s Worksheet (from Chapter 2) for yourself. Think about how “the fault” will affect all of the process variables. Try to anticipate the kinds of actions that the trouble shooter might take. What would the fault do to the system under those conditions? Give the results of experiments. Do not give explanations. Give correct information but do not be generous. If, for example, the fault occurs periodically, and you are asked to give the lab analysis for one sample taken, then assume Murphy’s law applies and give them the result when the system was operating normally. Insist that they write out all Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

260

8 Prescription for Improvement: Put it all Together

requests; write down the results opposite. Do not talk. . . . . just acknowledge that they are working on it by saying “Ahemmm, mmmmm,” Insist on their instructions or requests be written precisely. If they write “Inspect the instrument” then respond “It’s OK”. If they ask what you did, then say “I went out and looked at it.” Be tough. Do not offer more information than they asked for. b.

The Trouble Shooter

You have a challenging role to play. .

.

.

c.

you are to talk aloud so that the Observer can track what you are doing. Think about the process you will be using: are you searching for change? or for the basics?; are you clarifying the situation or testing an idea? You may feel frustrated because the Expert System is going to supply written responses to your requests for tasks to be done; the Expert System will not discuss things with you. He/she may say “Hmmm” or “mmmmmm” to you occasionally so that you do not feel as though you are talking to a wall, but the Expert System is there to provide an instant system written response to your requests. you are to display all the good problem-solving skills we have developed. Do this by verbally monitoring your progress, being active with pencil and paper to keep track of the route you are following. Indeed, you might wish to prepare a Trouble-Shooter’s Worksheet. Recall, from Chapter 2 (and Chapters 5 and 6), the performance characteristics of successful trouble shooters. Try to display those performance characteristics. you are to write out your requests for information from the Expert System. These should be written out precisely and unambiguously. Observer for trouble shooting

Your worksheet is the Feedback form given in Chapter 2, Worksheet 2-2. As the trouble shooter is tackling the problem, your task is to assess how well the problemsolving components are handled. This is challenging because the skills are difficult to identify – let alone observe and assess. To some extent your role is similar to the listener role in TAPPS, described in Section 5. 1. The feedback form is made to help you look at the mental process used by the trouble shooter. Try to focus more on the “actions taken”, and on the “talk-aloud description of the thought process”. Look at the organizational pattern used; listen for the monitoring of the process. Consider, “Does he/she confuse activities unknowingly?” Let the Expert System focus on “how well the trouble shooter wrote out the questions and tasks to be done”. Activity 8-1

Prepare for the activity ahead of time with each person in the triad selecting a Case and preparing for the role of Expert System. As expert system, make a copy of the Case to give to the trouble shooter. When it’s convenient, the triad members meet for at least 2 hours.

8.1 Approaches to Polish Your Skill

1) The person with surname first in the alphabet starts as trouble shooter. Next in the alphabet, as expert system and last in the alphabet as observer. Seat yourselves approximately as illustrated in Figure 8-1, although the barrier between the trouble shooter and expert system is imaginary.

Figure 8-1

2) 3)

The triad activity.

Refresh your memory as to how to play each role. Set the timer for 20 minutes, start with the expert system handing the Case to the trouble shooter. The trouble shooter reads the case aloud and then, by talking aloud proceeds to “solve the case”. He/she gathers information by writing out actions to be taken. These actions should be written one at a time with a

261

262

8 Prescription for Improvement: Put it all Together

4)

5) 6)

7) 8) 9)

written response being given by the expert system to each task. The expert system responds in writing and so the role-playing activities continue. Focus on the process and not on rushing to trying to solve the problem in the available time. In the time available, the trouble shooter may only complete the TroubleShooter’s Worksheet and never even ask a question of the expert system. That’s OK. When the 20 minutes has elapsed, the expert system reveals the root cause and possible solution. This is not a discussion. This is simply the expert system sharing the root cause about the case. Time = 2 minutes. The observer completes the feedback form and gently shares his/her feedback with the trouble shooter about what he/she observed. 4 minutes. The trouble shooter collects the evidence about the process: the case-problem statement, any worksheets, perhaps the Trouble-Shooter’s Worksheet, and the action-request form with responses that were written between the trouble shooter and the expert system. All three people write their reflections of what they learned from the activity. 5 minutes. Discuss this briefly with the people in your triad. 3 minutes. Rotate roles and repeat until everyone has had a chance to play all three roles and gather evidence about how they played the role of trouble shooter.

As an aside, although the primary goal of this activity is to improve your troubleshooting skill through the role as trouble shooter, most participants report that they learned the most from playing the role of expert system. Interesting. 8.1.2

Individual Activity

As an individual, read over the cases given in Section 8.2. The cases are graded according to degree of difficulty; some cases are written in a general context; they could occur in any process. Others are process specific. For the case you selected, you might start by completing the Trouble-Shooter’s Worksheet given in Chapter 2. Then, scan the list of diagnostic actions given, prioritize, select your choices in sequence and obtain the answer in sequence from the code in Appendix D. For example, in Case ’8 you might decide that your first activity is “put on safe-hold”. The code is 1880. From Appendix D, the result of this activity is 1880: “Not needed at the beginning until after we have collected data. Indeed if put on safe-hold we won’t be able to collect data to figure out what is wrong. $3000” Keep a record of the codes you use. Continue until you believe you have identified the root cause or have corrected the fault. Check your answer with the fault encountered in industry, recorded in Appendix E. Reflect on the problem-solving process used and compare the sequence of actions with those reported for the case in Appendix E. Total up the cost. You might also give yourself-feedback using the form given in Worksheet 2-2.

8.2 Cases to Help you Polish Your Skill

8.2

Cases to Help you Polish Your Skill

First, the cases are listed and guidelines provided for each. Then the cases are given together with a set of diagnostic tasks from which you can select. These tasks have been listed from the choices made by previous trouble shooters. I hope that I have included all the options you like. 8.2.1

Guidelines for Selecting a Case

The actual process problems have been selected carefully from our files of several hundred. These represent varying levels of difficulty, different types of situations (from startup to usual operations) and different varieties of equipment. Table 8-1 summarizes these options. In selecting the sequence of cases to work on, I recommend that you start simply and build up your confidence and skill. The criteria to use in selecting the case are: .

.

.

.

Degree of difficulty. Although this is a subjective rating, I think it is useful. The basis for the ratings is given in Appendix E. Start looking at cases at Level 3 and 4. The lower ratings start with Case ’19. Type of equipment involved. The key pieces of equipment (plus people) that are in the process are listed. For example, Case ’19 relates to filtration and pumps. If your experience with process equipment is being developed for this type of equipment, then you may wish to return to this case later. Alternatively, you can consult Chapter 3 and trouble shooting and suggestions for good practice early in your consideration of the case. As you scan the diagnostic actions listed for each case, you might find the Vendor files and “call to the vendor” activities helpful to do early. Type of process. Some cases apply to any process. Others include characteristics that are unique to the industry. As a start, you may want to select cases that are “general”. Alternatively, some information unique to certain processes may be obtained by consulting the diagnostic actions: MSDS, process description, and Handbook. Continuity. Some cases all relate to similar processes. For example, four cases relate to the depropanizer-debutanizer system. Two, to the ethylene process; five, to the ammonia-reformer. The interconnections are given in column six in Table 8-1. Once you are on a roll with one type of process you might want to continue with problems for the same type of process.

8.2.2

The Cases and Understanding the Choice of Diagnostic Actions for each Case

The cases are listed in Table 8-1. Each case has two parts: the case statement (and diagram) and the list of coded diagnostic actions from which you can select. Here

263

264

8 Prescription for Improvement: Put it all Together

are some more details about how I have designed the diagnostic actions (and the feedback provided for each from Appendix D). There are no diagnostic tests listed for the case that tells you the fault. The diagnostic actions provide you with sufficient evidence that you can identify the fault. Identifying the fault, correcting the fault and preventing the fault from reoccurring are left to you. But you have to select the most pertinent diagnostic actions to give you the key evidence. You are free to select any diagnostic action you wish from among those listed for each case. Every action you request incurs a cost. Unless otherwise specified, I have used $600/h as a cost for loss in production. To this needs to be added the cost of the people and equipment needed. For example, the diagnostic action of determining “what’s the weather today and in the past?” costs $50 (representing about a 10-minute activity). The diagnostic action of “sampling and analysis” would cost $3700 to $15 000/sample depending on the type of analysis involved. (This represents a 6–24-h activity). For each case I tried to assign reasonable costs for each action. The diagnostic actions fall into four classes. The first actions relate to safety! The second set of actions help you understand the background. The third set are factual gathering of information about the process as it operates now. The fourth set are various tests, calculations and activities to test hypotheses, and provide evidence as to the fault. As mentioned at the start of this section, the activities do not necessarily tell you the fault. Even the actions called Take Corrective Action may not correct the fault; they may, but they may not! Do not look for “the answer” to be among the options called Take “corrective” action. Make your selections based on logic combined with safety and economics. Here are more details of the four sets of activities that you may select. 1.

The first two activities relate to safety

The first priority is safety: recognize hazards (via MSDS) and take action. .

MSDS

Usually all engineers working on any site should know this information already. If you don’t for the particular case, MSDS information is available: this usually is expressed as the three National Fire Protection Association, NFPA, ratings for health, fire and spontaneous reaction/explosion for individual chemicals. The ratings range from 0 to 4 with 0 meaning negligible and 4 meaning extreme. Thus an NFPA rating of 0, 4, 0 would mean that the species is not an issue for health, it is extremely flammable and it is stable. I tried to include information about how one chemical might react with another when this is important. .

Immediate action for safety and hazard elimination. The first decision relates to safety. The initial evidence might suggest a health, a fire or an explosive hazard. Act now! There are four options: – Continue with the trouble shooting without implementing activities related to safety: there is no hazard.

8.2 Cases to Help you Polish Your Skill

–

–

–

2.

Put on safe-park: this keeps the process going but under conditions that are safe. This could mean isolating a distillation column and keeping it on total reflux; or reducing the throughput to conditions that previously did not pose a hazard. Safety interlock shut down, SIS: this should happen automatically if the control system has been designed with the four levels expected. However, sometimes this has to be actively initiated. This gives you a chance to reflect on the situation and decide if this should have happened. SIS plus evacuation: the SIS should happen automatically if the control system has been designed with the four levels expected. Now, because of the hazard posed we should add evacuation.

The second set of activities: background that you probably know

You probably know most of this information, except the weather and information related to maintenance. However, as you are developing your skill this section provides easy access to background information related to the case. .

.

.

.

.

More about the process. This is to provide some reassurance about what the process is about; this additional information might help. I have not included key information that is needed to solve the case in this activity. IS and IS NOT: this is based on given problem statement. This might help as feedback to you about how you have done this task. Why? Why? Why?: may be helpful to put some of the cases in the larger context. You can do this activity on your own and use this question to give you feedback. Weather Today and past. All cases refer to weather conditions in Ontario, Canada, where there are four seasons with snow and freezing weather for December through March; hot humid summers June to August. Many of the cases are sensitive to the weather. Maintenance: turnaround. Three conditions might apply: new plant startup, startup of an existing plant that has just been through its annual turnaround and operation after some maintenance has been done. – That the problem is with a first time or startup of a new process is usually given in the statement of the problem. Therefore, no separate question is posed related to this. – The Maintenance: turnaround activity relates to startups after the annual turnaround. During the turnaround the minimum is usually – the inspection of most pieces of equipment, – the replacement of worn parts, – the installation of changes to the process, repiping, and changing the operation to implement ideas to optimize or improve operation, – the calibration of sensors, – cleaning exchangers, and – changing the catalyst.

It is important to know “When and what done?”

265

266

8 Prescription for Improvement: Put it all Together .

.

Maintenance: routine. During routine maintenance something will have changed. It is important to know what and the extent. It could be that all the isolation valves were not correctly opened after the maintenance was completed; that the key was not correctly fitted into the drive shaft; that the pump was not primed. The answers I give to this question will not admit to such mistakes. However, this will open your mind to possible things to check and look at. What should be happening based on office records and files: the resources include the simulation/design computer background; records of the design; the collection of information from the equipment vendors and any internal reports on past tests, or trials done on the plant. Handbook data, mainly properties, for the conditions and species in the case are available. Troubleshooting suggestions are also available as are simple calculations you might do based on the given information in the case.

Design and simulation files (allowances made for fouling, overdesign and uncertainties). Here I tried to list all the possible pieces of equipment involved in the case. All equipment is designed according to the Codes and Standards; the information given here are decisions within a designer’s judgement. Vendor files: practical information and some specifications from the suppliers of heat exchangers, steam ejectors, pumps and equipment that would be purchased from a vendor. The information may be slightly different from the information in the design files. Commissioning data, P&ID, internal reports. Often new plants are constructed under contract. The contract may include penalty clauses stating a financial penalty that is paid if the process does not perform up to specification within a stated period of time. For insurance and governmental regulations, certain performance standards have to be met. Usually records are kept of the performance trials. However, don’t look for much information here for the startup of a new plant. Handbook: physical and thermal properties of the chemicals in the case: vapor pressure–temperatures, steam tables, and thermal properties may be given. Trouble-shooting files. Here I refer to specific sections of Chapter 3 that pertain to the equipment in the case. .

Calculations and estimations that can be done in the office based on the information given in the case statement and the diagram (if available). The given information varies from Case to Case. Nevertheless, some simple checks and calculations can provide neat insight. These include: – Pressure profile: in most cases we can obtain the pressures on either side of a barrier and then state the direction of flow that would incur if there is a leak in the barrier. – Mass balance: nice idea but you usually can’t do this at this early stage in the problem because the key data are not given in the case statement. – Energy balance: sink= source; the heat lost by one fluid = heat gained by the other. Often we are given enough information in the case statement to allow us to do this check.

8.2 Cases to Help you Polish Your Skill

– Thermodynamics: use the principles of thermodynamics to predict trends, in general. – Rate: use DTs to estimate whether nucleate or film boiling might predominate. Similarly, we might be able to estimate rates of mass and heat transfer. – Equipment performance: we might be able to estimate the performance of equipment; usually, however, we need more information than that given in the case statement. 3.

The third set are factual gathering of information about the process as it operates now.

.

What is the current operation

Unless you arranged over the phone to meet the operator at the piece of equipment, you will go to the control room first and check in. You might scan the data reported in the control room; discuss the operating activities with the operator and explore the operating procedures. Visit control room: control-room data: values now and from past records. Process operators: you can ask for information about what happened (from their perspective) and they might offer ideas as to what is the fault. The operators are usually a valuable source of information. Operating procedures: knowledge of the usual operating procedure to be used for this condition might help. .

Check with colleagues about hypotheses

The purpose of this book is to develop your confidence and trouble-shooting skill. To provide feedback about reasonable hypotheses (that you will have generated by this time in your journey through the case), I give this diagnostic check, but only for the Cases up to Case ’29. You may wish to create a symptom ‹ cause diagram similar to the one illustrated in Figure 6-5, in Section 6.5.5c but I give no feedback about this for the cases. 4.

The fourth set are various tests and activities selected to test hypotheses/or correct

This is the central action for each case. You may choose the diagnostic actions in any order. For example, you might suspect that the sensors are incorrect. One test is to gather data that checks for consistency. In the format of this text, I rarely include a section called “consistency tests”. I expect you to identify neat ways to check for consistency. Methods might include: 1) 2) 3)

to compare two sensors at the same or close-by location, to look for agreement between composition, temperature and pressure (say at the top or bottom of a distillation column), to check for agreement between temperature and pressure on a pressure– enthalpy diagram for pure refrigerant,

267

268

8 Prescription for Improvement: Put it all Together

4)

to check that the conditions on a stream are the same at two locations. This latter type of information we often obtain by contacting the operators of the utilities or of the plants upstream or downstream from our process called Contact with on-site specialists.

The diagnostic activities from which you can select are: .

.

.

. .

.

. . . . .

visit the site, go out on the plant, see, smell, feel and listen; and read gauges whose values are not shown in the control room, check that the diagram given with the case agrees with reality, Check diagram and P&ID versus what’s out on the plant, do on-site simple tests, such as “turn and seal” test on valves; and check for trends in the data. gather data needed to do key calculations, check the sensors. This includes checking the sensor’s response to change, using temporary instruments, and calibrating the sensors; consider the control system: put the system on manual, retune, call in instrument specialists. get information from vendors, gather samples and have these analyzed, do more complicated tests, such as gamma scans, tracer studies. open and inspect, and take corrective action.

Cautions must be given about the last two diagnostic options. Open and inspect is expensive and usually a last resort. Use this only when you have narrowed the problem down to a specific fault and piece of equipment. As mentioned at the beginning, the cause is not usually found in the list of Take “corrective” action. Sometimes this does corrects the fault. Sometimes it doesn’t. When you have completed your set of diagnostic actions, then you should have all the evidence needed for you to write down the fault, think about how to correct it and how to prevent its reoccurrence. Enjoy! Now here, in Table 8-1, is the list of cases that I have selected to help you develop your skill in trouble shooting. Table 8-1

Case number and name

Degree of difficulty

Equipment involved

Type of process

Continuity: related to Case ’

1

Ammonia startup

6

ammonia

2

Leak

4

reactor, startup heater and compressor pipe and storage tank

’29, 15, 36, 42, 50–52 ’29, 15, 36, 42, 50–52

3

Cycling column

4

distillation column plus auxiliaries

ammonia general

8.2 Cases to Help you Polish Your Skill Table 8-1

Continued.

Case number and name

Degree of difficulty

Equipment involved

Type of process

4 Platformer fires

4

refinery

5 Acid pump 6 Air dryer 7 The reluctant crystallizer 8 Depropanizer: the temperatures go crazy 9 Bleaching plant

5 8 10

heat exchanger, platformer reactor pump, storage tank adsorption, regeneration vacuum, crystallizer

8 7

10 Dry and not to dry

8

11 Lazy twin 12 Drop boxes

5 3

13 The lousy control system 14 Condenser that was just too big 15 The flooded boot

4

16 The case of the dirty vacuum gas oil 17 Is it hot or is it not?

3 9

5

distillation column plus auxiliaries conveying powder for adsorption in vacuum deoderizer crystallizer, screen, hydrocyclone, centrifuge and steam dryer pumps distillation, adsorption, regeneration dryers, exchangers, pumps distillation column, overhead condenser distillation column, vacuum, overhead condenser vacuum, condenser and pumps

4

distillation plus auxiliaries vacuum distillation

18 The streptomycin dilemma 19 The belt filter

7

evaporator

2

20 The fussy flocculator pump 21 The flashy flare

3 3

22 pH pump

4

filter, screen, pump, sedimentation tank pump, storage tanks, flocculation refinery, flare system, compressor pump, mixing

23 The hot TDI

4

24 Low production on the ethylene plant

4

polymerizer, mixer, cooling system distillation, adsorption, regeneration dryers, exchangers

Continuity: related to Case ’

general general general depropani- ’32, 38, 41,43, zer 45, 48 foods

foods

general ethylene

Case ’24

general

’30

fatty acids, food see Appendix C general, reformer, ammonia refinery foods, pharmaceutical food, pharmaceutical wastewater ’37 general, wastewater refinery general, wastewater polymer ethylene

Related to ’12

269

270

8 Prescription for Improvement: Put it all Together Table 8-1

Continued.

Case number and name

Degree of difficulty

Equipment involved

25 The case of the delinquent exchangers 26 The drooping temperature 27 IPA column

4 5

reformer reactor, furnace, refinery exchangers, pump furnace, pump general

28 The boiler feed heater

5

29 Reluctant reactor

5

30 The case of the reluctant reflux 31 Ethylene product vaporizer 32 What does the alarm mean? 33 Chlorine feed regulation 34 The cement plant conveyor 35 The cycling multiple effect evaporator 36 The really hot case

6

7

37 Mill clarifier

7

38 More trouble on the deprop

7

39 The case of the lumpy sunglass display

7

feed bin, molding machine, mold and mold design

40 The cool refrigerant

7

41 Ever increasing column pressure 42 The weak AN

7

refrigeration system, compressor, turbine, control, knockout pot distillation column plus auxiliaries storage tank, pump, reactor, cooling coil

5

6 6 6 6 6

7

Type of process

petrochemical, refinery shell and tube exchanger general to heat boiler feedwater ammonia reactor, compressor, separator, condensers, refrigeration unit distillation column plus general auxiliaries heat exchangers, boiling ethylene.

Continuity: related to Case ’

distillation column plus auxiliary equipment

sequence of distillation columns plus auxiliaries slurry, pump, control system, storage tank solids conveyer, bagging, dust filters, cyclones long tube evaporators, vacuum system reactor and heat exchanger, steam generation thickener, sludge pumps, control sequence of distillation columns plus auxiliaries

’36, 15, 42, 50–52 ’13

petro’8, 32, 38, 41, chemical 43, 45, 48 minerals processing general, ceramic glycerine reformer; ammonia pulp and paper petrochemical, refinery, polymer, injection molding of thermoplastics general

general ammonia

’29, 15, 42, 50–52 ’19 ’8, 32, 41, 43, 45, 48

’8, 32, 38, 43, 45, 48 ’29, 15, 36, 50–52

8.2 Cases to Help you Polish Your Skill Table 8-1

Continued.

Case number and name

Degree of difficulty

Equipment involved

43 High pressure in the debut!

7

44 Reactant storage

8

45 The deprop bottoms and the ISO dilemma

8

46 Not so cool chiller

8

47 The fluctuating production of acetic anhydride 48 The column that just wouldn’t work

8

’8, 32, 38, 41,45, sequence of distillation petrocolumns plus auxiliaries chemical, 48 refinery storage tank, heating general coil, steam traps ’8, 32, 38, 41 43, sequence of distillation petrocolumns plus auxiliaries chemical, 48 refinery pumps, exchangers, polymerizarefrigeration cycle tion petrofeed vaporizer, vacuum pump, absorber, chemical, acetic acid condensers, reactor sequence of distillation petro’8, 32, 38, 41, columns plus auxiliaries chemical, 43, 45 refinery feed bin, molding injection machine, mold and mold molding of design. thermoplastics vacuum distillation general, ’29, 15, 36, 42, column plus auxiliaries ammonia 50–52

8

49 The case of the faulty stretcher pedal

8

50 The cleanup column

9

51 More trouble on the cleanup column 52 Swinging loops

9 9

vacuum distillation column plus auxiliaries reactor, compressor, separator, condensers, refrigeration unit

Type of process

Continuity: related to Case ’

general, ammonia ammonia

[8, distillation column plus auxiliaries, depropanizer] The problem statement is given in Chapter 2, Section 2.4. Case ’8: Depropanizer: the temperatures go crazy

. .

. .

.

MSDS, 1495 Immediate action for safety and hazard elimination, Put on safe-park, 1881 Safety interlock shut down, 1531 SIS plus evacuation, 732 More about the process, 1010 IS and IS NOT, Where? upstream, 537 Where? downstream in the debutanizer, 979 Why? Why? Why?, Best goal? To keep bottom level steady and stop tray temperature cycling, 1335

271

272

8 Prescription for Improvement: Put it all Together . .

Weather, Today and past, 23 Maintenance: turnaround, When and what done?, 421

Maintenance: routine, When and what done?, 2 What should be happening, Design and simulation files (allowances made for fouling, overdesign and uncertainties) Condenser, E-25, 591 Distillation column, C8, 515 Thermosyphon reboiler, E-27, 380 Feed pump, F-25; F-26, 16 Turbine drive, 1193 Motor drive, 1072 Reflux pump, F-27, 286 Feed preheater, E-24, 1989 Feed drum, V-29, 1478 Overhead drum, V-30, 1124 Vendor files: Condenser, reboiler and preheater, 10 Steam traps, 481 Commissioning data, P&ID, internal reports, 1245 Handbook, Cox charts, 803 Trouble-shooting files, 1430 . .

.

.

Calculations and estimations (that can be done in the office before special tests are done and based on rules of thumb and information given in the case) Energy balance: sink= source, Estimate the steam flow to the reboiler based on the reflux rate and the fact that each kg steam boils 5 kg typical organic, 821 What is current operation

Visit control room: control-room data, Feed to the C8, depropanizer. FC/1, 563 Reflux flowrate, FIC/4, 155 Pressure drop Dp I/1, 78 Level bottoms LIC/2, 1436 Temperature bottoms TI/4, 1765 Temperature mid-column TIC/5, 1502 Temperature top, TI/3, 1947 Pressure on overhead drum, PIC/10, 1610 Process operators, This shift, 334 Previous shift, 33 Operating procedures, Column pressure, 659 Feed to the column, 1834

8.2 Cases to Help you Polish Your Skill .

Check with colleagues about hypotheses, 2395

Call to others on-site, Call operators of process supplying the feed, V29, about possible upsets and details of flowrate and composition of feed, 2371 Call operators of downstream units receiving the propane, 2757 Call operators of downstream unit receiving butane, 769 Visit site, read present values, observe and sense. Column pressure, PI-4, 303 Look at the flare; same size as usual?, 799 Pressure relief to flare PSV-1, 1177 Temperature mid-column TI- 8, 1025 Valve position for column feed, FV- 1, 1951 Valve position for steam to preheater E-24, 208 Valve-stem position on PV-10, 614 Level in feed drum, V-29; LI- 1, 885 Pressure on exit of pump F-26, PI-3, and compare with head-capacity curve at feed flowrate, 691 Listen to the check valve on turbine-driven pump, downstream of PI-2, 1273 Observe whether shaft is rotating for the feed pump F-26. Is the pump noisy?, 1906 Check diagram and P&ID versus what’s really out on the plant, 2769 On-site simple tests: Shut isolation valves around turbine-driven pump F-25, 1585 Put TIC/-5, FIC/-1 and -4 and LIC/-2 controllers on manual and try to steady out the column, 1737 Shut exit valve on discharge of pump F-26 and read pressure, 1316 .

Gather data for key calculations,

Pressure profile, Drum V-29 to pumps, F-25, 26, 1792 Dp across pump, converted to head, 1666 Pump F-25–26 exit to feed location, 63 Drum V-30, pump F-27 and reflux into column, 917 Pump F-27, 1868 Vapor from top of column to vapor space in V-30, 1171 Thermosyphon reboiler process fluid side, 133 Mass balance, over column, 442 Perform more complicated tests, Gamma scan over the stripping section to locate collapsed tray, 255 .

Take “corrective” action, Put column on “safe-park”, 1390

273

274

8 Prescription for Improvement: Put it all Together

[7, conveying powders, adsorption, vacuum deodorizer, foods] The problem statement is given in Chapter 5, Section 5.3.2. Case ’9: The bleaching plant

. .

.

. . .

MSDS, 1480 Immediate action for safety and hazard elimination, Put on safe-park, 1196 Safety interlock shut down, 1345 SIS plus evacuation, 2400 IS and IS NOT, What, 1416 When, 1896 Who, 2838 Where, 2245 Why? Why? Why?, Best goal? to suck Fullers earth into deodorizer, 1008 Weather, Today and past, 2977 What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) For the diameter and length of conveying line, the design flowrate of Fuller’s earth and the design conveying velocity, what vacuum is needed in the bleacher?, 2690 Hopper design, 2562 Piping for pneumatic conveying, 347 Vacuum system, 59 Handbook, Physical properties of Fuller’s earth, 938 Trouble-shooting files, 535 .

Calculations and estimations

Pressure profile, Is the pressure difference between atmospheric and vacuum (Dp= 80 kPa) sufficient to convey Fuller’s earth?, 1368 .

What is current operation

Visit control room: control-room data, none of the data are shown in the control room; all are on the unit, Process operators, When you say “nothing happened” and “you couldn’t see powder dumping into the liquid”, what did you hear and see? 1097 Operating procedures, 1925

8.2 Cases to Help you Polish Your Skill .

Check with colleagues about hypotheses, 1762

Call to others on-site, Utilities: any upsets in the steam? Check that the steam pressure at the boiler house is approximately the steam pressure at the ejectors, 2431 Visit site, read present values, observe and sense. Pressure gauge, 193 Check that there is Fuller’s earth in the hopper, 2920 Is the level of liquid in the bleacher so high that it covers the inlet line for the Fuller’s earth into the bleacher, 2630 Check diagram and P&ID versus what’s out on the plant, 1206 On-site simple tests: Check for leaks into the bleacher from around the agitator shaft, 2988 Rap the pipe and the side of the hopper to try to dislodge any bridging that might be occurring in the humid weather, 2767 Use “turn and seat” to check the valve on the conveying line and leave open, 2334 Sensors: check response to change, Pressure gauge on the bleacher, 2150 Sensors: calibrate, Replace/calibrate the pressure gauge, 240 Contact vendor supplier, Did other customers receive Fullers earth similar to batch number 4853 that we received, and if so, have they had any comments or queries?, 366 Are there any particular precautions we should be taking?, 831 Samples and measurements, Sample Fuller’s earth, 457 More ambitious tests Remove the Fuller’s earth from the hopper, and crack open the valve. Listen for air flowing into the bleacher and observe vacuum gauge, 2456 Insert porous tubes into the bed of Fuller’s earth and position them such that compressed air is blown in to try to fluidize the powder near the inlet of the conveying line, 1856 Open and inspect, Open conveying line and check for plugs, 1442 .

Take “corrective” action, Replace the valve, 2437 Relocate the inlet to the conveying line. Instead of using a pipe stuck into the bed, attach the inlet to the bottom of the conical hopper, 2750

[8, crystallizer, screen, hydrocyclone, centrifuge, steam dryer, foods] The problem statement is given in Chapter 5, Section 5.4.2. Case ’10: To dry or not to dry

. .

.

MSDS, 52 Immediate action for safety and hazard elimination, Put on safe-park, 308 Safety interlock shut down, 1331 SIS plus evacuation, 1103 More about the process, 2234

275

276

8 Prescription for Improvement: Put it all Together .

.

. .

IS and IS NOT, What, 2209 Where, 2279 When, 2642 Why? Why? Why?, Best goal? To get the exit crystal moisture content < 1.5%, 2779 Weather, Today and past, 290 Maintenance: turnaround, When and what done?, 492

Maintenance: routine, When and what done?, 270 .

What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Crystallizer, 20 Hydrocyclone, 427 Pump, 45 Surge vessel installed between the screen and the centrifuge?, 4 Surge vessel installed between the centrifuge and the dryer?, 405 Vendor files: Centrifuge, 816 Hydrocyclone, 510 Pump, 992 Rotary dryer, 1490 Feeder screw, 1224 Parabolic screen, 711 Commissioning data, P&ID, internal reports, 878 Trouble-shooting files, 533 .

What is current operation

Visit control room: control-room data, all data on the plant. Process operators, This shift; elaborate on the changes you have made, 1411 Operating procedures, What is the new cycle time on the centrifuge and how does this compare with before?, 1009 Cycle time for screening and washing for the screen, before and now 1041 .

Check with colleagues about hypotheses, 1311

Call to others on-site, Utilities: steam plant: any upsets, pressure delivered to our site, 2219 Stores: any available spare equipment: centrifuges, screens, 1804 Utilities: wash water for screen and centrifuge: change in quality, upsets, 2786 Visit site, read present values, observe and sense. Records of analysis of moisture content in feed to dryer. Samples taken at the beginning and end of a shift, 71

8.2 Cases to Help you Polish Your Skill

Check diagram and P&ID versus what’s out on the plant, 233 On-site simple tests: Reduce throughput for the unit and thus reduce feedrate to dryer, 360 Steam condensate trap on dryer working?, 1971 Estimate feedrate to the screen over the screening cycle and compare with feedrate before the change, 1514 Observe feed condition from the screen to the centrifuge during the wash cycle of the screen, 1114 Estimate feedrate from the screen to the centrifuge over the filter cycle and compare with the feedrate before the change, 1458 Estimate the feedrate to the centrifuge during the screen wash cycle and compare with the feedrate before the change, 561 Estimate the feedrate from the centrifuge to the dryer during the filter cycle and compare with the feedrate before the change, 909 Estimate the feedrate to the dryer during the centrifuge wash cycle and compare with the feedrate before the change, 545 Stop the centrifuge when the screen is in its wash cycle. Test the degree of compaction of the residual crystals under the peeler, 54 Gather data for key calculations Pressure profile, Recycle through hydrocyclone, 1671 Energy balance: sink= source, Measure amount of condensate by submerging exit line in bucket with preweighed amount of cold water. Does energy lost from condensation of steam= energy gained by evaporation?, 603 Equipment performance, Rating dryer, 1860 Screen, 1566 Centrifuge, 1977 Call to vendors, licensee, Rotary dryer, 482 Samples and measurements, Sample the crystal discharge from the screen every 30 s for three cycles of screen operation; measure % screen overs as a % of feed. Compare with results for old operation, 1968 Sample the discharge from dryer every 20 s for three cycles of screen operation; measure moisture content. Compare with data for previous operation, 1520 Sample the liquid-fine underflow from the screen every 30 s for the screening cycle. Compare with data from previous operation, 38 Draw a composite sample the wash water exit from the screen. Analyze for particle-size distribution. Compare with previous operation, 318 Draw a composite sample of the wash water from the centrifuge. Analyze for particle-size distribution. Compare with previous operation, 1911 Sample the feed to the dryer every 20 s for three cycles of screen operation; measure moisture content. Compare with previous operation, 196

277

278

8 Prescription for Improvement: Put it all Together

Return to previous operating conditions, Sample the discharge from dryer every minute for three cycles of screen operation; measure moisture content, 1705 Sample the feed to the dryer every minute for three cycles of screen operation; measure moisture content, 1032 Sample the crystal discharge from the screen every 30 s for three cycles of screen operation; measure % screen overs as a% of feed, 558 Open and inspect, Centrifuge, 913 Dryer, 689 Take “corrective” action, Revamp the condensate removal system: install a thermodynamic trap, slanted exhaust pipe to the trap and minimized the distance between the trap and heating tube, 795 Operate the centrifuge on former operation, 1194 Case ’11: The lazy twin [5, pumps, general] The problem statement is given in Chapter 6, Section 6.2.2. . .

.

. .

. .

MSDS, 376 Immediate action for safety and hazard elimination, Put on safe-park, 42 Safety interlock shut down, 490 SIS plus evacuation, 984 IS and IS NOT, What, 337 When, 2368 Who, 2912 Where, 2728 More about the process, 1811 Why? Why? Why?, Best goal? get the flow up to design rate when pump A is running, 1667 Weather, Today and past, 1415 Maintenance: turnaround, When and what done?, 1400

Maintenance: routine, When and what done?, 1822 .

What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Centrifugal pump A, 1595 Centrifugal pump B, 1510 Heat exchanger, 166 Check valves, 395 Vendor files: Centrifugal pumps A and B, identical and from the same vendor, 930 Heat exchanger, 540

8.2 Cases to Help you Polish Your Skill

Commissioning data, P&ID, internal reports, 551 Trouble-shooting files, 976 .

.

Calculations and estimations. None can be done based on the given information What is current operation

Visit control room: control-room data What is the temperature after the heat exchanger when pump A is on-line?, 386 What are the levels in the downstream equipment when pump A is on-line?, 470 Is FRC/-100 local or remote?, 506 “What is the pressure in the storage tank T-200?”, 973 Process operators, Current shift, 1470 Operating procedures, About use of pumps, 1006 .

Check with colleagues about hypotheses, 90

Call to others on-site, Contact downstream cat-cracking unit. Everything working as expected? is the behavior consistent with the flowrate signalled from FRC/-100?, 648 .

.

Visit site, read present values, observe and sense. Do either pumps A or B sound as if they are cavitating?, 1998 Look and see that valves V200, V201, V202, V205 and V206 are all open, as expected, 518 Read the controller output to valve V100 when pump A is running, 2325 Read the controller output to valve V100 when pump B is running, 2441 Observe the valve-stem position on F100 when pump A is pumping, 2086 Observe the valve-stem position on F100 when pump B is pumping, 2845 Observe whether the arrow on each valve V200, V201, V202, V205, V206 is in the direction of flow through the valve, 2935 Observe whether the direction of flow through both check valves agrees with valve installation, 1587 Check the tab on the orifice plate FRC-100 that it is the correct diameter and facing the correct direction, 1160 Pump A running hot?, 1617 On-site simple tests: Shut exit valve on pump B; read pressure on gauge PI-220 when pump is running. Convert to head of fluid, 1030 Shut exit valve on pump A; read pressure on gauge PI-210 when pump running. Convert to head of fluid, 1465 Put pump A on-line and check that the motor started. Repeat for pump B, 602 Test valves V200, V201, V202, V205 and V206 with “turn and seal” to ensure they are working properly, 1461

279

280

8 Prescription for Improvement: Put it all Together

What is the pressure at P210 and at P220 when pump A is on and pump B is off? Express this is head (so I can compare with the head-capacity curve), 550 What is the pressure at P210 and at P220 when pump A is off and pump B is on? Express this is head (so I can compare with the head-capacity curve), 170 Stop pump B and close valve V205. Start pump A. What is the flow on FRC/100?, 102 Stop pump A and close valve V201. Start pump B. What is the flow on FRC/100?, 1482 Check diagram and P&ID versus what’s really out on the plant, 2372 Gather data for key calculations Pressure profile, From Tank T-200 through pump A to cat cracker, 1432 From Tank T-200 through pump B to cat cracker, 7 Energy balance: sink= source, Read clamp-on ammeter, voltmeter and power-factor meters, assume density= design density, compare power for motor to drive pumps A and B, 140 Sensors: check response to change Change set point on FRC/-100; does the valve V100 respond?, 1081 Sensors: calibrate, Flowmeter, FRC/-100, 1552 Open and inspect, Pump B and ask maintenance to inspect and look for reasons why the pump is not functioning well, 1886 Pump A and ask maintenance to inspect and look for reasons why the pump is not functioning well, 213 Stop the process. Open and inspect valves V201 and V202, 453 . Take “corrective” action, Shut down the process. Replace valves V201 and V202, 276 Shut valve V205 whenever pump A is running, 2027 Stop the process. Replace check valves on the exit lines from both pumps, 678 Case ’12: Drop boxes [3, distillation, adsorption, regeneration dryers, evaporators, ethylene] The problem statement is given in Chapter 7, Section 7.2.2. . .

.

MSDS, 2667 Immediate action for safety and hazard elimination, Put on safe-park, 925 Safety interlock shut down, 1428 SIS plus evacuation, 2268 More about the process, About the dryers and dryer cycle, 1715 About the distillation, 264 About the sewer systems, 420 About the low-temperature condensation, 499

8.2 Cases to Help you Polish Your Skill .

.

. .

IS and IS NOT, What, 827 When, 2001 Where, 1571 Why? why? why?, Goal: “clear up the hazard as stated by the safety inspector”, 637 Weather, Today and past, 2494 Maintenance: turnaround, When and what done?, 2085

Maintenance: routine, When and what done?, 2982 .

What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Exchanger E131: cool regeneration gas to the dryer during last portion of regeneration: water tube side; regeneration fuel-gas shell side, 1773 Heater E130: heat regenerative fuel gas to dryer, 256 Exchanger E107and E108: precool feed from 38 C to –29 C, 444 Demethanizer overhead condenser: process fluid shell side; ethylene refrigeration on the tube side, 462 Demethanizer reboiler: process stream shell side; propylene on tube side, 987 Steam trap on regenerative gas heater, 2997 Knockout pot, 2990 Vendor files: Heat exchanger, 2518 Dryer system, 2022 Steam trap, 2533 Commissioning data, P&ID, internal reports, P&ID, 437 Sewer plot plan indicating which streams go to each “sewer gate” and which go to drop box B and which to drop box C, 1101 Trouble-shooting files, 2223 .

.

Calculations and estimations. no calculations can be done based on the limited information in the problem statement What is current operation

Process operators, Any changes in operation?, 2330 .

Check with colleagues about hypotheses, 1318

Call to others on-site, Safety inspector Where were the samples taken and what was the composition?, 1699 Operators of upstream styrene, ethylene and propylene plants: any upsets, any calls from the safety inspector, 2882 .

Visit site, read present values, observe and sense, Look at the flare, 2674

281

282

8 Prescription for Improvement: Put it all Together

Check diagram and P&ID versus what’s really out on the plant, 2164 On-site simple tests: Vent valves closed on process (or shell side) of exchangers E107, E108, condenser E114, E131?, 2477 Vent valves closed on utility side (or tube side) of exchangers E107, E108, condenser E114, E131?, 2364 Vent valves closed on shell and tube side of heater E130?, 2876 Drain valve on knockout pot shut?, 2617 For exchanger E131 cooler; block off the in and out water lines; open vent on tube side and note fluid leaking out, 1211 For the vent valves on the process (or shell side) of exchangers E107, E108, condenser E114, E131, test by “turn and seal”, 2433 For the vent valves on the utility side (or tube side) of exchangers E107, E108, condenser E114, E131, test by “turn and seal”, 2363 Test the vent valves on shell and tube side of heater E130 by “turn and seal”, 2812 Test the drain valve on knockout pot by “turn and seal”, 2557 Retest drop box B and C for explosive mixture at half-hour intervals for 2 hours, 554 Retest drop box A for explosive mixture at half-hour intervals for 2 hours, 119 Gather data for key calculations Pressure profile, Exchanger process fluid pressure relative to utility pressure: direction of leak, 2067 Exchanger utility pressure relative to atmospheric; direction of leak, 2472 Exchanger process fluid relative to atmospheric; direction of leak, 2814 Mass balance, Over demethanizer, 2698 Samples and measurements, Sample cooling water leaving cooler E131, 1488 Gas sample from drop boxes B and C using an evacuated bomb. Lab analysis for the overall concentration of light hydrocarbons and a breakdown of the hydrocarbon portion, 2434 Open and inspect, Cooler E131 and look for leaks in tubes, 2576 .

Take “corrective” action, Replace the vent valves on the shell side of E107, 108, 130, 131, 2923 Replace drain valve on KO pot, 168

Case ’13: The Lousy Control System [4, distillation, overhead condenser, general] The problem statement is given in Chapter 7, Section 7.2.5, Activity 7-9. . .

MSDS, 2266 Immediate action for safety and hazard elimination, Put on safe-park, 2439 Safety interlock shut down, 2932 SIS plus evacuation, 2510

8.2 Cases to Help you Polish Your Skill . .

.

. . .

More about the process, 497 IS and IS NOT, What, 2317 When, 1770 Where, 621 Why? why? why?, Goal: “to prevent too much stuff from going to the flare”, 232 Maintenance at turnaround, When and what, 1106 Weather, Today and past, 30 What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Fan: control system on the air flow to the condenser, 324 Condenser, air-cooled, 739 Distillation column, 1996 Reflux drum, 2070 Reflux pump, 2458 Vendor files: Condenser, air cooled, 2662 Reflux pump, 2910 Commissioning data, P&ID, internal reports, 2543 Trouble-shooting files, 2218 .

Calculations and estimations

Equipment, Rate condenser: heat load, 411 .

What is current operation

Visit control room: control-room data Control-room values now and from past records, Temperature at the top of the column, 2502 Temperature on liquid-gas exit from the condenser, 2685 Pressure at the top of the tower, 2722 Pressure on the overhead receiver, 778 Level in the overhead receiver, 1521 Process operators, This shift, 2179 Previous shift, 2356 .

Check with colleagues about hypotheses, 143

Call to others on-site, Operators of the upstream process, 2185 Operators of the downstream process, 1613 Visit site, read present values, observe and sense. Hot exhaust air recirculation to the intake of the air-cooled condenser, 220 Inspect the hydraulic configuration on the exit header from the condenser, 89 Look at the flare, 456

283

284

8 Prescription for Improvement: Put it all Together

Sounds around the condenser, 325 Check diagram and P&ID versus what’s really out on the plant, 1462 On-site simple tests: Temperatures and humidities of inlet and exit air for the air-cooled condenser, 1866 Lower the elevation of the exit from the header to decrease the amount of flooding, 2879 Gather data for key calculations Pressure profile, On process pipe from the top of the column to the overhead receiver, 2346 Across condenser, 2062 Equipment performance, Rating of air-cooled condenser, 2844 Fans, 1631 Reflux pump, 1876 Column, 1670 Sensors: check response to change Temperature sensor exit of condenser, 1087 Sensors: use of temporary instruments, Surface temperature near thermowell at exit of condenser measured by a contact pyrometer, 192 Sensors: calibrate, Temperature sensor on exit of condenser, 684 Control system, Put on manual; change the set point and note response of fan system, 487 Sampling and measurements Concentration: vent from the overhead receiver; concentration of heavies, 2587 Concentration: vent from the overhead receiver; concentration heavies, ten samples taken at one-hour intervals, 2152 Open and inspect, Air-cooled condenser: visual inspection plus air and water pressure tests, 1254 Fan and fan blades, 2293 .

Take “corrective” action, Operate on manual, 2793 Install hose and let spray of water fall over tubes, 367 Stop operation; open the headers and water-wash with high pressure hoses to remove scale, 1230

8.2 Cases to Help you Polish Your Skill

[3, distillation, vacuum, overhead condenser, fatty acids, food] The problem statement is given in Chapter 7, Section 7.2.5, Activity 7-9. Case ’14: The Condenser that was just too big

. . .

. .

MSDS, 51 More about the process, 273 Immediate action for safety and hazard elimination, Put on safe-park, 400 Safety interlock shut down, 88 SIS plus evacuation, 1500 Weather, Today and past, 1994 What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Condenser, 1660 Backup condenser, 2446 Booster ejector, 3000 Ejector, 2524 Wet vacuum pump, 1781 Barometric condenser, 21 Coil cooler in tank, 966 Reciprocating pump, 1271 Vendor files: Reciprocating pump. F1400, 1528 Ejectors and booster ejector, 1961 Wet vacuum pump F1401, 2163 Commissioning data, P&ID, internal reports, Any files?, 2004 Handbook, Vapor pressure of fatty acids, 2490 Trouble-shooting files, 2969 .

Calculations and estimations

Pressure profile, Estimate the pressure in C1400; vapor side of E1400; vapor side of E1401, 2511 Mass balance, Mass balance on the overhead pumped by F1400 compared with expectations based on feed composition and rate, 2781 Energy balance: Check coolant temperature for E1400, 2580 .

What is current operation

Visit control room: control-room data, P1, 2273 T4, 2377 T3, 390 Process operators, Flow pumped from F1400, 486 Trends in pressures and temperatures, 188 Liquid level in bottom of E1400 as feed to V1400, 871 Anything strange happening?, 1317 Was the vacuum pump started up according to operating procedures?, 1020

285

286

8 Prescription for Improvement: Put it all Together

Visit site, read present values, observe and sense. Cooling-water flow into E1403, 1322 Cooling-water flowrate and temperature to E1404, 1091 Read P3 and look for oscillations in pressure, 1851 Is the water level in the hot well above the exit from the downcomer from the barometric condenser? Water temperature?, 1594 Note whether steam valves to booster and regular ejector are full or partially open, 2201 Does the booster ejector sound as though it is “kicking out”?, 2094 Is the level in the boiling condenser E1400 dropping? Test by trying to add more liquid via the top up funnel at the top of E1400, 2655 Listen for cavitation in pump F1400, 326 On-site simple tests: Visually and audibly check for any leaks of air into the system, 716 Sensors: check response to change T4: 933 T3: 522 Sensors: use of temporary instruments, Put surface temperature sensor on suction line to F1400 and check value, 1155 Call to vendors, licensee, Reciprocating pump: what could cause the knocking its head off?, 1142 Sensors: calibrate, T4, 1920 T3, 1614 Samples and measurements, Feed: usual composition? and does this contain more volatiles than expected?, 1719 Open and inspect, Suction line from bottom of E1400 through to F1400, 2859 .

Take “corrective” action, Flow cold water over the outside casing of pump F 1400, 2939

(supplied by Mike Dudzic, B. Eng. 82, McMaster University) [2, filter, screen pump, deep thickeners, wastewater treatment] The following diagram shows a section of the sludge concentration and dewatering section of our wastewater treatment process. Water containing sludge is concentrated in a deep settling tank and then pumped to a continuous-belt filter. Polymer is added just upstream of the filter to improve dewatering. The process is shown in Figure 8-2. For the past three weeks the plant has been operating smoothly although lately the operator thinks that the flow to the filter has been a bit lower than normal. The operator is relatively new on this plant. He has been on the unit for three months. Today, the operator notes that the sludge has stopped flowing to the filter. The pump is running. If the fault is not corrected quickly, the sludge will build up in the settlers and shortly the whole process will be shut down. Case ’19: The case of the reluctant belt filter

8.2 Cases to Help you Polish Your Skill West thickener (idle)

East thickener Flush water

15 m

Air at 550 kPa

10 cm diam line

Closed

7.5 cm line

10 cm

Polymer addition 2.5 cm water

Heavy dewatered sludge

Filtered water Continuous belt filter Figure 8-2

The dewatering system for Case ’19.

Case ’19: Reluctant belt filter .

. .

. . .

Immediate action for safety and hazard elimination, Put on safe-park, 8 Safety interlock shut down, 330 SIS plus evacuation, 268 More about the process, 2830 IS and IS NOT, What, 2608 When, 2161 Who, 1554 Where, 274 Why? why? why?, Goal: “Get the sludge flowing to the filter”, 596 Weather, Today and past, 494 Maintenance: turnaround, When and what done? 722

Maintenance: routine, When and what done? 2660 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Thickener, 707 Centrifugal pump for handling sludge, 765 Strainer, 997 Continuous-belt filter, 2143 Vendor files, Centrifugal pump, 1518 Trouble-shooting files, 223

287

288

8 Prescription for Improvement: Put it all Together .

Calculations and estimations that can be done in the office based on the information given in the problem statement

Pressure profile, Calculate to see if 550 kPa g pressure is sufficient for air to actually backflow through the thickener, 1635 .

What is current operation

Process operators, When did the flow appear to decrease? 1330 How long had the east thickener been idle before you started using it? 1757 Have you ever seen anything like this before? and what did you do?, 2329 Operating instructions, Standard procedure: if there is low sludge flow; flush the exit lines from the thickeners with high-pressure water and high-pressure air, 2052 Check with colleagues about hypotheses, 1805 Call to others on-site, High pressure air; have there been any interruptions in service? what is the pressure of the air being delivered to site? 1258 Visit site, read present values, observe and sense. Is the shaft of the pump rotating? 322 Is the block valve V101, to the idle thickener closed? 731 Are the isolation block valves around the strainer and around the pump open? 294 Check diagram and P&ID versus what’s out on the plant, 2839 On-site simple tests: Flush the lines with high-pressure water and air for five minutes, 793 Test by “opening and closing” the block valves V100, V101, 2672 Test by “opening and closing” the block valves on the flush out lines V102, V103, 2216 Sensors: use of temporary instruments, Use a clamp-on ammeter to measure the amps to the pump motor and compare with design or usual value, 1855 Use a clamp-on ammeter, portable voltmeter and power-factor meters and calculate the power drawn by the pump motor and compare with design or usual value, 1648 Call to vendors, suppliers or licensee, Pump supplier. what might be going on here?, 1199 Open and inspect, The line between the thickener exit tee and the strainer; is it clear?, 1289 Strainer, 1576 Isolate, flush and when safe, open and inspect pump, 1923 .

Take “corrective” action, Shut-off the polymer to the filter, 798 Shut down. Block off and drain. Clean out the crud in the line from the bottom of the thickener to the screen, 249 Replace the block valve V101 on the bottom from the idle thickener, 2300 Install a recording ammeter on the pump motor, 2800

8.2 Cases to Help you Polish Your Skill

Case ’20: The case of the fussy flocculator pump (problem supplied by Jonathan Yip, B. Eng. McMaster University, 1997) [3, pump, storage tanks, flocculation, general] Pump 40P002 is a centrifugal pump that transfers wastewater from the buffer tank to the flocculation tank. The liquid overflows into the Dissolved Air Flotation, DAF, unit. Figure 8-3 shows the system. One day the operator noticed that the level in the flocculation tank was lower than normal and the resulting overflow to the DAF was less than expected. Pump 40 P002 just wasn’t performing as it should! Get it fixed. The manual ball valve on the exit line is wide open.

Figure 8-3

The flocculation system for Case ’20.

Case ’20: The case of the fussy flocculator pump .

.

. . .

Immediate action for safety and hazard elimination, Put on safe-park, 416 Safety interlock shut down, 999 SIS plus evacuation, 1995 IS and IS NOT, What, 195 Where, 1892 When, 2744 Why? why? why?, Goal: “Get the overflow to DAF to expected valve”, 2596 Weather, Today and past, 2497 Maintenance: turnaround, When and what done?, 2987

Maintenance: routine, When and what done?, 2624 .

What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties), Centrifugal pump, 1297 Vendor files: Centrifugal pump, 2525 Trouble-shooting files, 2798 .

What is current operation

Process operators, Anything change in this portion of the plant?, 604 Operating procedures, For pumping, 898 .

Check with colleagues about hypotheses, 2176

289

290

8 Prescription for Improvement: Put it all Together

Call to others on-site, Operators of DAF: is the flow less than expected?, 779 Visit site, read present values, observe and sense. PI gauge, 2322 Listen to the pump for sounds of cavitation, 2343 Is drive motor on the pump running hot? Touch with gloved hand, 2420 Is the pump running hot? Touch with gloved hand, 2411 Does the handle on the ball valve, on the exit line, move easily and smoothly?, 2243 Does the handle on the ball valve, on the entrance into the flocculation tank, move easily and smoothly?, 2460 Current level in the buffer tank; usual value?, 2029 Check diagram and P&ID versus what’s out on the plant, 1342 On-site simple tests: Close the discharge valve on the pump exit and read PI, 2764 Backflush the pigtail on the pressure gauge with city water to clear any blockage, 1099 Tachometer reading on drive shaft rpm for the pump, 2927 Stethoscope on the check valve, 2877 Measure motor amps with clamp-on ammeter, calculate power assuming volts, power factor and density, 2506 Gather data for key calculations Equipment performance, Pump, 641 Sensors: check response to change Pressure gauge PI response to partial closure of valve on exit line, 112 Sensors: calibrate, Pressure gauge PI, 413 Samples and measurements, Sample solution in buffer tank. Compare the density and composition with specifications, 676 Sample solution in the flocculation tank. Compare the density and composition with specifications and with specifications for solution in buffer tank, 932 Open and inspect, Shut down system, isolate. Open and inspect ball valve, 953 Shut down system, isolate. Open and inspect centrifugal pump: clearance between impeller and volute tongue; status of the wear rings; erosion of the impeller, key between the shaft and the impeller. Pluggage?, 1473 Shut down system, isolate. Open and inspect line to the pressure gauge for dirt and pluggage causing sluggish response, 1630 Shut down system, isolate. Open and inspect exit line from the valve to the flocculation tank, 2190 Shut down system, isolate. Open and inspect check valve, 2894 .

Take “corrective” action, Change the pressure gauge, 2569 Replace the ball valve on pump exit, 2852 Replace the check valve, 2965 Replace the impeller in the pump, 1727 Realign drive motor and pump shaft, 1644

8.2 Cases to Help you Polish Your Skill

(problem supplied by Mark Argentino, B. Eng. McMaster University, 1981) [3, refinery, flare system, compressor, refinery] The flare system at your refinery works as follows. When there is a high-pressure build-up in a vessel, tower, or exchanger a relief valve opens and the high pressure vapors, and maybe some liquid, flow into a pipe network that eventually ties into a common header that flows into a large knockout drum where the liquid is removed and the vapors are drawn overhead. Other sources of vapor in the flare system could be off-specification products sent to flare, hazardous vapors educted from pumps, or any hydrocarbon or non-hydrocarbon sources in the refinery that are not of any product value but cannot be vented to atmosphere. The typical composition of the flare gases is as follows Case ’21: The case of the flashy flare

CH4 C2H6 C3H8 C3H7

20% 15% 5% 5%

C4 s C5+ N2 H2S

3% 1% 50% 1%

The vapors then pass through a seal pot that is a vertical cylindrical drum with a smaller cylinder in the center. The vapors flow down through the central cylinder, out the bottom and into the annulus where they bubble through a liquid and are fed to the burner, flare, where they are burned. To visualize the operation of the seal, consider an analogous situation of a large drinking straw in a glass of water. When no air is blown down the straw, the water in the glass is at the same level as the water in the straw. As one blows down the straw the water level drops in the straw and rises in the water glass. This liquid differential in the system forms a certain liquid head that must be overcome for one to blow air out of the straw and bubble it through the water. As with the drinking-glass analogy, the seal pot has a certain level of liquid in it so that pressure must be applied on the inlet line (the straw in our analogy) for any vapors to pass through the seal and be burned. In this way it serves as a seal. The process is given in Figure 8-4. Last turnaround a flare-gas compressor was added to the system. The compressor takes suction just downstream of the knockout drum. The function of the compressor is to recompress the flare gases from 106.5 kPa abs to 480 kPa abs so that the vapor can be recovered as fuel, which along with purchased natural gas, is used in the boilers and fired heaters. The compressor has a spill back or kick back valve and line, which is a small pipe that reroutes some of the compressed vapors in the discharge line, via a control valve on pressure control, back to the compressor suction, and hence to the flare line to keep the suction pressure at 106.5 kPa abs. The compressor will shut down on high suction pressure = 112 kPa abs. In the cold Canadian winter kerosene is used as the sealant liquid. Nevertheless, there are freeze ups in lines and equipment failures; hence frequent flaring when there is a high release of vapors to the flare line when the pressure builds up in the flare line. The pressure builds up in the flare line because the compressor will only pump a fixed maximum amount of vapors. If there is a higher flow of flare gases then the restriction of flow at the seal pot causes a pressure increase up to the blowpoint. The maximum blow pressure is 107.8 kPa. That is the level of liquid sealing the dip tube in the seal pot is equivalent to a 6.8 kPa differential.

291

292

8 Prescription for Improvement: Put it all Together Flare gas Flare

Pressure relief from equipment & other sources to the flare system

KO pot

Liquid

PIC 2

5.5 kPagague

PIC 1

Kickback loop

PIC 3

Seal pot

Recompressed vapors for reprocessing

6.8 kPa

Flare gas compressor

To flare

To flare 6.8 kPa

When pressure exceeds 6.8 kPa

Figure 8-4

The flare system for Case ’21.

Today it is cold, dry and windy. The pressure gauge on the flare line reads 112 kPa abs; the compressor shuts down. The operator explains that “when the compressor shuts down, the accompanying surge in the flare gas flow is so great that the kerosene is all blown into the flare burner along with the usual flare vapors. The result is that the flare flashes and smokes like a giant fire.” All the neighbors are phoning in with smoke complaints. With the flare seal blown, all the flare gas goes to the flare. When the compressor is down we are losing $1000/h in non-recovered vapors. Fix it. Case ’21: Flashy flare . .

.

. . .

MSDS, 215, Immediate action for safety and hazard elimination, Put on safe-park, 84 Safety interlock shut down, 2341, SIS plus evacuation, 2900 IS and IS NOT: (based on given problem statement), What, 2076 When, 2641 Who, 1992 Where, 1730 Why? why? why?, Goal: “Prevent the compressor from shutting down”, 597 Weather, Today and past, 1505 Maintenance: turnaround, When and what done?, 1395

8.2 Cases to Help you Polish Your Skill

Maintenance: routine, When and what done?, 1451 .

What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Compressor, 75 Control, 471 Seal pot, 9 Knockout drum, 500 Vendor files: Compressor, 852 Commissioning data, P&ID, internal reports, 576 Handbook, Density of kerosene and color, 957 Trouble-shooting files, 625 .

Calculations and estimations (that can be done in the office before special tests are done)

Pressure differential, Height of kerosene in the seal pot = 6.8 kPa?, 1127 Equipment performance, Compressor, 1497 .

What is the current operation

Visit control room: control-room data: values now and from past records, Motors amps on the compressor, 1511 Process operators, 1974 .

Check with colleagues about hypotheses, 2906

Call to others on-site, Call operators on other units to see if excessive pressure in the flare line might be originating on units, 1842 Visit site, read present values, observe and sense. Pressure gauge on flare line, 2382 Pressure gauge by seal pot = pressure gauge on kick-back control= 112 kPa, 2610 Check diagram and P&ID versus what’s out on the plant, 70 On-site simple tests: Atmospheric pressure, 2883 Drain off the KO pot to verify that the liquid level is not > expected, 694 Sensors: check response to change Pressure gauge on flare line, 141 Sensors: use of temporary instruments, Surface sensor for temperature of suction gas to compressor, 410 Use clampon ammeter to measure amps when running under usual conditions, 2487 Sensors: calibrate, Pressure gauge on flare line, 1034 Control system, Control-valve stem moves when suction pressure < 106. 5 kPa abs?, 1284 Call to vendors, licensee, Compressor, 742

293

294

8 Prescription for Improvement: Put it all Together

Samples and measurements, Flare gas: composition, 1240 Liquid density of liquid from the knock-out drum, 1070 Kerosene in seal pot, density, 1590 Kerosene in seal pot, color, 1447 Open and inspect, Open seal pot and check for blockage, 2119 Open and inspect KO pot for frozen ice in the demister, 729 .

Take “corrective” action,

Isolate the seal pot. Attach compressed air to the feed line to the seal pot and blow out line through the flare, 2465 Case ’22: The pH control unit (used courtesy of Scott Lynn, University of California, Berkeley, CA) [4, pumps, control, storage tank, acid-base wastewater treatment] Concentrated hydrochloric acid is used to neutralize caustic wastes being fed to a newly built effluent treatment plant. The process is illustrated in Figure 8-5. The volumetric flowrate of wastes is approximately constant at 12.6 L/s but due to the nature of their source, the concentration varies from 1 to 10 g/L equivalent NaOH. The average is about 5 g/L. Control is usually good, but at times it becomes erratic and occasionally the acid flow stops altogether. Turning off and restarting the acid feed control system usually serves to get the acid flow going again, but this malfunction threatens to shut down the entire plant. Find the bug and get rid of it!

Caustic waste 12.6 L/s

5 cm diam pipe

2.4 m

Mixing tee

To effluent treatment pH IC 1

Storage tank for concentrated HCl 7560 L

6 m vertical

1.25 cm diam PVC Figure 8-5

Feed system to the effluent treatment process for Case ’22.

8.2 Cases to Help you Polish Your Skill

Case ’22: pH pump . .

.

. . .

MSDS, 2250 Immediate action for safety and hazard elimination Put on safe-park, 2167 Safety interlock shut down, 2035 SIS plus evacuation, 1740 IS and IS NOT, What, 2940 When, 2700 Who, 2081 Where, 2358 Why? why? why?, Goal: “To prevent erratic control of pH”, 2579 Weather, Today and past, 1953 Maintenance: turnaround, When and what done?, 2985

Maintenance: routine, When and what done?, 234 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Feed pump, 468 Vendor files: Feed pump, 592 Commissioning data, P&ID, internal reports, 998 Handbook, Density of 30% HCl, 530 Trouble-shooting files, 1326 .

Calculations and estimations based on given information Control system, Feed-forward control, 2909 Valve stiction and hysteresis, 426 Fluid dynamics, Calculate the residence time between the acid injection and pH sensor, 1425 Calculate if the flow is turbulent in the caustic waste line near the mixing tee, 1191 Velocity of waste in the line, 1403

.

What is happening

Visit control room: Process operators, This shift, 2509 Did the motor overload and trip off, 2677 .

Check with colleagues about hypotheses, 97

Call to others on-site, Effluent treatment: variability in flowrate or pH, 1290 Visit site, read present values, observe and sense.

295

296

8 Prescription for Improvement: Put it all Together

Control valve: does the control-valve stem move in response to a signal from the controller, 2002 Check that the block valves on the control valve are open, 1068 Check that the valve on the bypass around the control valve is shut; if not shut it, 1878 Is there a vent on the acid storage tank, 1835 Note the characteristics of the response: cycling? amplitude and frequency?, 1682 Sounds near the pump, 1150 Check diagram and P&ID versus what’s out on the plant, 2129 On-site simple tests: Check valve for stiction: remove backlash by doing a bump of 2 to 3%; then move controller output slowly via a slow ramp or bumps of 0.1%; observe controller output: pressure to the actuator, the valve stem and the pH output. Repeat ramping up and ramping down, 2467 Is period of oscillation in pH close to the time delay, 2799 Manually start with a high acid flowrate demand gradually decrease the flow demand, 700 Gather data for key calculations Pressure profile, For the feed acid from the storage tank to the injection location, 1003 Energy balance: Estimate power required for density of acid, 1499 Equipment performance, Rating pump, 2850 Sensors: check response to change pH, 184 Sensors: use temporary instruments, Use clamp-on ammeter to measure amps to pump, compare with expected, 896 Sensors: calibrate, pH sensor, 850 Control system, Put on manual, 498 Samples and measurements, Sample acid and check that density is 1.48, 1075 Open and inspect, Pump discharge line from pump to injection point, 1700 Pump suction line, 2100 .

Take “corrective” action, Replace pH sensor, 2500 Relocate acid injection line so that the acid enters the top of the caustic waste line, instead of the bottom, 2600 Install a check valve in the discharge of the acid pump, 378 To improve mixing, install a static mixer just after the mixing tee in the caustic waste line, 125

8.2 Cases to Help you Polish Your Skill

Relocate the pH sensor to 3.2 m from the injection point to give a 2-s residence time, 1218 Run cold water over the pump, 1300 (based on Barton and Rogers, 1997) [4, polymerizer, mixer, cooling system, polymer] In the manufacture of polyurethane prepolymer, toluene diisocyanate (TDI), at room temperature, 25 C, was charged to an 8-tonne reactor overnight. It was raining cats and dogs this morning. At the start of the 8 am shift, 1.7 Mg of polyol was added gradually over a 20-minute period as spelled out in the operating procedures. The mixer operated continuously. The temperature was monitored carefully. It rose to 127 C as expected. After 35 minutes the operator in the control room noted that the reactor temperature read 170 C or 43 C hotter than expected. The red light indicated that the stirrer stopped. The operator couldn’t get the stirrer restarted. Shortly thereafter the temperature sensor on the reactor read 200 C. Case ’23: The hot TDI

Case ’23: The hot TDI . .

.

.

. .

MSDS, 2252 Immediate action for safety and hazard elimination, Put on safe-park, 2168 Safety interlock shut down, 2034 SIS plus evacuation, 1739 IS and IS NOT, What, 2535 When, 2755 Who, 2422 Where, 14 Why? why? why?, Goal: “get the reactor temperature to 127 C and the mixer going”, 1596 Weather, Today and past, 1954 Maintenance: turnaround, When and what done?, 2986

Maintenance: routine, When and what done?, 2231 .

What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Custom designed reactor with coolant coils and stirrer, 2468 Commissioning data, P&ID, internal reports, P&ID, 991 Handbook, TDI, 531 Trouble-shooting files, 1325 .

What is happening

Visit control room: control-room sensors and historical data, TI in reactor and TRC in reactor, 1195

297

298

8 Prescription for Improvement: Put it all Together

Pressure in reactor, 2289 Pressure relief on top, 2893 Indication light that shows stirrer is turning; green means yes; red means stopped, 291 Process operators, Was the TDI charged correctly?, 1201 Is the feed cooling water cold?, 1422 Is the cooling-water flowing to the cooling coils?, 1618 Is there a power failure that might cause the mixer to fail?, 1986 Please describe the procedure you used to add the polyol, 2188 .

Check with colleagues about hypotheses, 843

Call to others on-site, Purchasing: did you change suppliers of the TDI or of polyol?, 1743 Utilities: any upsets or changes in the cooling water supplied to our site, 2274 Visit site, read present values, observe and sense. Anything obvious interfering with the mixer shaft preventing it from turning?, 2791 TRC valve position for the cooling-water valve controlling water to the jacket, 2824 Emergency block valve on vent (that bypasses the PCV in case the PCV fails to open under SIS), 723 Signal to the TRC control valve for the cooling water, 1261 Check diagram and P&ID versus what’s out on the plant, No diagram supplied; only verbal description, 1837 On-site simple tests: Glove test on temperature of reactor, 2996 Sensors: check response to change Temperature on reactor, 306 Temperature on cooling water, 49 Control system, Put on manual to control temperature, 651 Sensors: calibrate, Temperature sensors on reactor, 977 Temperature sensors on cooling water, 2736 Samples and measurements, Sample the polyol. Is it within specs?, 2522 Sample the TDI. Is it within specs?, 709 Open and inspect, Isolate, drain, vent, make safe to enter. See if the cooling coils are fouled, 543 Isolate, drain, vent, make safe to enter. See if there is anything preventing the mixer from turning, 1769 .

Take “corrective” action, Replace motor on mixer with motor with double the kW, 1503

8.2 Cases to Help you Polish Your Skill

(courtesy of John Gates, B. Eng. 1968, McMaster University) [4, distillation, adsorption, regeneration dryers, exchangers, ethylene] The part of the ethylene plant that relates to this problem concerns the drying section to remove moisture from the feed gas and the distillation train to separate the gas stream into the desired component section. Drying section: Three alumina dryers are installed. One dryer is regenerated while two are hooked in series on stream. For example, dryer V106 is being regenerated with V107 and 108 removing the moisture in the process stream to less than 4 ppm. Then V107 will be regenerated with V108 and 106 in series and so on. The cycle lasts 12 hours. The dried gas goes to a knockout pot (to remove any entrained material) and then is chilled, in exchangers E107 and 108, before entering the separation towers. During regeneration of the dryers, “fuel gas”, heated with 2.8 MPa steam, flows through the dryer in the direction reverse to normal flow. Once it is through the dryer the regeneration effluent gas is cooled and returned to the fuel-gas system. During regeneration the dryer temperature rises to 190 C and is maintained at this temperature for one hour. Then the fuel gas bypasses the heater and is sent directly to the dryer to cool it. For this plant, the “fuel gas”or “town gas” is purchased from an off-site, independent utility supply pipeline. This gas is primarily methane with some hydrogen, is supplied at a pressure of 1.1 MPa and with a specified moisture content of < 6 ppm. This company has recently been expanding its facilities and pipelines. The dryers are all appropriately manifolded and valved so that any dryer can be regenerated, bypassed or used. The regenerating dryer is separated from the line dryers by a gate valve. The separation is performed in a train of three distillation columns operating at about 3.2 MPa. These are a demethanizer, de-ethanizer and C2 splitter, T101, T102 and T103 respectively. The process is illustrated in Figure 7-1 (Case ’12). The overhead from Tower T101 is condensed with ethylene as refrigerant. The overhead from Tower T102 is condensed with propane as refrigerant. The overhead from Tower T103 is condensed with propylene as the refrigerant. The current situation. At low flowrates of 70 Mg/d of ethylene, we encounter no difficulties with production. Recently, we have had excessive pressure drops across the third column when all conditions were the same except that the production rate had increased to 150 Mg/d. At these higher rates of production, the pressure drop was so large that we could not operate satisfactorily. The lost capacity is worth $20 000/day. We cannot shut the plant down because the rest of the site uses ethylene as a raw material and our current inventories are very low. The usual operating conditions are given in Figure 8-6. Get our inventories up so that the rest of the site can function properly. Case ’24: Low production on the ethylene plant

299

P&ID for the ethylene plant for Case ’24.

8 Prescription for Improvement: Put it all Together

Figure 8-6

300

8.2 Cases to Help you Polish Your Skill

Case ’24: Low ethylene production . .

.

.

.

. .

MSDS, 320 Immediate action for safety and hazard elimination, Put on safe-park, 2948 Safety interlock shut down, 74 SIS plus evacuation, 98 More about the process, About the dryers and dryer cycle, 2888 About the low-temperature condensation, 2514 IS and IS NOT, What, 1800 When, 2200 Where, 2399 Why? why? why?, Goal: “remove the high Dp across the trays in the C2 splitter, T103 at high throughput”, 488 Weather, Today and past, 429 Maintenance: turnaround, When and what done?, 783

Maintenance: routine, When and what done?, 1484 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Exchanger E131: cool regeneration gas to the dryer during last portion of regeneration: water-tube side; regeneration fuel-gas shell side, 1773 Heater E130: heat regenerative fuel gas to dryer, 257 Exchanger E107and E108: precool feed from 38 C to –29 C, 451 Reboiler E115 on bottoms of De-ethanizer, 2130 Demethanizer overhead condenser: process fluid shell side; ethylene refrigeration on the tube side, 2885 Demethanizer reboiler: process stream shell side; propylene on tube side, 2999 Steam trap on regenerative gas heater, 2632 Distillation columns T101, 102 and 103, 2501 Knockout pot, 1777 Vendor files: Heat exchanger, 1569 Dryer system, 280 Steam trap, 824 Utilities: town gas supplier, 501 Commissioning data, P&ID, internal reports, 982 Handbook or Google, Gas hydrates, 1493 Trouble-shooting files, 2992

301

302

8 Prescription for Improvement: Put it all Together

Calculations and estimations based on information given in the problem statement Pressure profile, Exchanger Process fluid pressure relative to utility pressure for E 131, 107, 108, 113, 114 to show direction of leak, 2726 Exchanger Process fluid pressure relative to utility pressure for E 130 to show direction of leak, 159 Exchanger utility pressure relative to atmospheric to show direction of leak, 466 Exchanger process fluid relative to atmospheric, 2429 Column T101 Dp estimate across trays at 150 Mg/d, 2936 Column T102 Dp estimate across trays at 150 Mg/d, 2116 Column T103 Dp estimate across trays at 150 Mg/d, 2112

.

.

What is current operation

Visit control room: read instruments and past records, For 150 Mg/d; Temperatures: top and bottom for demethanizer T101, 1214 For 150 Mg/d; Temperatures: top and bottom for de-ethanizer T102, 435 For 150 Mg/d; Temperatures: top and bottom for C2 splitter T103, 1377 For 150 Mg/d, measured Dp across demethanizer column T101, 2432 For 150 Mg/d, measured Dp across de-ethanizer column T102, 2012 For 150 Mg/d, measured Dp across C2 splitter T103, 2840 Process operators, Any changes in operation other than increase in flowrate?, 2043 .

Check with colleagues about hypotheses, 162

Call to others on-site, Operators of upstream feed gas facilities: upsets? any moisture in feed? any changes in composition at the higher feedrates?, 559 Operators of fuel-gas system; any changes in the fuel gas we are sending you from the regeneration process for our adsorbers?, 675 Operators of the refrigeration units for the ethylene and propylene refrigeration loops: any changes or upsets?, 1129 Utilities: any changes in the 2.8 MPa steam supplied to site, 1286 Visit site, read present values, observe and sense, Look at the flare, 1912 Check diagram and P&ID versus what’s really out on the plant, 1767 On-site simple tests: Drain knockout pot after adsorbers, 2171 Sensors: check response to change Dp and pressure gauges on demethanizer T101, 2540 Dp and pressure gauges on C2 splitter T103, 2232

.

8.2 Cases to Help you Polish Your Skill

Sensors: calibrate, Calibrate the pressure gauges at the top and bottom of the demethanizer column T101, 2834 Calibrate the pressure gauges at the top and bottom of the de-ethanizer column T102, 2507 Calibrate the pressure gauges at the top and bottom of the C2 splitter T103, 371 Call to vendors, licensee, supplier, Call utility supplying town gas about changes, 2032 Call utility supplying town gas about specifications, 2050 Samples and measurements, Sample town gas entering battery limits. Analyze for water content. Samples every 30 min for 2 hours, 892 Sample town gas leaving heater E130 before entering the dryer for regeneration. Analyze for water content. Samples every 30 min for 2 hours, 1000 Sample town gas entering battery limits. Analyze for water content, 527 Sample town gas leaving heater E130 before entering dryer for regeneration. Analyze for water content, 1874 Sample process gas leaving KO pot as feed to tower T101. Analyze for water content, 2387 Sample process gas entering dryers. Analyze for water content, 2880 Sample process gas leaving the last dryer in the series. Analyze for water content, 2545 Perform more complicated tests, Gamma scan near the top of the demethanizer and of the C2 splitter to locate collapsed tray, 2790 Open and inspect, Shut down tower, vent to safety, open access hole near top and look for plugs in the downcomers or on the trays. Demethanizer column T101, 2370 Shut down tower, vent to safety, open access hole near top and look for plugs in the downcomers or on the trays. C2 splitter T103, 2972 Shut down column T102, isolate the reboiler E115 on the de-ethanizer, once conditions are safe, pull bundle, hydraulically pressure test to identify leaks in the tubes or between the tubesheet and the tubes, 1190 Heater E130; isolate, once conditions are safe, pull bundle, hydraulically pressure test to identify leaks in the tubes or between the tubesheet and tubes, 348 .

Take “corrective” action, Replace the tube bundle on reboiler E115 on the De-ethanizer. T102, 818

303

Figure 8-7

NAPHTHA FEED

E201

FRC 100

FURNACE

TI 201

TI

TI 202

E203

TO OTHER UNITS

E202

The naphtha process for Case ’25.

PI 100

REACTOR

TI 102

TI 101 TI

LOW PRESSURE SEPARATOR

TI 203

RECYCLE HYDROGEN

LC

V200

FRC 300

TO TOWERS

E204

V300

COOLING WATER FROM COOLING TOWER

EXCESS HYDROGEN

304

8 Prescription for Improvement: Put it all Together

8.2 Cases to Help you Polish Your Skill

[4, reformer, furnace, exchangers, pumps; refinery] In the reforming process of naphtha, naphthenes are dehydrogenated (cyclohexane to benzene and hydrogen) and paraffins are isomerized (n-hexane to 2.methylpentane) and dehydrocyclized (n-hexane to benzene and hydrogen). In the reactor, a relatively high hydrogen:hydrocarbon feed ratio is maintained to minimize coking. The exit gas from the reformer is a very useful source of heat at a temperature of about 415 C and is used to heat several streams for other units. Figure 8-7 shows the system. To keep up-to-date, we changed the catalyst so that the feed hydrogen: hydrocarbon flowrate could be reduced to about half its original value. Not only does this allow a reduction in the amount of hydrogen recycled, but this allows an increase in the naphtha flowrate to keep the space velocity in the current reactor the same as it was before we changed to the new catalyst. We have just started up the unit after the new catalyst has been installed. Immediately the operators of the other units phone to say that their process streams are no longer getting the amount of heating (via E202 and E 203) they used to get from the reformer before the shutdown. The exchangers are delinquent! Fix the problem. Case ’25: The case of the delinquent exchangers

Case ’25: The case of the delinquent exchangers on the naphtha reformer . .

.

.

. .

MSDS, 473 Immediate action for safety and hazard elimination, Put on safe-park, 243 Safety interlock shut down, 2774 SIS plus evacuation, 2966 IS and IS NOT, What, 2527 When, 27 Who, 496 Where, 564 Why? why? why? Goal: “to provide the usual amount of heat to the process streams of other units” 738 Weather, Today and past, 832 Maintenance: turnaround, When and what done? 951

Maintenance: routine, When and what done? 1351 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Exchangers: E201, 202, 203, 1036 Naphtha pump, 1474 Furnace, 1235 Reformer, 1526

305

306

8 Prescription for Improvement: Put it all Together

Vendor files: Heat exchangers E201, 202, 203, 1636 Pump, 1975 Commissioning data, P&ID, internal reports, 2941 Handbook, Approximate thermal properties of hydrogen and hydrocarbon vapors in the stream. Approx. Prandtl numbers, 2551 Trouble-shooting files, 2998 .

Calculations and estimations

Equipment performance, Exchangers E201, 202, 203, 2451 .

What is the current operation

Visit control room: control-room data, Into reformer, TI 203, 2111 Ex reformer, TI 101, 2404 Ex exchanger E203, TI 102, 356 Naphtha flow, FRC 100, 65 Process operators, On other units: Has anything else changed for you because of the shutdown, 922 On the reformer unit: anything changed with the new catalyst? 503 Operating procedures, Please walk me through the startup procedure you used, 874 .

Check with colleagues about hypotheses, 2780

Visit site, read present values, observe and sense. TI 201, 686 TI 202, 2418 PI 100, 2122 T on exit stream to other units from E202, 2816 T on exit stream to other units from E203, 2503 Visual check around exchanger E202 and E203, 2383 Look at the flare, 2645 Check diagram and P&ID versus what’s really out on the plant, 2228 On-site simple tests: Vent exchanger E202 to check if inert gas trapped in shell side, use gas sniffer designed to test for expected gas, 2059 Vent exchanger E203 to check if inert gas trapped in shell side, use gas sniffer designed to test for expected gas, 2212 Vent exchanger E201 to check if inert gas trapped in shell side, use gas sniffer designed to test for expected gas, 1751 Gather data for key calculations Pressure profile, Reformer exit to low-pressure separator, 2951 Exchanger: tube side, 2016

8.2 Cases to Help you Polish Your Skill

Naphtha from pump to reformer, 2316 Mass balance, Space velocity and mass flowrate through the reformer and the downstream exchangers, 2294 Energy balance, Overall energy balance on the system, 1086 Equipment performance, Check heat exchanged on exchanger E202 based on new conditions, 1807 Check heat exchanged on exchanger E203 based on new conditions, 1550 Sensors: check response to change TI 202, 1901 TI 203, 11 Sensors: use of temporary instruments, Surface or laser thermometer on the line between E202 and E203, 313 Sensors: calibrate, TI 202, 600 TI 201, 668 T on exit stream to other units from E202, 1276 T on exit stream to other units from E203, 1489 Control system, Controls related to the reformer; put on manual and check that working correctly, 851 FRC 100: put on manual and check that it is working well, 701 Call to vendors, licensee, Phone call to catalyst vendor: any unexpected behavior difference between new catalyst and old catalyst given the conditions we are using, 392 Samples and measurements, Reformer exit gas and check for hydrogen content, 1956 Open and inspect, For exchanger E202: pull the bundle and measure baffle spacing; check sealing strips, check that baffles are not loose. Look for fouling; check that vent works. Look for condensed liquid, 1555 For exchanger E201: pull the bundle and measure baffle spacing; check sealing strips, check that baffles are not loose. Look for fouling; check that vent works, 2866 For exchanger E203: pull the bundle and measure baffle spacing; check sealing strips, check that baffles are not loose. Look for fouling; check that vent works, 2586 .

Take “corrective” action, Open exchanger E202 and decrease the baffle spacing to increased the vapor flow across the bundle by 20%, 2711

307

Figure 8-8

Air Intake

T 10

PROCESS FLUID

The fired heater process for Case ’26.

Feed tank

FC 1

FC 5

Fuel oil

V300

P3

F 2

TC 3

T4

C.W.

F 7

Product tank

T7

308

8 Prescription for Improvement: Put it all Together

8.2 Cases to Help you Polish Your Skill

(used with permission from T. E. Marlin) [5, furnace, pump; general] The process shown in Figure 8-8 consists of a fired heater, which raises the temperature of a hydrocarbon stream via convective and radiative heat transfer, and a packed-bed reactor. The process has been working well for over a year. Recently, the market for the product is growing, and the plant would like to maximize the production rate. Therefore, the operators have been slowly increasing the feedrate. You happened to be in the control room one morning to collect some data when an operator asks for your assistance. She shows you the trend of selected variables seen in Figure 8-9. She is quite concerned; you better solve this problem fast! Case ’26: The drooping temperatures

T3

F2

FC 1

time 1 minute Figure 8-9

Trend plot of some process variables for Case ’26.

Case ’26: The case of the drooping temperature . .

.

. .

MSDS, 283 Immediate action for safety and hazard elimination, Put on safe-park, 340 Safety interlock shut down, 15 SIS plus evacuation, 476 IS and IS NOT, What, 1487 When, 2995 Who, 2636 Where, 218 Why? Why? Why? Goal: to operate plant safely and efficiently, 1444 Weather, Today and past, 2626

309

310

8 Prescription for Improvement: Put it all Together .

Maintenance: turnaround, When and what done? 1389

Maintenance: routine, When and what done? 1130 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Pump, 1999 Blower, 1782 Heat exchanger, furnace/ fired heater, 1532 Vendor files: Blower, 1005 Pump, 1381 Commissioning data, P&ID, internal reports, Packed-bed reactor: data about the capacity: can it handle the increased feedrate? 1401 Trouble-shooting files, 1729 .

What is the current operation

Visit control room: control-room data, Flow: Process liquid in. FC-1, 955 Flow: process liquid ex reactor F-7, 2851 Flow: Fuel gas F-2, 2542 Flow: Air to furnace FC-5, 1225 Temperature: exit process liquid TC-3, 1092 Process operators, This shift, 1452 Previous shift, 1255 Operating procedures, Please guide me through the procedures you use, 1894 .

Check with colleagues about hypotheses, 1111

Visit site, read present values, observe and sense. Temperature: process liquid out. T4, 1623 Pressure: on furnace P3, 2191 Stack gas: Look at quality of flue gas out of stack, 2101 Pump: process liquid: sound like cavitation? 2046 Valve: control: fuel oil, 2881 Valve: control air to furnace, 381 Temperature of the process fluid entering fired heater, T-10, 2428 Block valves around the fan fully open? 2140 Readings of the draft gauges upstream and downstream of the damper, 2878 Position of the damper; compare with expected 1/3 closed, 1279 Inspect orifice tabs the three flow meters to ensure that the plates were installed with sharp edge upstream, 1139 Check diagram and P&ID versus what’s really out on the plant, 171

8.2 Cases to Help you Polish Your Skill

On-site simple tests, Use laser temperature sensor or pyrometer to measure temperature of tubes in the radiant section, 443 Use laser temperature sensor or pyrometer to measure temperature of the tubes in the convection section using the observation port in that section, 820 Shut the valve on the process pump discharge and compare the measured pressure with the head-capacity curve for that pump, 690 Change set point on air flow, FC-5, and note response, 553 .

Gather data for key calculations

Pressure profile, On process pipe through the furnace, reactor and product tank, 1902 On fuel-oil line bringing fuel into the furnace, 1601 On the combustion air line from the intake to the burner, 1801 Direction of leak of process fluid versus furnace, 661 Direction of leak of air: into the furnace or out of furnace? check pressure P3, 897 Mass balance, On process liquid, 2301 Energy balance: On combustion: temperatures, 2003 On combustion: excess air, 101 Equipment performance, Heat exchanger, 307 Air blower, 150 Process liquid pump, 801 Sensors: calibrate, Temperature: process liquid out. T4, 287 Pressure: on furnace P3, 469 Control system, Temperature: Controller: Signal output from controller TC/3, 2574 Put the process fluid temperature controller, TC-3, on manual control to control under the high flowrate conditions, 2740 Air-flow controller: signal output from controller FC/5, 2077 Samples and measurements, Oxygen in flue gas, 929 Open and inspect, Furnace, 1431 Blower, 1189 .

Take “corrective” action, Put TC3 on manual. Reduce output from valve V300, 1938 Install sensor in the flue gas to measure percentage oxygen, 2379

311

312

8 Prescription for Improvement: Put it all Together

Case ’27: The IPA column [5, distillation column plus auxiliary equipment; petrochemical or refinery] Recently, the water-cooled condenser on the IPA column was replaced by an aircooled condenser. Variable-pitch fans underneath the horizontal bank of finned tubes direct air upwards across the tubes. Since the installation, the ratio of IPA out to IPA fed to the column is 0.70. Previously we could account of 0.995. Where is the IPA going? This costs us $4,000/h. Case ’27: The IPA column: where is it going? . .

. .

. . .

MSDS, 2529 Immediate action for safety and hazard elimination, Put on safe-park, 1 Safety interlock shut down, 271 SIS plus evacuation, 122 More about the process, 2134 IS and IS NOT: What, 398 When, 1758 Who, 1002 Where, 1708 Why? why? why? Goal: “to prevent the apparent loss of IPA.”, 2162 Weather, Today and past, 447 Maintenance: turnaround, When and what done? 1259

Maintenance: routine, When and what done? 1589 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Fan: control system on the air flow to the condenser, 284 Condenser, air-cooled, 126 Distillation column, 2000 Lines insulated? 485 Vendor files: Condenser, air cooled, 2181 Commissioning data, P&ID, internal reports, Commissioning report of tests done before startup, 2993 Handbook, Pertinent properties of IPA, 1166 Trouble-shooting files, 1017 .

.

Calculations and estimations: more data are needed before calculations can be made. What is the current operation

Visit control room: control-room data: values now and from past records, Temperature at the top of the column, 1501 Temperature on liquid exit from the condenser, 1685

8.2 Cases to Help you Polish Your Skill

Pressure at the top of the tower, 1722 Pressure on the overhead receiver, 777 Level in the overhead receiver, 521 Reflux flowrate, 1385 Feedrate change? 2394 Bottoms feedrate change? 2846 Process operators This shift, 1179 Previous shift, 1356 .

Check with colleagues about hypotheses, 2926

Call to others on-site Of upstream suppliers of feed to IPA unit, 86 Of downstream vent scrubber unit, 1291 Of downstream processors of IPA, 377 Visit site, read present values, observe and sense. Hot exhaust air recirculation to the intake of the air-cooled condenser, 18 Sounds around the condenser, 338 Look at the flare, 239 Check diagram and P&ID versus what’s really out on the plant, Same information as given in more about the process, 2134 On-site simple tests: Temperatures and humidities of inlet and exit air for the air-cooled condenser, 106 Tape accessible flanges on the line to the top of the condenser and after the condenser. Test for leaks with soap solution. Test valve stems with soap solution, 1879 Record the vertical dimensions around the exit of the condenser, 1135 Gather data for key calculations Pressure profile, On process pipe from the top of the column to the overhead receiver, 926 Across condenser, 1062 Mass balance, On process liquid, IPA, 1508 Energy balance: Over the condenser: heat extracted in air = heat of condensation for all IPA? 1026 Equipment performance, Rating of air-cooled condenser, 844 Fans, 632 Column, 670 Sensors: check response to change Temperature at top of column, 50 Temperature of the IPA leaving the condenser, 200 Sensors: use of temporary instruments, Temperature and humidity of inlet and exhaust air for the condenser, 1275

313

314

8 Prescription for Improvement: Put it all Together

Air velocity to the face of the condenser for the maximum and minimum blade pitch, 1423 Sensors: calibrate, Temperature at top of the column, 351 Temperature of IPA leaving condenser, 1808 Control system, For fan, put on manual and increase blade pitch to maximum, 1069 Ask control specialist about the quality of this type of control, 311 Use clamp-on ammeters, voltmeter and power-factor meters to obtain data to estimate power to the fan, 475 Call to vendors, licensee, Air-cooled condenser. Describe symptoms to the vendor, 1517 Samples and measurements Sample bottoms of column; measure concentration of IPA and compare with past data, 1658 Sample feed to column; measure concentration of IPA and compare with past data, 1172 Concentration: vent from the overhead receiver; concentration IPA, single sample, 222 Concentration: vent from the overhead receiver; concentration IPA, ten samples taken at one-minute intervals, 152 Concentration: liquid from bottoms; IPA concentration. Three samples, at 2min intervals, 461 Concentration: feed to the column; IPA concentration. Three samples, at 2min intervals, 956 Open and inspect, Air-cooled condenser: visual inspection plus air and water pressure tests, 1652 Fan and fan blades, 251 .

Take “corrective” action, Run cold water over the tubes for an 8-hour shift, 1916 Tip the tube bank so that it is no longer level “so that the condensate can run out easily”, 495 Repipe exit line so that the syphon is removed. “The condensate flows directly from the level of the bottom tubes to the overhead receiver”, 1943

Case ’28: The boiler feed heater (adapted from case of P. L. Silveston, University of Waterloo) [5, shell and tube exchanger, steam heated, condensate traps; general] Waste flash steam from the ethyl acetate plant is saturated at slightly above atmospheric pressure. It is sent to the shell side of a shell and tube exchanger to preheat boiler feed water to 70 C for a nearby boiler house. Condensate is withdrawn through a thermodynamic steam trap at the bottom of the shell. The water flows through 1.9-cm nominal tubes. There are 100 tubes. See Figure 8-10. “When the

8.2 Cases to Help you Polish Your Skill

system was put into operation 3 hours ago everything worked fine,” says the supervisor. “Now, however, the exit boiler feed water is 42 C instead of the design value. What do we do? This difficulty is costing us extra fuel to vaporize the water in the boiler. “ Fix it.

STEAM TEMP

BOILER FEED WATER

PI 100 TI 100

CONDENSATE RETURN TO HEADER

TO BOILER

DRAIN VALVE

Figure 8-10

CLOSED CLOSED

The boiler feedwater heater for Case ’28.

Case ’28: The boiler water preheater . .

.

. . .

MSDS, 349 Immediate action for safety and hazard elimination, Put on safe-park, 6 Safety interlock shut down, 319 SIS plus evacuation, 266 IS and IS NOT: (based on given problem statement), What, 2861 When, 1320 Where, 2391 Why? why? why?, Goal: “to heat the boiler water to 70 C”, 254 Weather, Today and past, 491 What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Exchanger, 2765 Float steam trap, 995 Piping: steam, 1990 Piping: water, 1270 Handbook, Approximate thermal properties of steam and ethyl acetate. Approx. Prandtl numbers. Saturation temperature for assumed 200 kPa g steam, 1504 Trouble-shooting files, 224

315

316

8 Prescription for Improvement: Put it all Together

Calculations and estimations (that can be done in the office before special tests are done and based on information in the problem statement and rules of thumb) Flowrate, Estimate flow of boiler feed water based on tube area and estimated velocity in the tubes, 1237 Energy, Difference in heat load, assuming inlet water temperature = 18 C, 1788 Heat transfer rates at new and design conditions based on assumed steam pressure of 200 kPa g, 2199 Estimate steam condensation, 2699

.

.

What is the current operation

Visit control room: control-room data: values now and from past records. Data only available on site Process operators, This shift, 1329 .

Check with colleagues about hypotheses, 2686

Visit site, read present values, observe and sense. Inlet steam pressure, 323 Inlet steam temperature, 177 Check diagram and P&ID versus what’s really out on the plant, 2126 On-site simple tests: Open bypass on the steam trap, 477 Open air vent on top of the exchanger; leave the vent open for 10 min. Read the temperature gauge on the exit boiler feed, 364 After the vent has been opened for 10 min, close the vent and read the boiler feed exit temperature after 3 h of operation, 876 At the end furthest from the steam inlet, remove a vertical section of the insulation about 10–30 cm wide. Tap the side with metal moving up the exchanger and listen for a change in sound indicating a liquid-vapor interface, 617 Collect data for key calculations Pressure profile, Exchanger: shell side, 1067 Exchanger: tube side, 66 Steam: from ethyl acetate plant to header to boiler feed heater, 465 Mass balance, On steam-condensate; measure the condensate by measuring the volume collected over a timed period, 2005 Energy balance: sink= source, Amount of steam= amount of heat picked up by water, 108 Equipment performance, Rating exchanger, 144 Type of steam trap, 1632 Upstream Y strainer on condensate line upstream of the trap, 1930

8.2 Cases to Help you Polish Your Skill

Sensors: check response to change Steam temp TI 100, 3 Exit water temperature, 210 Sensors: use of temporary instruments, Water temp in, 502 Sensors: calibrate, Steam temp TI 100, 368 Exit water temp, 635 Call to vendors, licensee, Heat exchanger, 1519 Steam traps, 1533 Samples and measurements, Measure amount of condensate to get a measure of the steam. Compare the energy loss from the steam to the energy gained by boiler water, 388 Open and inspect, Exchanger: visually inspect. Water and air tests for leaks, 1577 Exchanger: pull the bundle and measure baffle spacing; check sealing strips, check that baffles are not loose, 57 .

Take “corrective” action, Replace the float trap with an inverted bucket trap with an air vent, 1924 Operate with the bypass around the trap partially open, 480 Stop operation; pull the bundle and clean tubeside and shell side, 1944

(courtesy of W. K. Taylor, B. Eng. McMaster, 1966) [5, reactor, compressor, separator; ammonia] Ammonia is produced on two interconnected reactor loops as given in Figure 8-11. Feed gas consists of hydrogen and nitrogen in the proper 3:1 ratio with about 1% methane as an inert. In this ammonia-synthesis reaction about 10% conversion occurs per pass through the reactor. Feed gas is compressed to 34.5 MPa abs and fed to a common header that feeds two reactor loops. Liquid product is condensed and removed from the system; gas is recycled back to the loop via the recycle stage compression. The reactor operates at 500 C. There is an internal gas-gas heat-exchanger within the reactor. The DT due to exothermic reaction is about 50 C. Each compressor is a multistage reciprocating constant speed machine rated at about 3000 kW. Bypass valve B is operated to control the Dp across the recycle stage that must not exceed 3.5 MPa. Opening valve B lowers the Dp and the flow of recycle gas to the loop. The recycle flow is about five times the flow of fresh feed. Bypass valve A is operated to trim the loop: closing this valve forces more gas over to the reactor; opening the valve causes gas to bypass the loop. Valve A is used to control the reaction temperature. If too much gas is fed to the reactor and the catalyst is inactive, the high flow might extinguish the reaction. Similarly if the flow to the reactor is too low, the reaction will go further because of the longer reaction time; the reactor will overheat because there is not enough flow to carry away the heat of reaction. Normally valve A is open slightly during plant operation. Methane is an inert coming in with the feed. The methane concentration is kept about 15% in the loop gas to the reactors by maintaining a small purge.

Case ’29: The Reluctant reactor

317

Figure 8-11

Feed gas

3 stages of compresson

The Ammonia synthesis reactors.

Kickback

South compressor

LIC

Kick back

B

FR 1

Feed gas

3 stages of compression

Liquid ammonia

Cooling water

Reactor

Electric startup heater

Manual Indicator controller, A

LRC

Refrigerant

Purge

H2, AR 1

South loop

B

Recycle compressor

North compressor

Manual Indicator controller, A

LRC

Refrigerant

Purge

H2, AR 1

North loop

LIC

Liquid ammonia

Cooling water

Reactor

FR 1

318

8 Prescription for Improvement: Put it all Together

8.2 Cases to Help you Polish Your Skill

The pressures, levels in the separators and the temperature profile in the reactor are shown in the control room. The design provides operating flexibility. If one compressor breaks down the other machine can feed both loops thus keeping the reactors at operating temperatures while repairs are done. This avoids costly startup expense. Isolation block valves on the compressors are not shown on Figure 8-11. Furthermore, both loops are equalized in pressure thus evening out any slight variations introduced by the compressors. The problem: The plants are in the final phase of startup after a turnaround shutdown. The compressors are sending feed gas and recycled gas to the reactors. Startup pressure is 7 MPa. The electric cal rod heaters, used to heat up the reactor catalyst bed during startup, are on. Heat-up normally proceeds at 50 C/h and at 8 am the reactors were up to 300 C. By noon the reactors should be at 500 C and in steady production and the plant in normal operation with all the gas vents closed. It is now 3 pm and the loops have only heated up 25 C to 325 C and have had no increase in temperature over the past hour. “Get these plants producing. This is costing us $20 000/h. “ Case ’29: The reluctant reactor . .

. .

. . . .

MSDS, 352 Immediate action for safety and hazard elimination, Put on safe-park, 48 Safety interlock shut down, 2482 SIS plus evacuation, 2018 More about the process, 483 IS and IS NOT: (based on given problem statement), What, 1319 When, 996 Who, 17 Where, 1978 Why? why? why?, Goal: “to heat up the reactors to 500 C”, 446 Weather, Today and past, 2991 Maintenance: turnaround, When and what done?, 2512 What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Reactor, 2452 Internal heat exchanger, 2010 Reciprocating compressor, 1832 Refrigeration system, 1507 Condensers: refrigerant, 1733 Condensers: water, 1960 Gas-liquid separators, 1600 Valves A and B, 1011

319

320

8 Prescription for Improvement: Put it all Together

Vendor files: Reciprocating compressors, 178 Refrigeration system, 484 Commissioning data, P&ID, internal reports, 452 Handbook, Thermal properties of hydrogen, nitrogen and ammonia, 385 Trouble-shooting files, 751 Calculations and estimations (that can be done in the office before special tests are done) Mass balance, Purge rate and methane buildup, 624 Energy balance: Energy transfer from Cal Rod heaters for startup (from experience), 532

.

.

What is the current operation

Visit control room: control-room data: values now and from past records, Pressure at exit of compressors, in the loop, 1459 Hydrogen concentration in the North loop, 1412 Hydrogen concentration in the South loop, 1117 Temperature in the catalyst bed, North reactor, 1048 Temperature in the catalyst bed, South reactor, 1546 Temperature leaving catalyst bed, North reactor, 1815 Temperature leaving catalyst bed, South reactor, 1946 Dp across the reactor, North loop, 1664 Dp across the reactor, South loop, 2415 Flow of liquid ammonia from North loop, 2897 Flow of liquid ammonia from South loop, 2959 Process operators, Anything surprising other than the temperatures?, 2558 Refrigeration condensers colder than usual?, 2226 Cooling-water condensers colder than usual?, 1720 Operating procedures, Please tell me about the startup procedures you used, 1969 .

Check with colleagues about hypotheses,

Visit site, read present values, observe and sense. Amps to cal rod heaters, 1564 Check that sample valves are closed, 1789 Check diagram and the P&ID versus what’s actually out on the plant, 1216 On-site simple tests: Increase the pressure in the loop to about 10 MPa, 1362 Control system, Over ride indicator control on board; go to site and manually open valve A fully. North loop. Check temperature after one hour, 1073 Over ride indicator control on board; go to site and manually open valve A fully. South loop. Check temperature after one hour, 912 Sensors: calibrate, Temperature sensors, reactor, North loop, 577 Temperature sensors reactor, South loop, 2836

8.2 Cases to Help you Polish Your Skill

Samples and measurements, Cooling water, North loop condenser. Analyze for ammonia, 2601 Feed gas. Analyze for hydrogen concentration, 2327 Continuity check on cal rod for reactor in North loop, 333 Continuity check on cal rod for reactor in North loop, 73 Open and inspect, Open and inspect the reactor and especially the top with the cal rod heaters. North loop, 277 Open and inspect the reactor and especially the top with the cal rod heaters. North loop, 1281 .

Take “corrective” action, Open the kickback to reduce the pressure in the loop, 1085 Open the purge line to the maximum, 660 Shut down one compressor and provide loop gas to both North and South loops from one compressor. Allow one hour of operation, 640

(courtesy Esso Chemicals) [6, distillation column plus auxiliaries; general] The 20-tray column, operating at 520 kPa g, has just started up for the first time. The control configuration is illustrated below. The pump can deliver only 2/3 of the design value of the reflux, and although the reflux valve is fully open, the level in the accumulator continues to increase. Do something quickly before the reflux drum is full and sort out the problem. The financial penalty is high because of the clauses in the construction contract for producing specification product within the commissioning period, and because of insurance issues and government regulations. Figure 8-12 illustrates the system. Case ’30: The case of the reluctant reflux

PSV 1

PC 100

TC 200

PI 201

LC 202

Figure 8-12

The overhead system for the column in Case ’30.

To Flare

321

322

8 Prescription for Improvement: Put it all Together

Case ’30: Reluctant reflux . .

.

. . .

MSDS, 272 Immediate action for safety and hazard elimination, Put on safe-park, 450 Safety interlock shut down, 972 SIS plus evacuation, 780 IS and IS NOT: (based on given problem statement), What, 2930 When, 2621 Where, 2013 Weather, Today and past, 1530 Maintenance: turnaround, When and what done?, 1988 What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Reflux pump, 1058 Condenser, 1450 Reflux drum, 1364 Vendor files: Reflux pump, 1137 Commissioning data, P&ID, internal reports, 1823 Trouble-shooting files, 1828 Calculations and estimations (that can be done in the office before special tests are done) Pressure profile, Recheck NPSH supplied, 993 From reflux drum to column via reflux pump, 825 Equipment performance, reflux pump, 26 Design of pressure-control system, 288 Design of level-control system, 138

.

.

What is the current operation

Visit control room: control-room data: values now and from past records, Overhead temperature, 460 Overhead composition from routine lab analyses, 752 Level in reflux drum, 1864 Process operators, This shift. What’s happening? 1606 Visit site, read present values, observe and sense. Control valve on the reflux pump discharge line: valve position and other features of valve; for example, put in backwards, 1513 Note direction of rotation of the shaft and confirm that this corresponds with the arrow on the casing, 1985 Ask maintenance if the impeller might have been put in backwards, 2484 Difference in pressure between PC 100 and PI 200 and compare with usual Dp across condenser, 2071

8.2 Cases to Help you Polish Your Skill

Level indication on controller 200, 2320 Listen for sounds of cavitation around the reflux pump, 2213 Check diagram and P&ID versus what’s out on the plant, 236 On-site simple tests: Shut the valve on the discharge side of the reflux pump and read the pressure; compare with head-capacity curve, 2499 Listen with a stethoscope to check valve. Note if the direction through the valve is the same as the flow direction, 2960 Turn the valve stems on suction and discharge block valves on the pump, 2664 Sensors: check response to change LC202, 2582 PI 201, 2186 TC 200, 2409 Sensors: calibrate, LC 200 on reflux drum, 681 PI 200; pressure on the reflux drum, 610 Control system, Put pressure-control system on manual, 1875 Put level-control system on manual, 1675 Put temperature control on manual, 1169 Samples and measurements, Feed to column and analyze for the amount of light components, 961 Overhead vapor from column and analyze for heavies, 900 Open and inspect, Control valve, 595 Reflux pump, 29 Suction line from reflux drum to tower, 292 Nozzle where line attaches to column, 407 Line from pump to column, 1375 .

Take “corrective” action, Install larger-size impeller in the pump, 1775 Replace motor on air-cooled condenser, 2339 Replace the level sensor on the reflux drum, 2733 Reduce feed to the column, 2611 Replace the check valve on the discharge line of the pump, 2550 Install vent break line on condenser, 1266

(courtesy of C. J. King, Chemical Engineering Dept., University of California, Berkeley) [6, heat exchangers, boiling, steam; general] Our New Jersey petrochemical complex includes an ethylene plant that supplies 6.8 Mg/h of ethylene through a pipeline to various consumers. It is important that we maintain a steady flow of ethylene to our users, and, as a result, our plant contains a large storage sphere of liquid ethylene. The ethylene must be a vapor, however,

Case ’31: Ethylene product vaporizer

323

324

8 Prescription for Improvement: Put it all Together

when it enters the pipeline and must be at a temperature close to ground temperature (slightly more than 1.6 C) so as to avoid thermal stresses on the pipeline. For these reasons we have installed an ethylene vaporizer between the sphere and the pipeline. The ethylene is vaporized by condensing n-butane, which in turn is vaporized by steam. The cascade vaporization system is required so as to avoid undue thermal stresses across the heat-exchange surfaces. Under normal operation, a small amount of ethylene, about 1.6 Mg/h is sent through the vaporizer, but the vaporizer frequently is called upon to provide more, or all, of the total ethylene supply. Before the vaporizer, the liquid ethylene is pumped up to 4.4 MPa gauge and metered through a flow-control valve; the ethylene pressure in the vaporizer is roughly the pipeline pressure of 3. 9 MPa gauge. The butane pressure controller set point, PIC, can respond to anywhere from 0.58 to 0. 97 MPa. A set point of 0.8 MPa has been used successfully at all ethylene flow rates during the past year, although the outlet ethylene temperature has been slow to recover following a change in the ethylene flow rate. In the past few months we have found it necessary to increase the PIC set point. Even so, we found yesterday when the ethylene unit came down that the vaporizer could not handle the full ethylene flow without tripping the low-temperature shutoff switch at the pipeline entry, which is set at 1.6 C. This situation will cost us $6000/h plus inestimable customer good will if we stop the flow, or else it may well necessitate expensive and time-consuming pipeline repair if we continue. Figure 8-13 illustrates the layout.

Figure 8-13

The ethylene vaporizer in Case ’31.

8.2 Cases to Help you Polish Your Skill

Case ’31: Ethylene product vaporizer . .

. .

. .

MSDS, 34 Immediate action for safety and hazard elimination, Put on safe-park, 2612 Safety interlock shut down, 2714 SIS plus evacuation, 2065 More about the process, 2980 IS and IS NOT: (based on given problem statement), What, 1108 When, 566 Where, 809 Weather, Today and past, 950 Maintenance: turnaround, When and what done?, 1359

Maintenance: routine, When and what done?, 1063 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Butane vaporizer, 609 Ethylene vaporizer, 120 Vendor files: Butane-steam system, 414 Commissioning data, P&ID, internal reports, 293 Handbook, Temperature-vapor pressure data for ethylene and butane, 989 Trouble-shooting files, 967 Calculations and estimations (that can be done in the office before special tests are done) Pressure profile, If there is a leak in the ethylene-butane system, the direction of the leak would be, 672 If there is a leak in the steam-butane system, the direction of the leak would be, 1365 Rate, Butane-steam system: Is the boiling nucleate or film boiling?, 1147 Ethylene-butane system: Is the boiling nucleate or film boiling?, 1717 .

.

What is the current operation

Visit control room: control-room data: values now and from past records, Past records, 2427 Process operators, Please tell me what happened before the system shut down, 2248 Tell me a bit about operation over the past few months, 2832 Call to others on-site, Steam utilities; any changes?, 2952

325

326

8 Prescription for Improvement: Put it all Together

Waste water treatment, 2528 Visit site, read present values, observe and sense. Steam pressure, 2603 Ethylene pressure in pipeline, 2807 Ethylene temperature in pipeline, 2549 PIC setting, 2109 Level in the butane, 2407 Valve position on condensate line and size of control valve relative to line size, 2498 Steam goes into the top channel of the butane evaporator, 2040 Ethylene feed goes into the bottom channel of the ethylene evaporator, 1987 Inspect insulation and look for cracks or holes in the external surface, 1776 Check diagram and P&ID versus what’s out on the plant, 331 On-site simple tests: Open vent on butane evaporator for several minutes; close and observe operation, 1549 Open vent on ethylene evaporator for several minutes; close and observe operation, 1745 Use stethoscope and listen to the sound of the fluid flowing through the condensate valve, 1138 Increase the level of butane in the butane evaporator as shown on the level indicator, 1257 Open the bypass on the condensate control valve, 656 Sensors: check response to change PIC, 699 Level indicator on butane, 105 PI on ethylene line, 353 TI on ethylene line, 2390 Temperature shut-off sensor switch, 2469 Sensors: use of temporary instruments, Use a contact pyrometer to measure temperature of the ethylene pipeline, 2919 Control system, Reduce ethylene flowrate. Put PIC on manual and note response when increase the PIC setting, 2572 Retune the PIC control system, 2020 Sensors: calibrate, PIC, 2154 PI ethylene line, 1553 TI ethylene line, 1963 Level indicator on butane, 1238 Temperature shut-off sensor switch, 1104 Samples and measurements, Butane from the shell side of the evaporator and analyze for ethylene and water contamination, 881

8.2 Cases to Help you Polish Your Skill

Condensate and analyze for butane contamination, 958 Ethylene and analyze for butane contamination, 55 Butane from storage and analyze for contamination; compare with specs, 252 More ambitious tests, Drain all the butane and replace with fresh butane, 1914 Open and inspect, Ethylene evaporator for fouling, 1812 Ethylene evaporator, pressure test and look for leaks, 1541 Butane evaporator, pressure test and look for leaks, 2869 Butane evaporator for fouling, 2055 .

Take “corrective” action, Replace the pressure gauge PIC, 2048 Install new control valve on condensate line; the size is same as line size, 1153

(courtesy of T. E. Marlin, Chemical Engineering, McMaster University) [6, sequence of distillation columns with auxiliaries; depropanizer-debutanizer] The process is the depropanizer-debutanizer described in Case ’8, Chapter 2. A P&ID is given in Figure 2.4 accompanying Case ’8. The operation of the upstream process is being modified to accommodate a new catalyst and modified feed composition. The upstream units have been on-line and operating smoothly for nearly a shift. Suddenly the high-pressure alarm on the debutanizer, column C-9, rudely disrupts the quiet. And you thought everything was going smoothly! Case ’32: The alarming alarm

Case ’32: The alarming alarm . .

. .

.

. .

MSDS, 1495 Immediate action for safety and hazard elimination, Put on safe-park, 111 Safety interlock shut down, 440 SIS plus evacuation, 2934 More about the process, 1010 IS and IS NOT, What, 2681 When, 2520 Who, 316 Where, 1407 Why? Why? Why?, Best goal? To find out why the high-pressure alarm is sounding, 2335 Weather, Today and past, 1060 Maintenance: turnaround, When and what done?, 1921

Maintenance: routine, When and what done?, 2928 .

What should be happening,

327

328

8 Prescription for Improvement: Put it all Together

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Condenser, E-28, 2006 Distillation column, C9: 2367 Thermosyphon reboiler, E-30, 1382 Reflux pump, F-29, 1112 Overhead drum, V-31, 1455 Vendor files: Condenser, reboiler, 1448 Steam traps, 1244 Commissioning data, P&ID, internal reports, 1771 Handbook, Cox charts, 1910 Trouble-shooting files, 1653 .

Calculations and estimations

Energy balance: sink= source, Estimate the steam flow to the reboiler based on the reflux rate and the fact that each kg steam boils 5 kg typical organic, 2132 .

What is current operation

Visit control room: control-room data, Feed to C9, debutanizer, FI/2, 2227 Feed to the C8, depropanizer. FC/1, 2304 Overhead liquid product butane, flowrate FIC/7, 2011 Reflux flowrate, FIC/6, 2461 Pressure drop Dp I/2, 1724 Level bottoms LIC/4, 1957 Level in feed drum, V-31; LIC-5, 1506 Temperature bottoms TI/12, 1870 Temperature top, TI/11, 1152 Analyzer A-1, 1357 Pressure on overhead, PIC/19, 1052 Process operators, This shift, 1408 Call to others on-site, Utilities: any change in the cooling-tower operation that might affect our site, 1323 Visit site, read present values, observe and sense. Column pressure, PI-12, 1173 Pressure relief to flare PSV-3, 717 Temperature of the feed to the column TI-10, 619 Cooling-water temperature in to the condenser, TI-13, 910 Cooling-water temperature out of the condenser, TI-14, 802 Valve position for column feed, FV-2, 258 Valve position on PIC/19; downflow from the condenser to the reflux drum, 13 Signal to PIC/19, 359 Pressure on exit of reflux pump F-29, PI-20, and compare with head-capacity curve at feed flowrate, 135 Listen to the reflux pump F29 for sounds of cavitation, 455

8.2 Cases to Help you Polish Your Skill

Observe whether shaft is rotating for the reflux pump F-29. Check that the direction of rotation is consistent with arrow on the housing, 767 Check diagram and P&ID versus what’s out on the plant, 2769 On-site simple tests: Open bleed valves to fuel on the shell of the condenser for 10 minutes, then shut, 606 Tap the side of the condenser along a vertical line to listen for change in sound associated with liquid level in the condenser, 858 Gather data for key calculations, Pressure profile, Drum V-31 to pumps, F-29, 2863 Dp across reflux pump, converted to head, 2984 Drum V-31, pump F-29 and reflux into column, 2723 Pump F-29, 2594 Thermosyphon reboiler process fluid side, 2531 Sensors: check response to change, Temperature sensor at the top of column, TI11, 963 Pressure sensor at top of column PI-12, 586 Pressure sensor on overhead line for PIC/19; PT-19, 1307 Pressure sensor on reflux pump exit line, PI-20, 1013 Sensors: use of temporary instruments, Measure the surface temperature on the outside of the condenser, 1252 Sensors: calibrate, Temperature sensor at the top of column, TI-11, 1802 Pressure sensor at top of column PI-12, 1997 Pressure sensor on overhead line for PIC/19; PT-19, 1785 Pressure sensor on reflux pump exit line, PI-20, 2202 Control system, Put overhead pressure control, PIC/19; reflux control FIC-6 controllers on manual and try to steady out the column, 2159 Sample and analyze, Feed to the upstream depropanizer for the amount of C4 and C5s, 2103 Concentration of effluent from upstream reactor where catalyst was changed. Analyze for C4 and C5 and compare with previous, 2491 .

Open and inspect, Condenser, 2901 Pressure control valve PV-19, 2752 Reflux pump, F-29, 2602 Line from the top of the column to the condenser, E-28, 2742 Sensor tap for pressure on the overhead line, PT-19, 2818

.

Take “corrective” action Replace the pressure control valve PV-19, 2260 Replace the pressure gauge PT-19, 2139

329

330

8 Prescription for Improvement: Put it all Together

Direct water from the fire hose onto the outside shell of the overhead condenser, 401 Reduce the flowrate to C8 and thus reduce the flowrate to C9, 2054 Case ’33: Chlorine feed regulation (courtesy of Scott Lynn, Chemical Engineering Department, University of California, Berkeley) [6, slurry, pump, control system, storage tank; minerals processing] A copper mine is treating its crushed ore with a dilute solution (5%) of sodium hypochlorite to improve the recovery of molybdenum disulfide by flotation. Sodium hydroxide solution of the appropriate strength is reacted with chlorine gas in a 5.5cm diameter pipe that serves as the reactor. Flow is continuous and relatively constant at 3.15 L/s. The pipe carries the bleach solution to the slurry tank. An oxidation-potential probe at the pipe outlet is used to regulate the flow of chlorine. The system has just been installed, as shown in Figure 8-14 and serious trouble has been encountered with the chlorine feed regulation. The flow of gas can be readily controlled manually but fluctuates wildly when put on automatic. A time recording of the oxidation potential, OPRC, is shown in Figure 8-15. Please correct this problem quickly!

Water

1.5 m

PFC

7.6 m

OPRC

Liquid chlorine 50% NaOH

FRC

Figure 8-14

The chlorine feed system for Case ’33.

Slurry treatment tank

8.2 Cases to Help you Polish Your Skill

1 min 100 mV

Automatic operation Manual operation

Figure 8-15

Sample recording chart for the OPRC for Case ’33.

Case ’33: Chlorine feed regulation . .

.

. .

MSDS, 110 Immediate action for safety and hazard elimination, Put on safe-park, 399 Safety interlock shut down, 846 SIS plus evacuation, 2899 IS and IS NOT: (based on given problem statement), What, 2588 When, 2758 Where, 1859 Weather, Today and past, 1642 Maintenance: turnaround, When and what done?, 1274

Maintenance: routine, When and what done?, 1148 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Orifice plate on sodium hydroxide line, 1027 Orifice plate in water feed line, 697 Control valve on chlorine feed line, 542 Control valve on caustic feed line, 145 Control valve on water feed line, 2794 Caustic feed pump, 2813 Pipe-reactor, 2943 Source of water, 2546 Chlorine injection line, 2079 Commissioning data, P&ID, internal reports, 1621 Handbook, Density and viscosity of 50% sodium hydroxide, 1915 Trouble-shooting files, 1787 .

Calculations and estimations (that can be done in the office before special tests are done) Implications of cycling, observed frequency of fluctuation, 1167

331

332

8 Prescription for Improvement: Put it all Together

Observed amplitude of fluctuation, 1379 Mass balance, Flowrates of caustic and chlorine for reaction, 1096 Fluid mechanics, mixing and residence time, Estimate the Reynolds number in the “reactor”, 1019 Estimate velocity in reactor, 781 Estimate residence time before chlorine addition, 981 Estimate residence time between chlorine injection and OPRC sensor, 2889 Rate, Of reaction to form sodium hypochlorite, 2922 Equipment performance, Control system: check the degrees of freedom, 2682 Control system: and stability, 225 .

What is the current operation

Visit control room: control-room data: values now and from past records, Flowrate of water, 99 Flowrate of caustic, 2848 Check that the output from OPRC fluctuates rapidly and significantly, 2609 Process operators, 345 Call to others on-site, Operators of upstream units providing water: any variation in the composition, temperature or flowrate?, 176 Visit site, read present values, observe and sense. Valve movement of chlorine when operating on automatic, 1349 Valve movement of caustic feed control valve, 1446 Valve movement of the water, 1249 Check that the bypass valves on all control valves are shut and that the block valves are fully open, 2299 Sounds of cavitation in the caustic pump, 2099 Check diagram and P&ID versus what’s out on the plant, 2403 On-site simple tests: Check valve for stiction: remove backlash by doing a bump of 2 to 3%; then move controller output slowly via a slow ramp or bumps of 0.1%; observe controller output: pressure to the actuator, the valve stem and the pH output. Repeat ramping up and ramping down, 1561 Gather data for key calculations Temperature of the water such that the water may be flashing in the orifice meter sensor in the PFC loop, 1547 Check that the oscillation is close to the time delay for the concentration to flow from the mixing point to the analyzer, 1931 Sensors: check response to change OPRC, 616 Sensors: use of temporary instruments, Use a contact pyrometer to measure the temperature variation in feed water, 945

8.2 Cases to Help you Polish Your Skill

Use a contact pyrometer to measure the temperature of the chlorine storage tank to ensure it is < 50 C, 593 Use clamp-on ammeter to measure amps to pump, compare with expected, 62 Sensors: calibrate, OPRC, 439 Control system, Retune control system, 267 Place controller on manual; observe the magnitude of the noise on the measurement, 2261 Call to vendors, licensee, or suppliers, Suppliers of chlorine, 2457 Suppliers of caustic, 2449 Samples and measurements, Caustic and check against the specs from the supplier, 2198 Chlorine and check against the specs from the supplier, 2656 Water and check for high levels of silt or humic acid from spring runoff, 1774 Open and inspect, Check that orifice plates are not installed backwards in the two locations, 1909 Open reactor pipe and check that it is clear (free of obstructions) and that the chlorine injection line is centered, 1948 .

Take “corrective” action, Shutdown the process. Insert “static mixer” just downstream of chlorineinjection point, 1598 Replace the water controller with FFC (flow fraction control, a ratio) instead of PFC, 1209 Shutdown the process. Change control loop on water line to be feed-forward control, 663 Shut down the process. Add a support structure to the chlorine feed line to remove the vibration of the feed line, 899 Redesign the caustic-water mixing area to provide better mixing. Shut down the process and install the improved system, 749 Insert “static mixer” just downstream of caustic injection point, 295

Case ’34: The cement plant conveyor [6, solids conveyor, bagging, dust filters, blowers, fans, cyclone: ceramic] This plant produces dry mixes of mainly cements and coarser aggregates. The raw materials are added – one batch at a time – to a mixing hopper, transported by a high-pressure, batchwise, conveying system to a solids blender or mixing vessel. The mixed material is conveyed from the mixer to a packaging hopper feeding a bagging machine. The finished, mixed product is packaged in 25-kg bags. The Quality Control Department checks regularly on the finished product to ensure that the material has adequate strength by forming casts and determining the break strength of the casts. Two or three days ago, the QC Department found that the strengths of the cast samples had decreased substantially. What’s going on?

333

334

8 Prescription for Improvement: Put it all Together

Case ’34: The cement plant conveyor .

. .

Immediate action for safety and hazard elimination, Put on safe-park, 262 Safety interlock shut down, 418 SIS plus evacuation, 919 Weather, Today and past, 5 Maintenance: turnaround, When and what done?, 516

Maintenance: routine, When and what done?, 41 .

What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Specs for blend, 771 Specs on mixing time, 1143 Handbook, Pertinent properties of cement blends, 1917 Trouble-shooting files, 1626 Calculations and estimations (that can be done in the office before special tests are done) Equipment performance, Dust collection, 212 Fans, 441 Conveying from mixer to packaging hopper, 25 Mixer/ blender, 569 Packaging, 1372 Materials hoppers and star valve feeder, 1083 Blower and batch-conveying system from the bins to the mixers, 1001

.

.

What is the current operation

Visit control room: control-room data: values now and from past records. No sensors or data in control room. Process operators, Mixing: What did you do differently three days ago?, 137 Packaging: What did you do differently three days ago?, 728 This shift on mixers, 538 This shift on packaging, 504 Previous shift on mixers, 941 Previous shift on packaging, 1417 Check diagram and P&ID versus what’s out on the plant, 1260 Control system, For dust collector, 865 For packaging, 585 Call to vendors, licensee, Packaging unit, 336 Solids mixing unit, 432

8.2 Cases to Help you Polish Your Skill

Baghouse, 391 Pressurized conveying system from the feed hopper to the solids mixer, 812 Samples and measurements, Monitor sampling of raw material, QC. Is the particle-size distribution within specs? QC perform tests, 1312 Monitor sampling. after the mixer. Sample to QC: Measure strength of casts ; particle-size distribution; composition of components, 1078 Monitor sampling. From the package Sample to QC: Measure strength of casts ; particle-size distribution; composition of components, 60 Sample air leaving bag filter and measure concentration and size of particulates. Compare with previous, 2886 Open and inspect, Feed hopper, 508 Solids mixer, 790 Conveyor hopper, blower and conveying line, 115 Dust collector on top of packaging hopper, 31 Packaging unit, 382 .

Take corrective action, Bang side of the hopper to the bagging machine to break any bridging, 2075 Repipe the bagging machine so that the dusty air flows down through the bags instead of up through the bags. The goal is to prevent fluidization of the fines in some of the bags, 2496

Case ’35: The cycling triple-effect evaporator [6, long tube evaporators, vacuum system, steam; glycerine] We are starting up a new plant, and it is like a zoo out there. We have three, multiple-effect evaporators to concentrate glycerine. However, we can’t seem to get the system to behave. It will not steady out. The flowrates and pressures in the three evaporators all seem to cycle. Clear up the problem. All pressures are absolute pressures. Figure 8-16 shows the process. Figure 8-17 illustrates the ejector system used to pull the vacuum.

335

8 Prescription for Improvement: Put it all Together

336

0.61 kg/s

Steam

0.54 kg/s

0.43 kg/s

7.9 kPa 0.64 kg/s

PI

to high vacuum

FR

1.19 kg/s

0.65 kg/s

cooling water to/from the tower

PI

228 kPa

drain

steam trap

TI

109ºC

Figure 8-16

PI

PI

164 kPa

115 kPa

98ºC

80ºC

TI

TI

TI

93.5ºC

barometric leg 0.43 kg/s to product 0.24 kg/s

TI

FR

1.80 kg/s

70ºC

Three multiple effect evaporators for Case ’35.

water

steam

water

vacuum system

steam

from the system

barometric legs Figure 8-17

The high vacuum system for the multiple effect evaporators for Case ’35.

steam

8.2 Cases to Help you Polish Your Skill

Case ’35: Cycling triple effect . .

.

. .

MSDS, 1871 Immediate action for safety and hazard elimination, Put on safe-park, 1611 Safety interlock shut down, 1744 SIS plus evacuation, 2331 IS and IS NOT: (based on given problem statement), What, 2128 When, 2095 Where, 2964 Weather, Today and past, 2668 Maintenance: turnaround, When and what done?, 2571

Maintenance: routine, When and what done?, 2508 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Vacuum system, 2746 Three evaporators, 2489 Two preheaters, 2344 Steam trap, 2147 Feed pump, 2097 Vendor files: Triple effect vacuum unit, 2041 Commissioning data, P&ID, internal reports, 2255 Handbook, Steam tables, 2450 Trouble-shooting files, 1818 Calculations and estimations (that can be done in the office before special tests are done) Pressure profile, If there is a leak: stages 1, 2 and 3; which direction would it leak?, 1759 If there is a leak in the preheaters, which direction would it leak?, 1853 Mass balance, Steam: condensate per stage, 1028 Process liquid out from the third stage, 1635 Energy balance: sink= source, estimate heat loads based on steady state information given, 1079 Rate, boiling range: nucleate or film, 1128 Equipment performance, three stages, 1178 .

.

What is the current operation

Visit control room: control-room data: values now and from past records. Sensors out on the plant. Process operators, Please tell me about the startup, 1228

337

338

8 Prescription for Improvement: Put it all Together

Contact on-site personnel, Steam utilities: steam conditions to ejector system, 1278 Cooling water to the barometric condensers; is the cooling water hotter than usual, 1328 Visit site, read present values, observe and sense. Glycerine feedrate, FR, 1378 Pressure PI stage 1; 228 kPa, abs, 1427 Pressure PI stage 2; 164 kPa, abs, 1476 Pressure PI stage 3; 115 kPa abs, 814 Pressure PI inlet to vacuum system. 7.9 kPa abs, 611 Temperature TI inlet to preheater, 962 Temperature TI exit stage 1.98 C, 915 Temperature TI exit stage 2.80 C, 864 Temperature TI into stage 1; 109 C, 866 Temperature TI exit preheater 1: 93.5 C, 571 What is the cycle time, 509 Steam flowrate to stage 1, 44 Steam pressure to the ejectors, 139 Check diagram and P&ID versus what’s out on the plant, 1304 On-site simple tests: Knock on the side of the vertical third effect and listen for a change in sound that corresponds to the cycle, 281 Gloved hand to test the temperature of trap and upstream and downstream of the steam trap. Do this over the full cycle, 431 Sensors: use of temporary instruments, Use a stethoscope to listen to the steam going to each of the three ejectors in the vacuum system. Listen for “vibration”, 472 Call to vendors, licensee, Triple-effect evaporator system, 983 Steam-trap vendor, 770 Control system, Put control system on manual, 860 Samples and measurements, Sample glycerine feed before pump and just as it enters the first stage; analyze for glycerine content and compare with specifications, 2265 Immerse steam-trap discharge line in weighed amount of cold water and measure condensate flow over three successive one-minute periods, 2108 Open and inspect, Triple effects and pressure test for leaks, check for blockages, evidence of excessive foaming, and fouling of inside or outside of tubes, 2280 Preheaters: and pressure test for leaks, check for blockages, and fouling of inside or outside of tubes, 2207 Centrifugal feed pump, 2931 Barometric condensers; look for blockage that could have flooded the condenser, 2678

8.2 Cases to Help you Polish Your Skill

Take “corrective” action, Raise the level of the overflow baffle on the hot well to give better seal of the downcomer, 2904 Retune the control system, 2136 Replace the valve on the steam line to the first ejector after the booster ejector, 1703

.

(courtesy W. K. Taylor, B. Eng. 1966, McMaster University) [7, reactor, heat exchanger, steam drum; reformer, ammonia] The heat from exit gas from the secondary reformer is used to generate steam according to the scheme shown in Figure 8-18. The temperatures and pressures for making 100 kg/s (as measured by F4) are as follows: the location, the design value and the current value for temperature and pressure, respectively. Case ’36: The really hot case

TI 4

PI 4

PC 6 FI 4 AAA

STEAM DRUM

BOILER FEED WATER NATURAL CIRCULATION

A TI LC 1 TI

TI

TI

WASTE HEAT BOILER

STEAM SUPERHEATER

TI TI 2 PI 1

TI 1

TIC 3

PI 3

BYPASS VALVE CONTROLLED BY TIC-3

BYPASS AND ALTERNATIVE COOLING

TI 5

METHANE STEAM AIR HP STEAM HP AIR COOLING WATER

Figure 8-18

Steam generation by the secondary reformer exit gas for Case ’35.

339

340

8 Prescription for Improvement: Put it all Together

Location

Design T, C;

Reads, C

Design P, MPa

Reads, MPa

Reactor exit Inside boiler Boiler exit Superheat feed Superheat exit

T1 = 1000 T2 = 750 T3 = 600 T4 = 325 (sat) T5 = 353

1000 820 730 325 443

P1 = 4.5 ? P3 = ? P4 = 12.0 ?

4.55 ? 4.5 12.0 ?

We cannot seem to control the exit gas temperature TI-3; the TI-3 controller output to bypass is 100% closed. The steam is much too superheated and could damage the turbine where it is used. Correct the fault. Case ’36: The really hot case . .

. .

. .

MSDS, 113 Immediate action for safety and hazard elimination, Put on safe-park, 116 Safety interlock shut down, 43 SIS plus evacuation, 1327 More about the process, 2981 IS and IS NOT: (based on given problem statement), What, 1018 When, 2408 Where, 2170 Weather, Today and past, 2015 Maintenance: turnaround, When and what done?, 2971

Maintenance: routine, When and what done?, 2564 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Reformer, 2712 Steam drum, 2853 Waste-heat boiler, 2953 Steam superheater, 2651 Vendor files: Waste-heat boiler, 2253 Steam superheater, 2453 Reformer catalyst, 2056 Commissioning data, P&ID, internal reports, 2017 Handbook, Steam tables, 2435 Trouble-shooting files, 555 .

Calculations and estimations (that can be done in the office before special tests are done)

8.2 Cases to Help you Polish Your Skill

Pressure profile, Direction of leak: waste-heat boiler, 902 Direction of leak: steam superheater, 520 Dp for process gas through the boiler and compare with rule-of-thumb, 37 Energy balance: sink= source, For waste-heat boiler: Heat loss in gas = steam generated, 428 For waste-heat boiler: heat load in first section vs second, 1742 For superheater: heat loss in gas = superheat, 1958 Rate, Boiling regime (film or nucleate), 1778 Equipment performance, Estimate performance of boiler based on given data, 1538 Estimate performance of superheater based on given data, 2967 .

What is the current operation

Visit control room: control-room data: values now and from past records, Temperature TI-1, 2526 Process operators, What has been done? 2688 Call to others on-site, Downstream users of superheated steam, 341 Upstream feed to the reformer: any change, 1449 Visit site, read present values, observe and sense. FI-4: flow of steam, 2946 Output signal from TIC-3 to baffle, 2599 Check diagram and P&ID versus what’s out on the plant, 1134 On-site simple tests: Are the thermocouple temperatures in the reformer bed consistent with the exit temperature TI-1?, 2120 Sensors: check response to change Temperature TI-5; superheated steam exit superheater, 2352 Temperature TI-2; process gas inside boiler, 2047 Temperature TI-3; process gas boiler exit, 1677 Pressure PI-1, 1984 Pressure PI-3, 53 Pressure PI-4, 354 Flowmeter FI-4, 804 Sensors: use of temporary instruments, Use contact or laser pyrometer to measure TI-5, 1413 Use contact or laser pyrometer to measure TI-1, 1861 Sensors: calibrate, Temperature TI-5; superheated steam exit superheater, 854 Temperature TI-2; process gas inside boiler, 40 Temperature TI-3; process gas boiler exit, 459 Pressure PI-1, 2470 Pressure PI-3, 2350 Pressure PI-4, 2950 Flowmeter FI-4, 2634

341

342

8 Prescription for Improvement: Put it all Together

Control system, Put TIC-3 on manual and open the bypass valve. Note changes is TIC-3 and TI-5, 2495 Retune the TIC-3 control-baffle system, 2976 Put PC-6 on manual. Open the valve and watch steam drum pressure and PI4, 584 Call to vendors, licensee, Catalyst in the secondary reformer, 2360 Samples and measurements, Analyze the material found, 2083 Open and inspect, Orifice plate for FI-4 to check if the plate was in backwards. Tab is not clearly marked, 1556 Isolate, drain the superheater. Pull the bundle and inspect for fouling; inspect the inside of the tubes for fouling, 1904 Isolate, drain the superheater. Pressure test the tubes with water and look for leaks. Plug any leaking tubes, 627 Isolate, drain the boiler. Pressure test the tubes with water and look for leaks. Plug any leaking tubes, 128 Isolate and drain the boiler. Pull the bundle and check for fouling on the outside of the tubes, 2810 Isolate and drain the boiler. Check the location of the bypass valve. Inspect the inside of the tubes for fouling, 2647 .

Take “corrective” action, Correct the linkage between the TIC-3 and baffle, 2149 Replace the sensor TIC-3, 1105

Case ’37: The mill clarifier (courtesy D. F. Fox, B. Eng. 1973, McMaster University) [7, thickener, sludge pumps, control; pulp and paper] The mill clarifier is 45 m in diameter and designed for 90% reduction in suspended solids from an inflow of 0.53 m3/s. The retention time is 2 hour. Parshall flumes measure the total inlet and outlet flows. There are 24-hour composite samplers on the influent and overflow effluent lines. The system is shown in Figure 8-19. Our target is to have the clarifier effluent < 50 ppm. The operator cannot understand why the clarifier effluent has been around 200 ppm for the last 20 days. The sludge pump has been “wide open”, yet all that occurs is flooding at the belt filter. The feed concentration is about 400 ppm. That’s the problem. Get it fixed pronto.

8.2 Cases to Help you Polish Your Skill

Figure 8-19

The mill clarifier for Case ’37.

Case ’37: The mill clarifier .

.

. .

Immediate action for safety and hazard elimination, Put on safe-park, 438 Safety interlock shut down, 121 SIS plus evacuation, 2975 IS and IS NOT: (based on given problem statement), What, 2676 When, 2801 Where, 2567 Weather, Today and past, 1656 Maintenance: turnaround, When and what done?, 1959

Maintenance: routine, When and what done?, 1839 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Clarifier, 1203

343

344

8 Prescription for Improvement: Put it all Together

Sludge pump, 1456 Belt filter, 1045 24-h composite sampler, 636 Parshall flumes, 954 Vendor files: Parshall flumes, 507 Commissioning data, P&ID, internal reports, 61 Trouble-shooting files, 317 .

What is the current operation

Visit control room: control-room data: values now and from past records, Torque, 493 Records of 24 h sampler on feed over the past week, 47 Records of 24 h sampler on effluent over the past week, 204 Influent flowrate based on Parshall flume, 403 Bottoms flowrate from sludge pump to belt filter, 703 Overflow flowrate based on Parshall flume, 579 Process operators, What was done this shift? 861 Previous shift, 1310 Call to others on-site, Operators of upstream units about possible upsets, 1154 Visit site, read present values, observe and sense. Sludge pump: cavitation? 1402 Sludge pump exit valve wide open or mid-range? 1133 Rake and skimmer turning at expected rpm, 1056 Belt filter: flooding? 1752 Check diagram and P&ID versus what’s out on the plant, 2373 On-site simple tests: Clear any stuff out of the sampler lines on 24 h composite sampler of feed and resample via 24-h composite, 1580 Sensors: use of temporary instruments, Use point gauge on both Parshall flumes and use this information to determine total flowrate, 1516 Sensors: calibrate, Recheck zero setting for the Parshall flumes, 2204 Samples and measurements, Grab samples of feed whenever torque starts to increase. Five successively at 20-min intervals. Analyze for suspended solids and compare with 400 ppm reading, 2107 Feed to the belt filter. Grab sample when torque increases and when increase pump output. Analyze for suspended solids and compare with 3%, 2353 Open and inspect, Sampler lines; look for plugs in sampler line, 2463 Stop process; drain clarifier and inspect rake, 2961

8.2 Cases to Help you Polish Your Skill .

Take “corrective” action, Turn off sludge pump and increase rake speed until torque just starts to increase; then start sludge pump, 2622 Decrease rake speed; keep sludge pump at usual design flowrate, 2534 Reduce the sludge-pump flowrate to 5 L/s, 1819 Increase vacuum on the belt filter, 1615

(courtesy of T. E. Marlin, Chemical Engineering, McMaster University) [7, sequence of distillation columns with auxiliaries; depropanizer-debutanizer] The process is the depropanizer-debutanizer described in Case ’8, Chapter 2. A P&ID is given in Figure 2-4 accompanying Case ’8. There has been a lot of trouble on this unit. So last January, we did a detailed study, and simulation, of the deprop and debut units. Our conclusion was that this unit was operating on spec. for the usual range of feedstocks with all equipment components operating very close to the design values. It’s a hot steamy day in August; you are glad that you are in your air-conditioned office. The phone rings! The laboratory analysis of the vapor product on the deprop indicates a loss of propane to the fuel system that is 212 times the design value. This loss of valuable product to fuel gas is costing lots of money. Fix it fast. Case ’38: More trouble on the deprop!

Case ’38: More trouble on the depropanizer . .

. .

. . .

MSDS, 1495 Immediate action for safety and hazard elimination, Put on safe-park, 315 Safety interlock shut down, 1315 SIS plus evacuation, 2315 More about the process, 1010 IS and IS NOT, What, 2978 When, 2627 Where, 2336 Why? Why? Why?, Best goal? To reduce the propane loss to fuel gas, 1360 Weather, Today and past, 1761 Maintenance: turnaround, When and what done?, 19

Maintenance: routine, When and what done?, 269 .

What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Condenser, E-25, 87 Distillation column, C8: 478 Thermosyphon reboiler, E-27, 969 Reflux pump, F-27, 2486 Overhead drum, V-30, 2970

345

346

8 Prescription for Improvement: Put it all Together

Vendor files: Condenser, reboiler and preheater, 1475 Steam traps, 1110 Commissioning data, P&ID, internal reports, 1043 Handbook and Google, Cox charts, 2024 Trouble-shooting files, 85 .

What is current operation

Visit control room: control-room data, Feed to the C8, depropanizer. FC/1, 2831 Reflux flowrate, FIC/4, 2036 Overhead product flowrate of propane, FIC/5, 1633 Pressure drop Dp I/1, 1926 Level bottoms LIC/2, 387 Temperature bottoms TI/4, 173 Temperature mid-column TIC/5, 980 Temperature top, TI/3, 789 Pressure on overhead drum, PIC/10, 513 Process operators, This shift, 552 Previous midnight shift, 1352 Operating procedures, Column pressure, 1185 Call to others on-site, Utilities: what is the temperature of the water leaving the cooling tower; what is the flowrate to our unit, 1908 Operators on downstream propane unit: amount and quality of propane received, 872 Visit site, read present values, observe and sense. Column pressure, PI- 4, 2405 Pressure relief to flare PSV-1, 2384 Look at the flare, 2994 Temperature mid-column TI- 8, 2123 Valve position for column feed, FV- 1, 2701 Valve position for steam to preheater E-24, 2606 Valve-stem position on PV-10, 2504 Level in feed drum, V-29; LI- 1, 2942 Isolation valves around the condenser, 2516 Feel temperature of water line going into condensers, 2008 Check diagram and P&ID versus what’s out on the plant, 2769 On-site simple tests: Open vents to fuel gas on each condenser for 10 min; then close, 1512 Gather data for key calculations, Pressure profile, Drum V-29 to pumps, F-25, 26, 959 Dp across pump, converted to head, 605 Pump F-25–26 exit to feed location, 857 Drum V-30, pump F-27 and reflux into column, 549

8.2 Cases to Help you Polish Your Skill

Pump F-27, 1464 Vapor from top of column to vapor space in V-30, 1966 Thermosyphon reboiler process fluid side, 1676 Mass balance, Over column, 1558 Energy balance: Heat load for condenser. Heat load condensed = heat load picked up in cooling water, 2911 Sensors: check response to change, Temperature at the top of column C8: TI 3, 1955 Pressure at the top of column C8, PI 4, 1175 Sensors: use of temporary instruments, Use a surface temperature sensor to measure the temperature of cooling water going into the condensers, 1012 Sensors: calibrate, Temperature at the top of column C8: TI 3, 1263 Pressure at the top of column C8, PI 4, 327 Samples and measurements, Sample feed and analyze for the amount of propane, 93 Sample overhead concentration in gas to flare from top of drum every 10 minutes for 1 hour. Analyze for propane, 2442 Sample overhead concentration from the tower every 10 minutes for 1 hour. Analyze for propane, 2871 Open and inspect, Condensers, 2554 .

Take “corrective” action, Direct water from the fire hose onto the outside shell of the overhead condensers, 2174 Increase column pressure to 1.78 MPa, 424 Decrease feed to the column to 80% usual flowrate, 774

(from D. R Winter, Universal Gravo-plast, Toronto, 2004) [Difficulty 7; involves feed bin, molding machine, mold and mold design. Context: injection molding of thermoplastics] The customer wants a display stand for sunglasses. The display required several contoured plates (35 cm by 18 cm) molded of polypropylene. These are mounted vertically and held in place at the edges. The display is to be installed in windows of drugstores so that the sunglasses on display are illuminated by sunshine during the day and by fluorescent tubes behind the molded plate during darker hours. The customer requested a special pearlized pigment be added so that the display had pizzaz. The feed was a mixture of polypropylene resin, a UV stablizer and the pearlescent pigment. The feed was dried by contacting it with hot air at 93 C for three hours and was used within 20 minutes of drying. The injection-molding machine was a reciprocating screw; L:D of 20:1; compression ratio 2.5:1 and the shot size for this job was 70% of the machine capacity. The processing temperature was 230 C. The Case ’39: The case of the lumpy sunglass display

347

348

8 Prescription for Improvement: Put it all Together

mold pressure was 4.2 MPa, consistent with the part width and the flow length. A mold was created of aluminum with appropriate gate and vent locations and cooling arrangements for the mold. The nozzle, sprue and cold runner system were carefully designed. The molding cycle was established, a molded sample approved by the customer and production started. The first 2000 parts had been shipped when the customer called and complained that all the parts looked fine in reflected light but in transmitted light, lumps were evident in the panels. Production was stopped. Your job is to eliminate the lumps that appear when the panels are viewed with transmitted light. . .

. . .

.

MSDS, 1572 Immediate action for safety and hazard elimination, Put on safe-park, 1799 Safety interlock shut down, 1298 SIS plus evacuation, 298 More about the process, 12 More about the product and the mold, 2238 IS and IS NOT: (based on given problem statement), What, 2481 When, 2820 Where, 2649 Weather, Today and past, 2272

Maintenance: routine, When and what done?, 1831 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Injection-molding machine, 1983 Mold, 1445 Feed for this product, 1004 Commissioning data, P&ID, internal reports, Prototype activity. Check if the prototype had lumps when viewed by transmitted light, 911 Handbook, Data sheets for resin, 667 Trouble-shooting files, 511 .

What is the current operation

Visit control room: control-room data: values now and from past records, Fill cycle, 136 Cool cycle, 2738 Open cycle, 2521 Melt feed temperature into mold, 2310 Injection pressure, 1934 Injection rate, 1545 Extruder: rear-barrel temperature, 1857

8.2 Cases to Help you Polish Your Skill

Extruder: head temperature, 1183 Backpressure, 1084 Mold pressure, 730 Hold pressure, 594 Mold temperature, 92 Process operators, 425 Operating procedures, Shutdown procedures used, 279 Were the feed materials from the same batches of resin, UV stabilizer and pearlized pigment as was used for the prototype?, 1945 .

Check with colleagues about hypotheses, 2416

On-site simple tests: Clean possible dirty machine by running a charge of acrylic through the extruder; then try again, 2087 Use another molding press; run the process under the prototype conditions, 2724 Redry the resin for three hours at 93 C and dry additives and use immediately, 2575 Reduce the screw rpm by 5% and use a pyrometer to measure melt temperature, 2771 Change melt temperature, increase by 20 C to 251 C and keep the cooling time the same, 2187 Change melt temperature, increase by 20 C to 251 C and increase the cooling time, 2009 Use the standard conditions but try an old batch of dried polypropylene from a different supplier together with the UV stabilizer and pearlized pigment, 1662 Use the standard conditions, use the polypropylene we have been using and the pearlized pigment but delete the UV stabilizer, 1718 Use the standard conditions, use the polypropylene we have been using and the UV stabilizer but delete the pearlized pigment, 1624 Clean the hopper; add a cover over the hopper to prevent atmospheric dirt from falling into the feed, 1285 Reduce the injection pressure to 7.5 MPa, 1367 Redesign of product and mold, Increase the diameter of the gates by 10%, 631 Sensors: check response to change Hot-melt temperature sensor at nozzle, 1337 Exit cooling-water temperature on top mold, 304 Exit cooling-water temperature on the bottom mold, 241 Pressure sensor at nozzle, 1841 Sensors: use of temporary instruments, Use a pyrometer or laser sensor to measure melt temperature and compare with temperature sensor, 2663 Sensors: calibrate, Hot-melt temperature sensor, 2635

349

350

8 Prescription for Improvement: Put it all Together

Call to vendors, licensee, or suppliers, Properties of polypropylene, 77 Properties of UV stabilizer, 183 Properties of pearlized pigment, 2359 Samples and measurements, Sample resin and analyze for moisture, 2206 More complicated tests, Change mold from a QC-7 aluminum mold to a mold P-20, 2873 Open and inspect, Check the shutoff valve for dirt or contamination, 2592 .

Take “corrective” action, Replace the shutoff valve, 2849

Case ’40: The cool refrigerant (courtesy T. E. Marlin, Chemical Engineering Department, McMaster University) [7, turbine, compressor, KO pot, refrigeration system; general] The propylene refrigeration system, given in Figure 8-20, was operated successfully for several years. Since the steam turbine and refrigerant compressor had spare capacity, they modified the process by adding an additional heat exchanger, E101, to cool the process stream in E101, as shown in Figure 8-21. When the process was started up, the design values could not be obtained for either of the streams to be cooled. The temperatures, given in C, are:

For E100: T5: actual 100, design 101; T6, actual 10; design 5. For E101: T7: actual 70; design 68; T8, actual 4; design 10. Into the compressor: T2: actual –5; design –4. F 1

H.P. Steam

PC 1

K.O. Drum L 1

periodic flow

T 2

C.W. steam turbine

L 2

compressor

T 5

LC 3

E 100

Figure 8-20

The original propene refrigeration system for Case ’40.

T 6

chilled process

8.2 Cases to Help you Polish Your Skill

Because T6 is too high and T8 is too low, the plant cannot produce saleable material. In checking over the design calculations, you realize that the system should work. What’s going on here? F 1

H.P. Steam

PC 1

K.O. Drum L 1

periodic flow

T 2

C.W. steam turbine

L 2

compressor

T 5

LC 3 T 6

E 100

chilled process warm T 7

LC 4 T 8

E 101

Figure 8-21

The modified refrigeration system for Case ’40.

Case ’40: The cool refrigerant . .

.

. .

MSDS, 914 Immediate action for safety and hazard elimination, Put on safe-park, 677 Safety interlock shut down, 744 SIS plus evacuation, 949 IS and IS NOT: (based on given problem statement), What, 1993 When, 1683 Where, 1827 Weather, Today and past, 1918 Maintenance: turnaround, When and what done?, 1212

chilled

351

352

8 Prescription for Improvement: Put it all Together

Maintenance: routine, When and what done?, 1466 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) E100 chiller, 1031 E101 chiller, 567 Vendor files: E100 chiller, 883 E101 chiller, 2485 Commissioning data, P&ID, internal reports, 2215 Handbook, Propylene temperature–pressure data and latent heat, 2913 Trouble-shooting files, 2544 Calculations and estimations (that can be done in the office before special tests are done) Pressure profile, Direction of leak: unable to tell because the pressures of the process streams are not known. Direction of leak from propylene–cooling water in condensers, 2658 Energy balance, Heat load for E100; actual versus design, 2708 Heat load for E101; actual versus design, 2157 Propylene flow: design versus actual calculated from heat loads on E100 and E101, 1734 Rate, Does propylene boil as nucleate or film form, 2208 Equipment performance, E100: does heat flow in the correct direction?, 1707 E101: does heat flow in correct direction?, 1582 UA calculations for E100 actual versus design, 190 UA calculations for E101 actual versus design, 740 .

.

What is the current operation

Visit control room: control-room data: values now and from past records, Temp and pressure of propylene: T2 and P1; are these on pH diagram for saturated propylene gas; if not might it be sensors? or contamination? 1140 Level: E100, LC/ 3, 69 Level: E101, LC/4, 362 Level: KO pot, L/1, 449 Level: liquid propylene drum, L/2, 944 Flow of propylene, F/1, 643 Propylene pressure at compressor suction, PC/1, 713 Process operators, Please tell me what has happened to far, 1369 Operating procedures, 1125

8.2 Cases to Help you Polish Your Skill

Contact with on-site specialists, Operators of unit that provides and receives the process stream from E100: flows and temperatures constant and consistent with what’s on this unit?, 1046 Operators of unit that provides and receives the process stream from E100: pressure in the lines and have you detected any contamination in your process streams, 1809 Operators of unit that provides and receives the process stream from E101: pressure in the lines and have you detected any contamination in your process streams, 1932 Operators of unit that provides and receives the process stream from E101: flows and temperatures constant and consistent with what’s on this unit?, 2338 Utilities: steam pressure, flow and degree of superheat, 2144 Utilities: cooling-water temperature, flow to propylene condensers, 2414 Visit site, read present values, observe and sense. Steam leave off the top of the steam header?, 2125 Valve-stem position on valve LC/3, 2783 Valve-stem position on valve LC/4, 2843 Check diagram and P&ID versus what’s out on the plant, 2583 On-site simple tests: Observe movement of the valve stem on LC/3 when the set point is changed, 1180 Observe movement of the valve stem on LC/4 when the set point is changed, 615 Open vent bleed on the top of E100 for 5 minutes, close and observe change, 842 Open vent bleed on the top of E101 for 5 minutes, close and observe change, 688 Open vent bleed on the top of propylene drum for 5 minutes, close and observe change, 124 Consistency check, Temp and pressure of propylene: T2 and P1; are these on pH diagram for saturated propylene gas; if not might it be sensors? or contamination?, 1140 Sensors: check response to change LC/3, 394 LC/4, 247 Control system, Put LC/3 on manual and adjust level to control temperature T/6, 94 Put LC/4 on manual and adjust level to control temperature T/8, 2389 Retune controller LC/3, 2890 Retune controller LC/4, 2683

353

354

8 Prescription for Improvement: Put it all Together

Sensors: calibrate, LC/3, 2597 LC/4, 2072 Samples and measurements, Sample propylene and analyze for contaminants, 657 More complicated tests, Reduce the propylene pressure such that the target range of cooling occurs when neither bundle is completely covered with liquid; adjust the levels in E100 and E101 separately until target temperatures are achieved, 130 Open and inspect, Propylene drum to check the condition of the vortex breaker at the bottom exit nozzle, 265 E100 and look for fouling inside or outside tubes, 1649 E100; isolate and pressure test for leaks, 2263 E101 and look for fouling inside or outside tubes, 2789 E101; isolate and pressure test for leaks, 2019 Check for plugs, obstructions or junk in the line from the liquid propylene reservoir and E100, 1638 .

Take “corrective” action, Replace LC/3 on E100, 1967 Replace LC/4 on E101, 1094 Replace control valve on propylene to E100, 775

(courtesy T. E. Marlin, Chemical Engineering Department, McMaster University) [7, distillation plus auxiliaries; general] The column for the new process was started up for the first time one month ago. From the beginning, the pressure control was not very good; the pressure seemed to deviate from its set point more than in other columns in the plant. However, the product purities seemed to be within specifications, so you didn’t worry about it. Today, the plant operator calls, “We’ve got a problem on that new column! The pressure controller is not working! The pressure in the column is higher than the set point yet the controller output is 0%! Yes, the production rate and the specifications on all lines are OK. But this pressure is really worrying me.” The column is shown in Figure 8-22. Case ’41: The ever-increasing column pressure

8.2 Cases to Help you Polish Your Skill

Figure 8-22

The column discussed in Case ’41.

Case ’41: Ever-increasing pressure . .

.

. .

MSDS, 379 Immediate action for safety and hazard elimination, Put on safe-park, 180 Safety interlock shut down, 479 SIS plus evacuation, 58 IS and IS NOT: (based on given problem statement), What, 758 When, 964 Where, 679 Weather, Today and past, 546 Maintenance: turnaround, When and what done?, 1131

Maintenance: routine, When and what done?, 1483 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties)

355

356

8 Prescription for Improvement: Put it all Together

Column, 2706 Reboiler, 2559 Control system, 2956 Condenser, 2867 Commissioning data, P&ID, internal reports, 2695 Handbook, Options for controlling overhead pressure based on Chin’s classic article from 1979 Hydrocarbon Process, 2184 Type of hydrocarbon based on overhead pressure and temperature, 2410 Trouble-shooting files, 2423 Calculations and estimations (that can be done in the office before special tests are done) Pressure profile, Direction of flow if there is a leak: condenser, 2069 Direction of flow if there is a leak: reboiler, 129 Flow from top of column to the reflux drum, 328 Equipment performance, Control system, 445 Condenser, 205

.

.

What is the current operation

Visit control room: control-room data: values now and from past records, Current overhead temperature, 886 History of overhead temperature, 736 History of overhead pressure, 1434 Current overhead pressure; PC/1, 630 Process operators, When you first started up two months ago, was the control valve on the CW almost closed?, 534 Contact with on-site specialists, Utilities: cooling water: temperature and flowrate, 1935 Operators of upstream plant supplying the feed to the column: any change in light ends in composition or change in flowrate, 2385 Visit site, read present values, observe and sense. Check for consistency between PC/1 and the pressure on the top of the column, 2205 Valve-stem position on the cooling water, 1109 Look at the flare, 1247 Amount of steam flowing to the reboiler, 1313 Bottom pressure, 1299 Read the controller output signal to the cooling-water valve, 1619 Is the control valve on the cooling water fail open or fail closed? does the direction arrow on the valve agree with the direction of flow?, 2938 Bottom temperature, 1454 Check diagram and P&ID versus what’s out on the plant, 1119

8.2 Cases to Help you Polish Your Skill

On-site simple tests: Open vent on the condenser for 2 min to bleed off any accumulated inerts, close and check performance, 1753 Does the valve stem on the cooling water respond to increase in set point, 1543 Increase the resistance in the vent break line to prevent uncondensed gas from bypassing the condenser, 1888 Tap the side of the condenser in a vertical line and listen for the change in sound suggesting the location of the vapor-condensate interface, 2631 Sensors: check response to change Does PC/1 respond to change in set point, 2160 Sensors: use of temporary instruments, Use a contact surface temperature probe to measure inlet and outlet temperatures of the cooling water to the condenser, 2063 Sensors: calibrate, Pressure PC/1, 2720 Control system, Retune the PC/1 control system, 2921 Samples and measurements, Sample the feed; analyze for light ends and non-condensibles and compare with previous records, 2539 More complicated tests, Stop the process; crack open the flange on the vent break line and insert a resistance disk with a smaller diameter hole to increase the resistance in the “bypass” line, 2652 Open and inspect, Control valve on the cooling water; inspect for quality of trim, plugged, 1559 Condenser and inspect for fouling inside the tubes, 1123 .

Take “corrective” action, Reduce the feedrate to the column, 1371 Direct fire hose water over the condenser, 231

(courtesy W. K. Taylor, B. Eng. 1966, McMaster University) [7, storage tank, reactor, pump, heat exchanger; ammonia] Ammonium nitrate is formed by the exothermic neutralization of nitric acid with ammonia. The nitric acid is produced as a 56% solution and is pumped from storage tanks to the reaction leg of the primary neutralizer. In the reaction leg, the acid is mixed with superheated ammonia vapor and also with an ammonia-bearing offgas stream from the urea plant. The heat of reaction produces a large amount of steam, and the liquid, which is removed to a secondary neutralizer is an 85% solution of ammonium nitrate. Usually about 50% of the ammonia needed to neutralize the acid is provided by the off-gas stream from the urea plant. The usual temperatures in the system are: nitric acid in storage: 40–50 C; superheated ammonia vapor entering neutralizer: 90–100 C; the urea plant off-gas: 100–110 C; the AN product

Case ’42: The weak AN

357

358

8 Prescription for Improvement: Put it all Together

solution: 130–140 C; the temperature inside the neutralizer: 135 C. The process is shown in Figure 8-23. steam

Primary Ammonium Nitrate Neutralizer

A.N. product to secondary neutralizer

TI

Nitric Acid Storage Tank

TI

Flow meter Acid Transfer Pump

Figure 8-23

urea plant off gas

FCV

PI

cooling water

C.W. return

NH3 vapor

The process for AN for Case ’42.

The urgent phone call poses today’s problem. The temperature inside the neutralizer-reactor has been falling. The most recent analysis of the AN shows a concentration of about 79%. “This weak AN will cause problems in the downstream processing units where the solution is concentrated in a falling film evaporator to about 99% and then prilled. Let’s get this fixed quickly. The evaporator can’t concentrate the solution enough for prilling and the whole process will shut down. “ Case ’42: Weak AN . .

. .

. .

MSDS, 2716 Immediate action for safety and hazard elimination, Put on safe-park, 2692 Safety interlock shut down, 2214 SIS plus evacuation, 2406 More about the process, 2870 IS and IS NOT: (based on given problem statement), What, 237 When, 332 Where, 1332 Weather, Today and past, 1082 Maintenance: turnaround, When and what done?, 1440

Maintenance: routine, When and what done?, 880 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties)

8.2 Cases to Help you Polish Your Skill

Storage tank, 906 Check the sensors on the off-gas and ammonia line to ensure that there is no mercury used in the sensor, 396 Design calculations for the orifice plate measuring the nitric acid flowrate, 1940 Transfer pump, 1796 Neutralizer, 2826 Cooling coil, 2628 Vendor files: Transfer pump, 2957 Commissioning data, P&ID, internal reports, 1338 Handbook, Heat of formation of ammonium nitrate, 1186 Trouble-shooting files, 1116 Calculations and estimations (that can be done in the office before special tests are done) Pressure profile, Direction of leak: cooling-coil reactant, 743 Direction of leak: steam heating coil to acid in storage tank, 645 Acid flow from the storage tank into the neutralizer, 142 Rate, Of neutralization reaction, 2448 Of cooling, 2492 Equipment performance, Neutralizer and sources of water, 2937 .

.

What is the current operation

Visit control room: control-room data: values now and from past records, Temperature on acid storage tank, 2749 Pressure at exit of pump, 2638 Acid flowrate, 2105 Temperature in neutralizer, 2306 Temperature of AN leaving the neutralizer, 1306 Cooling-water temperature out, 1149 Steam generation from overhead of neutralizer, 1050 Process operators, Please describe what has happened so far, 750 Has this ever happened before and what did you do, 570 Contact with on-site specialists, Utilities: cooling water, 199 Operators of downstream plant receiving our AN: check on temperature and concentration of AN received, 448 Supplier: urea plant off-gas: change in flow, temperature or concentration, 263 Supplier: ammonia plant supplier: change in flow, temperature or concentration, 1829 Purchasing; have we changed suppliers for the nitric acid, 1735 Visit site, read present values, observe and sense.

359

360

8 Prescription for Improvement: Put it all Together

Check that the location of valve stem on acid flow-control valve is mid-range, 1599 Level of liquid in the neutralizer from the level gauge, 1731 Do the tabs on the orifice plate, used to measure the acid flowrate, indicate that the orifice is facing the correct direction, 2290 What sounds do you hear by the neutralizer, 2795 Are there sounds of cavitation from the transfer pump?, 2532 Are there any water/steam lines hooked up to any of the process equipment or lines feeding the neutralizer, 1250 Are there any water/steam lines hooked up to the neutralizer, 1350 Does the steam off the top of the neutralizer go into the top or bottom of the steam header, 1100 Check diagram and P&ID versus what’s out on the plant, 698 On-site simple tests: Check that valve on bypass of acid flow-control valve is shut and block valves fully open, 849 Shut the valve on exit of transfer pump and compare pressure (head) with head on head-capacity curve for this impeller rpm, 948 Consistency checks, Does composition of the off-gas add up to 100%, 250 Does temperature and composition of AN leaving neutralizer agree with conditions downstream, 296 Trend checks, Do both the temperature and the concentration of AN decrease in the neutralizer?, 1838 Gather data for key calculations Mass balance, Based on the measured flows of reactants, products, 82 Based on the measured flows of off-gas and pure ammonia, calculate the fraction of the required amount of the ammonia that comes from the off-gas and compare with design, 422 Energy balance: sink= source, Heat of reaction= heat removed in cooling water, 2347 Sensors: check response to change Temperature of AN in the neutralizer, 1697 Flow-control valve of nitric acid feed, 1548 Sensors: use of temporary instruments, Contact or laser sensor of temperature of the cooling water entering and leaving the cooling coil, 1877 Contact or laser sensor of temperature of the contents of the neutralizer, 2841 Contact or laser sensor of temperature of urea feed to the neutralizer, 2264 Contact or laser sensor of temperature of the ammonia vapor feed to the neutralizer, 2090 Contact or laser sensor of temperature of the AN leaving the neutralizer, 1849

8.2 Cases to Help you Polish Your Skill

Contact or laser sensor of temperature of the acid in the storage tank, 1749 Sensors: calibrate, Temperature sensor on the AN neutralizer, 346 Temperature sensor on the AN at the exit nozzle from the neutralizer, 118 Control system, Retune the flow control on the nitric acid, 1200 Call to vendors, licensee, or suppliers, Supplier of nitric acid: concentration and additives, 1293 Samples and measurements, Product AN at exit nozzle from the neutralizer; analyze concentration of AN, 811 Nitric acid in storage tank; analyze concentration, 2233 Nitric acid at nozzle at the entry into the neutralizer; analyze concentration, 2466 Nitric acid at sampler before acid-transfer pump; analyze concentration, 2775 Nitric acid at flow-control valve; analyze concentration, 2581 Ammonia vapor at nozzle into neutralizer; analyze concentration of ammonia, 2321 Urea off-gas at nozzle into neutralizer: analyze concentration, 1485 Open and inspect, Cooling coil; shut down; pressure test for leak in the cooling coil in the neutralizer by installing a gauge; pressurizing with air, isolating and note decrease in pressure with time, 838 Cooling coil; shut down; drain neutralizer, pressure test for leak in the cooling coil in the neutralizer by isolating and hydraulically pressurize coil; look for water leaks, 791 Cooling-coil fouling: Shut down; drain, make safe for entry. Visually inspect. Look for fouling on outside of coil and at entry and exit on the inside, 355 Trombone, steam heating coil in acid storage tank. Isolate and statically pressure test with air, gauge and observe pressure decrease with time, 936 Condition of gas sparge rings on both off-gas and ammonia. Shut down; drain, make safe for entry. Visually inspect, 1472 Acid transfer pump; isolate, drain, inspect, 2155 .

Take “corrective” action, Interchange the off-gas and ammonia feeds to nozzles on the neutralizer so that the off-gas is at the lowest nozzle, 2703 Relocate the ammonia entry nozzle and sparge so that this is below the cooling coils; install two layers of static mixers above the gas sparge to improve mixing, 2945 Install a central draft tube in the neutralizer to improve mixing in the neutralizer, 1399

361

362

8 Prescription for Improvement: Put it all Together

Case ’43: High pressure in the debut! (courtesy of T. E. Marlin, Chemical Engineering, McMaster University) [7, sequence of distillation columns with auxiliaries; depropanizer-debutanizer] The process is the depropanizer-debutanizer described in Case ’8, Chapter 2. A P&ID is given in Figure 2-4 accompanying Case ’8. This is the startup of the unit after the annual turnaround. During the turnaround several valves were replaced, pumps were dismantled and reassembled, the column internals were inspected, the heat exchangers, condensers and reboilers cleaned and the sensors calibrated. This was done for both the debutanizer, C-9, and the depropanizer, C-8. Shortly after the unit starts up there is an upset on the depropanizer and we have just resolved that in Case ’41. This Case ’41 had symptoms that the depropanizer overhead product has far too much C4 and the debutanizer overhead has far too much C3. (You don’t have to work Case ’41 before you work on this one.) Now the operator panics and calls you again. This time she thinks that the pressure is so high in the debutanizer that the safety relief valve has popped! That means butane is spewing out; the flare should be flashing! Now what? Case ’43: Debutanizer pressure relief . .

. .

.

. . .

MSDS, 1495 Immediate action for safety and hazard elimination, Put on safe-park, 111 Safety interlock shut down, 440 SIS plus evacuation, 2934 More about the process, 1010 IS and IS NOT, What, 389 When, 1481 Who, 1962 Where, 1772 Why? Why? Why?, Best goal? To manage the purported high pressure on the debutanizer, 1321 Weather, Today and past, 1065 Maintenance: turnaround, When and what done?, 2381 What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Condenser, E-28, 2006 Distillation column, C9: 2367 Thermosyphon reboiler, E-30, 1382 Reflux pump, F-29, 1112 Overhead drum, V-31, 1455 Vendor files: Condenser, reboiler, 1448 Steam traps, 1244 Commissioning data, P&ID, internal reports, 2051

8.2 Cases to Help you Polish Your Skill

Handbook and Google, Cox charts, 2114 Trouble-shooting files, 2809 .

What is current operation

Visit control room: control-room data, Feed to C9, debutanizer, FI/2, 365 Feed to the C8, depropanizer. FC/1, 601 Overhead liquid product butane, flowrate FIC/7, 994 Reflux flowrate, FIC/6, 1491 Pressure drop Dp I/2, 1102 Level bottoms LIC/4, 1701 Level in feed drum, V-31; LIC-5, 1551 Temperature bottoms TI/12, 2401 Temperature top, TI/11, 2902 Analyzer A-1, 2570 Alarm on PIC/19, 2505 Pressure on overhead, PIC/19, 2443 Process operators, This shift, 278 Call to others on-site, Utilities: any change in the cooling tower operation that might affect our site, 22 Visit site, read present values, observe and sense. Column pressure, PI-12, 458 Pressure relief to flare PSV-3, 202 Observe flare, 417 Temperature of the feed to the column TI-10, 505 Cooling-water temperature in to the condenser, TI-13, 960 Cooling-water temperature out of the condenser, TI-14, 735 Valve position for column feed, FV-2, 599 Valve position on PIC/19; downflow from the condenser to the reflux drum, 853 Signal to PIC/19, 1301 Pressure on exit of reflux pump F-29, PI-20, and compare with head-capacity curve at feed flowrate, 1253 Listen to the reflux pump F29 for sounds of cavitation, 1016 Observe whether shaft is rotating for the reflux pump F-29. Check that the direction of rotation is consistent with arrow on the housing, 1991 Check diagram and P&ID versus what’s out on the plant, 2769 On-site simple tests: Open bleed valves to fuel on the shell of the condenser for 10 minutes, then shut, 1605 Tap the side of the condenser along a vertical line to listen for change in sound associated with liquid level in the condenser, 1509 Gather data for key calculations, Pressure profile, Drum V-31 to pumps, F-29, 2863

363

364

8 Prescription for Improvement: Put it all Together

Dp across reflux pump, converted to head, 2984 Drum V-31, pump F-29 and reflux into column, 2723 Pump F-29, 2594 Thermosyphon reboiler process fluid side, 2531 Energy balance: sink= source, Estimate the steam flow to the reboiler based on the reflux rate and the fact that each kg steam boils 5 kg typical organic, 104 Sensors: check response to change, Temperature sensor at the top of column, TI-11, 1634 Pressure sensor at top of column PI-12, 2151 Pressure sensor on overhead line for PIC/19; PT-19, 2319 Pressure sensor on reflux pump exit line, PI-20, 2031 Sensors: use of temporary instruments, Measure the surface temperature on the outside of the condenser, 329 Sensors: calibrate, Temperature sensor at the top of column, TI-11, 464 Pressure sensor at top of column PI-12, 28 Pressure sensor on overhead line for PIC/19; PT-19, 702 Pressure sensor on reflux pump exit line, PI-20, 607 Control system, Check if alarm on PIC/19 is faulty or if signal to control room is faulty, 813 Put overhead pressure control, PIC/19; reflux control FIC-6 controllers on manual and try to steady out the column, 901 Sample and analyze, Current feed to the debutanizer for the amount of C3, 2488 .

Open and inspect, Condenser, 2916 Pressure control valve PV-19, 2552 Reflux pump, F-29, 2513 Line from the top of the column to the condenser, E-28, 2825 Sensor tap for pressure on the overhead line, PT-19, 1219

.

Take “corrective” action Replace the pressure control valve PV-19, 1460 Replace the pressure gauge PT-19, 1055 Replace high-pressure alarm on PIC/19, 1347 Direct water from the fire hose onto the outside shell of the overhead condenser, E-28, 786 Reduce the reboiler duty, 2983

[8, steam coil, storage tank, controller; general] Reactant must be stored at 50 – 1 C. To achieve this a steam-heated coil is placed inside the open, 2.25-m3 vertical cylindrical storage tank as illustrated in Figure 8-24. The instrument engineers designed a control system that satisfies the temperature requirement. Although the sensor, valve and controller are the fanciest on the market, the best it can do is keep the temperature 50 –10 C according to the controlCase ’44: Reactant storage

8.2 Cases to Help you Polish Your Skill

ler chart. The control engineers have worked on tuning the control loop for a week but find nothing wrong. This reactant ties up $2,000 worth of production per hour. Solve the problem. 100

Feed Steam main

30

Condensate main

T

60

40

Time Sample Chart Bucket Steam Trap

Effluent Condensate to Boiler House

Figure 8-24

The reactant storage vessel for Case ’44.

Case ’44: Reactant storage .

. .

. . .

Immediate action for safety and hazard elimination, Put on safe-park, 2671 Safety interlock shut down, 2145 SIS plus evacuation, 32 More about the process, 2805 IS and IS NOT: (based on given problem statement), What, 467 When, 1424 Who, 1132 Where, 1157 Weather, Today and past, 1858 Maintenance: turnaround, When and what done?, 2115 What should be happening

365

366

8 Prescription for Improvement: Put it all Together

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Controller, 2412 Steam control valve, 2718 Heating coil, 2811 Steam trap, 2696 Vendor files: Steam traps, 2547 Commissioning data, P&ID, internal reports, 2124 Handbook, Steam tables: Temperature for saturated steam at 0. 205 MPa g and 1.5 MPs-g, 2014 Trouble-shooting files, 2254 .

.

Calculations and estimations (that can be done in the office before special tests are done) not sufficient data given. What is the current operation

Visit control room: control-room data: values now and from past records, Steam pressure in main, 1220 Feed flowrate into the tank, 1409 Process operators, General comments about the situation, 1492 Temperature variation in the feed entering, 1037 Call to others on-site, Utilities: steam pressure and quality of steam, 1014 Visit site, read present values, observe and sense. Pressure in the condensate header, 727 Check if insulated and quality of the insulation, 1810 Check that the bypass valve on the condensate trap is closed and that the block valves are open, 935 Check diagram and P&ID versus what’s out on the plant, 2287 On-site simple tests: Open bypass on the steam control valve, 517 Open bypass on condensate trap, 2421 Open bypass on steam control valve; block off steam valve; open drain, 612 Give the bucket trap a sharp hit to dislodge any crud that might interfere with the mechanism, 402 Shut valve on exit of trap for several minutes and then open slowly to “reseal” the trap, 79 Note position of valve stem on steam valve over “two cycles” of the temperature variation, 2917 Gather data for key calculations Pressure profile, Pressure drop across the control valve on inlet side; estimate, 2462 Estimate the pressure to push the condensate up vertically to the condensate header, 2092 Mass balance, estimate the mass balance on the steam, 1668 Energy balance: sink= source, steam required = amount of heat up the reactant, 1903

8.2 Cases to Help you Polish Your Skill

Sensors: use of temporary instruments, Use contact or laser pyrometer to measure the temperatures upstream and downstream of the trap; compare with expected 200 and 134 C, 2777 Use contact or laser pyrometer to measure feed temperature into the vessel over two cycles, 2577 Use the stethoscope to listen to the bucket trap over “two cycles” of the temperature variation and compare with expected “loud initially, followed by lower-pitch bubbling and then no noise” for each cycle, 2979 Use two contact or laser pyrometers to measure the temperatures upstream and downstream of the trap over “two cycles” of the temperature variation and compare with expected 200 and 134 C, 1702 Sensors: calibrate, Temperature sensor in vessel, 1207 Pressure sensor on steam main, 705 Control system, Put controller on manual, 2687 More complex tests, Add visible tracer to observe mixing patterns in vessel, 1560 Open and inspect, Block off steam trap; operate on bypass; open strainer and check for clogging. If clogged, clean and restart, 588 Steam heating coil, 156 Bucket steam trap; then clean if necessary and restart, 408 .

Take “corrective” action, Insulate vessel, 666 Repipe steam connections to coil so that steam comes in the top and condensate to trap is out the bottom of the coil, 1093 Add a portable, top-entry propeller mixer, 1494 Relocate the feed pipe entry under the surface, 2898 Replace bucket trap with a float trap, 2530 Install a check valve on the line to the condensate header, 1763

(courtesy of T. E. Marlin, Chemical Engineering Dept., McMaster University) [8, sequence of distillation columns plus auxiliaries; petrochemical, refinery] The process is the depropanizer-debutanizer described in Case ’8, Chapter 2. A P&ID is given in Figure 2.4 accompanying Case ’8. To be able to sell your products, your plant must obtain ISO certification. As a result, you have established a routine analysis of various streams in the depropanizer-debutanizer system. The initial laboratory analyses indicate too much variability in the mole fraction of propane in the bottoms of the depropanizer, C-8. For the last day, the mole fraction has been about 0.04, while the target is 0.015. Before the new procedures, we never knew that we were operating the plant so poorly, so no one cared! Now everyone is frustrated that you have found this fault. If you cannot obtain ISO certification, the company will not be able to sell products to key customers. Resolve the problem. Case ’45: The deprop bottoms and the ISO dilemma

367

368

8 Prescription for Improvement: Put it all Together

Case ’45: The deprop bottoms and the ISO dilemma . .

. .

.

. .

MSDS, 1495 Immediate action for safety and hazard elimination, Put on safe-park, 1882 Safety interlock shut down, 1531 SIS plus evacuation, 732 More about the process, 1010 IS and IS NOT, What?, 1477 When?, 2237 Where?, 2537 Why? Why? Why?, Best goal? To reduce the propane concentration in the bottoms to within specifications, 2211 Weather, Today and past, 2303 Maintenance: turnaround, When and what done?, 1421

Maintenance: routine, When and what done?, 2044 .

What should be happening,

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Condenser, E-25, 591 Distillation column, C8: 515 Thermosyphon reboiler, E-27, 2380 Feed pump, F-25; F-26, 16 Turbine drive, 1193 Motor drive, 1072 Reflux pump, F-27, 286 Feed preheater, E-24, 1989 Feed drum, V-29, 1478 Overhead drum, V-30, 1124 Vendor files: Condenser, reboiler and preheater, 10 Steam traps, 481 Commissioning data, P&ID, internal reports, 245 Handbook, Cox charts, 803 Trouble-shooting files, 1430 .

Calculations and estimations

Equipment performance, Nucleate or film boiling in reboiler? check DT, 642 Energy balance: Estimate the steam flow to the reboiler based on the reflux rate and the fact that each kg steam boils 5 kg typical organic, 2821 .

What is current operation

Visit control room: control-room data, Control temperature on the C8, depropanizer. TC/5, 1563 Bottoms temperature. TI/4, 1156

8.2 Cases to Help you Polish Your Skill

Steam flow to reboiler. FIC/3, 2078 Debutanizer overhead analyzer for C3: AC/1, 436 Level in bottoms, LC/2, 103 Flow of bottoms to the debut, FI/2, 2251 Feed flowrate to column. FIC/1, 766 Process operators, This shift; variation in AC/1?, 1334 Previous shift: variation in AC/1, 433 Call to others on site, Operators of facility receiving the butane as overheads from the debutanizer: propane contamination in product, 219 Visit site, read present values, observe and sense, Column temperature on tray ’9, TI-8, 1303 Observe the flare, 2815 Valve-stem position on LV/2, bottoms flowrate from the deprop, C-8, 2989 Valve-stem position on FV-3; steam to reboiler, 2614 Check diagram and P&ID versus what’s out on the plant, 2769 On-site simple tests: Put control of steam to reboiler E-27 on manual and control based on TC-5, 2585 Trend check, do trends in the lab sample results follow the same trends in the readings from AC/1?, 652 Control system, Retune the control system for the bottoms of the deprop, C-8. TC/5, FC/3 and LC/2, 894 Operator adjust the set point on TC-5 based on the measurement of A-1, 201 Sensors calibrate, Calibrate sensor on the top of the debut. A/1, 454 Calibrate temperature sensor, TC/5 on tray ’9 of the column, 952 Samples and measurements, Sample feed to C-8, the deprop and check for > usual concentration of C3, 1202 Sample the bottoms from C-8 once per hour for three hours. Flush sample line well before collecting each sample. Analyze in lab for C3, 1107 Draw three samples of the bottoms from C-8; send one sample to our lab and others to two independent labs. Analyze for C3, 1453 Are the lab sample results consistent with the readings from AC/1?, 2102 Open and inspect, Reboiler: pull bundle and check for fouling on both inside the tubes and outside, 1952 Column at tray ’9 and check that TI-8 and TC-5 are “well situated” and the wells are not corroded; that the downcomer is sealed and the tray is level, 1651 Column at bottoms and check that vortex breaker is present and that thermowell for temperature sensor TI-4 is “well situated” and the well is not corroded. Level sensor is OK?, 1051

369

370

8 Prescription for Improvement: Put it all Together .

Take “corrective” action Install a C3 analyzer on the feed into the column, 2751 Replace the thermocouple at TC-5; retune the controller, 1391 Replace the steam trap on the condensate leaving reboiler, E-27, 1007

Case ’46: The not so cool chiller (courtesy of Scott Lynn, Chemical Engineering Department, University of California, Berkeley) [8, pump, exchanger, ethylene refrigeration; polymerization] The system below is designed to prepare butene in methyl chloride solvent for polymerization. This reactor feed is to be dried, condensed and cooled to–29 C with ammonia refrigerant, and finally chilled to –96 C with boiling C2H4 in a thermosyphon chiller. The ethylene tube-and-shell chiller will cool the reactor feed to –96 C at only half the design flowrate or to only –60 C at the design flowrate. This reduces polymer production by 0.45 Mg/h at an out-of-pocket loss in profit of $1/kg. Get this going correctly. The process is shown in Figure 8-25.

Feed at design rate, 450 kPa g; 25C

FRC 1

-101C

TI 4

PRCV set at 22 kPa g

FRC 2

Ethylene head tank

Alumina dryer

Liquid ethylene

LRC 1

-101C Ammonia

-29C

TI 3

TI 1

TI 2

Chiller Thermosyphon Feed pump

Figure 8-25

The chiller for the reactant feed for Case ’46.

To Reactor

8.2 Cases to Help you Polish Your Skill

Case ’46: Not so cool chiller . .

. .

. .

MSDS, 423 Immediate action for safety and hazard elimination, Put on safe-park, 148 Safety interlock shut down, 1282 SIS plus evacuation, 1029 More about the process, 1136 IS and IS NOT: (based on given problem statement), What, 2337 When, 2096 Where, 2896 Weather, Today and past, 2633 Maintenance: turnaround, When and what done?, 2332

Maintenance: routine, When and what done?, 1723 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Chiller, 1616 Feed pump, 1223 Drier, 1061 Ammonia condenser, 647 Feed drum, 823 Ethylene head tank, 312 FRC controller, 1937 Vendor files: Feed pump, 2924 Commissioning data, P&ID, internal reports, 2680 Handbook, Methyl chloride: vp at 25 C and compare with pressure on vapor of 450 kPa g, 2555 Heat capacities of butene, water, methyl chloride and ammonia, 2640 Trouble-shooting files, 2827 Calculations and estimations (that can be done in the office before special tests are done) Pressure profile, From inlet to drier at 450 kPa g to feed drum, 2598 Direction of leak: ammonia condenser, 343 Direction of leak: chiller, 134 From the vapor from the heads tank to the ethylene compressor, 792 From feed drum to reactor, 613 Energy balance: sink= source, vaporisation load: design load vs; current full load and half-load, 907 Rate, In chiller, nucleate vs film boiling on shell side for full versus half-load, 863 Equipment performance, Chiller: rating comparison of UA for full load vs half-load, 746 .

371

372

8 Prescription for Improvement: Put it all Together

Chiller: tube side and shell side transfer coefficient, 1429 Compressor-condenser in the refrigeration loop, 1187 .

What is the current operation

Visit control room: control-room data: values now and from past records, Flowrate of vapor to the driers, FRC, 1095 Feed temperature to chiller, TI, 1302 Temperature of the process fluid exit of chiller, 1397 Temperature of the ethylene vapor off the chiller, 1780 Flow of ethylene off top of head tank, FR, 1942 Setting on PRCV on ethylene head tank, 2222 Process operators, Please tell me about the operation so far, 2471 Contact with on-site specialists, Contact operators of reactor; temperature and flowrate to reactor consistent with the temperatures and flowrates from chiller unit, 2193 Unit supplying and receiving the ethylene, 2788 Visit site, read present values, observe and sense. Feed pump: sounds like cavitation? 2584 Valve position for vapor feed at FRC/1, 2697 Tab on orifice: indicating that the plate has been put in correctly, 2914 Check that the block valve on bypass around the FRC on the vapor feed to the dryers is CLOSED, 2217 Liquid level in ethylene head tank, 2084 Valve position for liquid ethylene feed, 2042 Temperature on vapor line off top of ethylene head tank, 1625 Check that the two valves on the vapor bypass line are fully closed. Check that the “flow direction” through the valve is correct, 1817 Check that the valve on the ethylene thermosyphon line is fully open and that the “direction of flow” through the valve is correct, 714 Check that the bypass valve on the feed pump is fully closed and that the “direction of flow” through the valve is correct, 875 Check that the valve on the pump discharge is fully open. Check that “flow direction” on the valve is correct, 1869 Check diagram and P&ID versus what’s really out on the plant, 1711 On-site simple tests: Adjust the PRCV to a slightly higher pressure and note temperatures, 1693 Complete a “turn and seal” test on two valves on the vapor bypass line. Ensure the valves are left shut, 1592 Complete a “turn and seal” test on the valve on the ethylene thermosyphon line. Ensure the valve is fully open, 1929 Complete a “turn and seal” test on the bypass around the pump. Ensure the valve is fully closed, 2739 Tap side of the ethylene head tank to try to detect liquid level. Compare height with level gauge reading, 2616 Tap side of the feed drum to try to detect liquid level, 2173

8.2 Cases to Help you Polish Your Skill

Tap side of the chiller to try to detect the liquid level. Compare with the level of exchanger tubes. Are the tubes completely covered by the ethylene, 2340 Consistency checks, Check that temperature and pressure measured at top of chiller are consistent for “pure” ethylene, 1821 Temperature sensors: TI/3 on ethylene chiller agree with TI/4 on top of head tank, 1688 Process temperature from chiller exit agree with downstream unit, 1529 Sensors: check response to change Temperature sensor of the ethylene vapor TI/3, 1071 Temperature sensor of the process liquid out of chiller; TI/2, 1213 Sensors: use of temporary instruments, Surface or laser temperature sensor for the overhead ethylene from the chiller, 127 Use clamp-on ammeter to measure the amps on the motor driving the feed pump; compare with specs, 397 Sensors: calibrate, Temperature sensor of the ethylene vapor TI/3, 756 Temperature sensor of the process liquid out of chiller; TI/2, 1333 Control system, Retune the FRC controller on the feed to the drier, 1042 Retune the LRC controller on the ethylene heads tank, 2298 Samples and measurements, Sample ethylene and analyze for contamination, 2440 Sample process liquid from the feed drum. Analyze for water and ammonia, 2875 More complex tests, Steam regenerator for the drier system. Isolate, once conditions are safe, hydraulically pressure test to identify leaks in the tubes or between the tubesheet and the tubes. Any leak of steam into the drier system might possibly load the adsorbent with water, 2538 Gamma scan to determine the interface location in the chiller, 2947 Open and inspect, Chiller, check for fouling. Isolate, drain, once conditions are safe, pull out the bundle and check for fouling inside and outside tubes; use optic fiber probe as needed, 2091 Feed pump. Isolate, when conditions are safe, open and inspect, 2256 Line from the heads tank to the chiller for “thermosyphon operation”: isolate, when safe, open and inspect for blockages. Use optic fiber probe as needed, 1161 .

Take “corrective” action, Replace the two valves on the vapor bypass around the chiller, 776 Replace the globe valve on the thermosyphon line from the head tank with a ball valve, 214

373

374

8 Prescription for Improvement: Put it all Together

[8, reactor, vaporizer, condensers, liquid ring vacuum pump, control; petrochemical] To produce acetic anhydride from acetic acid, via the Wacker process, about 0.6 kg/s of acetic acid is vaporized and sent into a series of three cracking coils that are operated under vacuum. The three cracking coils are inside a furnace as is illustrated in Figure 8-26. But producing acetic anhydride is not a simple process. The acetic acid is cracked in the furnace to produce ketene and water. The reaction is reversible. If the water is not condensed and removed rapidly from the reactor effluent, the reverse reaction could occur and the product from the reactor would be acetic acid – the reactant you started with. If the removal of the water is successful, then gaseous ketene leaves the reactor condenser system and reacts with fresh liquid acetic acid in an absorber to produce acetic anhydride. The condensers for the water are a series of Liebig condensers using first water and then brine as the coolant. A further complication is that ketene dimerizes and forms a gunk that plays havoc with the operation of the reciprocating vacuum pumps. Case ’47: The fluctuating production of acetic anhydride

ACETIC ACID VAPOUR

VACUUM PUMPS

AcOH LIQUID

AcOH

STEAM FURNACE

Figure 8-26

WATER

WATER

The furnace reactor and condensers to produce acetic anhydride in Case ’47.

PI 201

DR

FI 201

TO CRACKING FURNACE

A 201

r

FRC 202 PI 101

LC

LIQUID ACETIC ACID

Figure 8-27

STEAM MAIN TO CONDENSATE HEADER The vaporizer for the feed to the furnace in Case ’47.

8.2 Cases to Help you Polish Your Skill

The system has experienced serious upsets. The amount of acetic anhydride product fluctuates greatly. Some suspect the vaporizer that is shown in Figure 8-27. Some think that the condensation of the water from the ketene is the source of the fluctuations. Still others point to the vacuum pump. Perhaps it is the absorber. Sort it out! and sort it out fast! Case ’47: Fluctuating production of acetic anhydride . .

.

. .

MSDS, 339 Immediate action for safety and hazard elimination, Put on safe-park, 203 Safety interlock shut down, 2653 SIS plus evacuation, 2080 IS and IS NOT: (based on given problem statement), What, 2361 When, 2061 Where, 1654 Weather, Today and past, 1542 Maintenance: turnaround, When and what done?, 1204

Maintenance: routine, When and what done?, 1021 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Cracking reactor, 46 Condenser, 117 Vacuum pumps, 363 Orifice meter FRC/201, 474 Absorber-reactor, 985 Vaporizer, 622 Steam trap, 541 Gas-fired furnace for the reactor coils, 1433 Vendor files: Vacuum pump, 1038 Commissioning data, P&ID, internal reports, 1023 Handbook, Properties of acetic acid, 1267 Trouble-shooting files, 1363 Calculations and estimations (that can be done in the office before special tests are done) Pressure profile, Pressure drop through reaction coil, 1681 Direction of flow if there is a leak: in the condenser, 1981 Mass balance, on acetic anhydride produced, 2104 .

375

376

8 Prescription for Improvement: Put it all Together

Energy balance: sink= source, On condensers: heat removed in water/brine = heat to be removed from condensing reactant, 2454 Thermodynamics, Vapor-liquid equilibrium in KO pots between condensers, 2963 Rate, Rate and catalyst addition in reaction coils, 2670 Boiling characteristics in vaporizer: nucleate or film, 2553 Equipment performance, Condensers: heat-transfer coefficients, 2515 .

What is the current operation

Visit control room: control-room data: values now and from past records, Steam pressure PI 101, 2973 Acetic acid flowrate FRC 201, 2590 Process operators, Please describe the operation so far, 2153 Contact with on-site specialists, Steam plant: pressure and steadiness of steam supplied to battery limits, 1607 Steam plant: condensate return header, 1927 Source of acetic acid; specs agree with those used in the design, 229 Cooling water to condensers, 56 Visit site, read present values, observe and sense. Inspect the tab on the orifice plate to determine if the plate is in backwards and to note the size of the orifice, 370 Does condensate discharge into the top of the condensate header, 931 Variation in the pressure and flowrate of the fuel to the cracking furnace, 768 Upstream and downstream distance from the orifice plate to see if distance meet usual design specs, 598 Duration of the cycle of the variation in production, 1164 Check diagram and P&ID versus what’s out on the plant, 2787 On-site simple tests: Glove test on the steam trap to estimate the temperature of the trap upstream and downstream, 1706 Stethoscope on trap to listen is discharge of condensate, 2203 Tap the side of the evaporator to try to discern the level of acid in the vaporizer, 2057 Consistency checks, Pressure on steam header on-site and the pressure at the steam drum, 2402 Sensors: calibrate, LIC on vaporizer, 855 FRC 201, 1256 Pressure in acetic acid vapor line PI 201, 1872 Steam pressure to coil PI 101, 2220 Control system, Controller on manual and check the FRC 201, 2604

8.2 Cases to Help you Polish Your Skill

Controller on manual and read PI 201, 2933 Controller on manual and check for steadiness in the downstream production, 560 Call to vendors, licensee, Vacuum pump, 2905 More complicated tests, Gamma scan on vaporizer to locate level relative to the coil, 715 Open and inspect, Evaporator and look for fouling, 2493 Isolate coil in vaporizer and pressure test for leaks, 2148 Orifice plate FRC 101 to see if it is put in backwards, 2702 Isolate, inspect cracking coils for carbon formation inside the tubes, 2548 .

Take “corrective” action, Retune the control system on the vaporizer, 253 Replace the bucket trap with a float trap, 889

(courtesy of T. E. Marlin, Chemical Engineering, McMaster University) [8, sequence of distillation columns with auxiliaries; depropanizer-debutanizer] The process is the depropanizer-debutanizer described in Case ’8, Chapter 2. A P&ID is given in Figure 2.4 accompanying Case ’8. This is the startup of the unit after the annual turnaround. During the turnaround several valves were replaced, pumps were dismantled and reassembled, the column internals were inspected, the heat exchangers, condensers and reboilers cleaned and the sensors calibrated. This was done for both the debutanizer, C-9, and the depropanizer, C-8. Shortly after the unit starts up the operator is in a panic because, according to the laboratory analysis, the depropanizer overhead product has far too much C4 and the debutanizer overhead has far too much C3. The plant is losing $1000s per hour and the plant manager is furious. Fix the problem. Case ’48: The column that just wouldn’t work

Case ’48: The column that just wouldn’t work . .

. .

. .

MSDS, 1495 Immediate action for safety and hazard elimination, Put on safe-park, 227 Safety interlock shut down, 2874 SIS plus evacuation, 2619 More about the process, 1010 IS and IS NOT, What, 2907 When, 2240 Where, 1680 Why? Why? Why?, Best goal? To meet specs on overhead and bottoms, 1366 Weather, Today and past, 1065

377

378

8 Prescription for Improvement: Put it all Together .

Maintenance: turnaround, When and what done?, 181 Was the vortex breaker in the reflux drum V-30 in “good” shape, 24

.

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Condenser, E-25, 725 Distillation column, C8: 519 Thermosyphon reboiler, E-27, 1673 Reflux pump, F-27, 1524 Overhead drum, V-30, 2324 Vendor files: Condenser, reboiler and preheater, 2915 Reflux pump, F-27, 1755 Steam traps, 1535 Commissioning data, P&ID, internal reports, 1863 Handbook and Google, Cox charts, 1162 Trouble-shooting files, 664 .

What is current operation

Visit control room: control-room data, Feed to the C8, depropanizer. FC/1, 1820 Reflux flowrate, FIC/4, 1712 Overhead product flowrate of propane, FIC/5, 1628 Pressure drop Dp I/1, 2309 Level bottoms LIC/2, 2026 Level in reflux drum: LIC/3, 2705 Steam to reboiler, FIC/3, 2760 Temperature of feed into the column, TIC/2, 2974 Analyzer A-1; concentration of C3 in bottoms of C8 and feed to C9, 1573 Temperature bottoms TI/4, 1515 Temperature mid-column TIC/5, 1077 Temperature top, TI/3, 1374 Pressure on overhead drum, PIC/10, 673 Process operators, This startup shift, 943 Any change to the feedstock to this unit?, 974 Operating procedures, Column temperature too high; then increase the reflux, 321 Call to others on-site, Utilities: is the steam-production rate and steam temperature and pressure the usual values?, 123 Utilities: is the cooling-water flowrate and temperature to our unit what we expect?, 39 Operators of plants downstream receiving the butane, propane and pentane, 2731 Operators of upstream plant providing feed, 1732

8.2 Cases to Help you Polish Your Skill

Visit site, read present values, observe and sense. Column pressure, PI-4, 412 Pressure relief to flare PSV-1, 2425 Look at the flare, 882 Temperature mid-column TI-8, 2183 Valve position for column feed, FV-1, 1766 Valve position for reflux, FV-4, 1913 Valve position for propane overhead product to downstream processing, FIC-5, 1695 Pressure on the reflux pump exit, F-27 as shown on PI-11, 2277 Is there noise of cavitation around reflux pump F-27, 2127 Are the leads to the motor drive on reflux pump F-27 correct so that the motor turns in the “correct” direction, 2925 Check that the direction of rotation of the pump shaft is correct, 2593 Valve position for steam to preheater E-24, 2138 Valve-stem position on PV-10, 2366 Level in feed drum, V-29; LI-1, 2106 Isolation valves around the reflux control valve, FIC/4, 1655 Isolation valves around the reflux pump, F-27, 1747 Check diagram and P&ID versus what’s out on the plant, 2769 On-site simple tests: Open bypass around the reflux control valve, FIC/4; note any change in overhead temperature TI-3 and flow of reflux FIC/4, 1217 Gather data for key calculations, Pressure profile, Drum V-29 to pumps, F-25, 26, 529 Dp across pump, F-25, 26 converted to head, 96 Pump F-25–26 exit to feed location, 463 Drum V-30, pump F-27 and reflux into column, 164 Pump F-27, 868 Vapor from top of column to vapor space in V-30, 582 Thermosyphon reboiler process fluid side, 1405 Mass balance, Over column, 1287 Energy balance: Heat load for condenser. Heat load condensed = heat load picked up in cooling water, 1015 Steam load to reboiler based on 1 kg steam boils 5 kg organic, 1088 Water flowrate to overhead condenser consistent with steam flowrate at bottoms. 15 L water / kg steam, 1426 With data from hand-held voltmeter, clamp-on ammeter and power-factor meters calculate the power used by pumps F-25,–26 and compare with specs, 890 Sensors: check response to change, Temperature at the top of column C8: TI 3, 572 Analyzer A-1, 76

379

380

8 Prescription for Improvement: Put it all Together

Flowmeter for reflux, FE-4, 375 Pressure at the top of column C8, PI 4, 747 Sensors: calibrate, Temperature at the top of column C8: TI 3, 1690 Analyzer A-1, 2257 Flowmeter for reflux, FE-4, 2270 Pressure at the top of column C8, PI 4, 2312 Control systems, Output signal from controller to reflux valve FIC/ 4, 1246 Put reflux control system, FIV/4, on manual; increase the flowrate, 1795 Samples and measurements, Sample feed and analyze composition and compare with usual, 2158 Sample overhead concentration in vapor from column every 10 minutes for 1 hour. Analyze for composition, 2196 Sample bottoms concentration from the tower every 10 minutes for 1 hour. Analyze for composition, 2088 .

.

.

More ambitious tests, Gamma scan on column to look for collapsed tray(s), 2730 Open and inspect, Condensers; E-25, check baffles and for fouling, 2694 Access hole to column and check if trays are level and sealed; for those trays that can be seen from access hole, 2855 Reflux pump: check for damage to impeller, wear rings are not worn, F-27, 2891 Reflux control valve: check stem, trim, FV/4, 2803 Take “corrective” action, Shut down the column; take out all the trays and add weirs to “seal” the selfsealed downcomers, 2284 Replace the check valve on the exit line of the reflux pump, F-27, 2349 Replace the control valve on the reflux, FIC/4, 2039 Reduce feedrate to the column to 1/2, 1647 Put on safe hold, 1604

(from D.R. Winter, Universal Gravo-plast, Toronto, 2004) [Difficulty 8; involves feed bin, molding machine, mold and mold design. Context: injection molding of thermoplastics] The client manufactures hospital stretchers, sometimes called gurneys. The plastic component the client wishes molded is a foot pedal that would stop the stretcher from rolling. The peal is 24 cm long with a central hub about which the pedal functions like a teeter-totter so that the pedal can be pushed from either direction. The foot part of the pedal is 6.5 cm across and ribbed. The pedal is 2.6 cm thick with webs of 3 mm thickness. The central hub is 3.2 cm in diameter with wall thicknesses of 1 cm. Ribs support the under part of the pedal because the torque created Case ’49: The case of the faulty stretcher pedal

8.2 Cases to Help you Polish Your Skill

is substantial. A metal core was used in the mold to create the hole for the axle, over which the pedal was fitted. The resin selected was an alloy of polycarbonate and ABS that provides an impact strength of 640 J/m. A mold was designed, tests were done and the customer approved the prototype. Production was started. After several months in the field, some pedals were breaking in Hong Kong, California, British Columbia. The breaks seemed to be independent of the hospital location. An analysis of the broken pedals suggested that the breakage occurred at the hub; inspection showed weld or knits lines. Solve the problem! Figure 8-28 is a sketch of the pedal.

Figure 8-28

The pedal.

Case ’49: The case of the faulty stretcher pedal . .

. . .

.

MSDS,300 Immediate action for safety and hazard elimination, Put on safe-park, 947 Safety interlock shut down, 523 SIS plus evacuation, 1584 More about the process, 2282 More about the product and the mold, 2093 IS and IS NOT: (based on given problem statement), What, 2729 When, 2523 Who, 191 Where, 1222 Weather, Today and past, 1343

Maintenance: routine, When and what done?, 1791 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Injection molding machine, 2473 Mold, 2704

381

382

8 Prescription for Improvement: Put it all Together

Feed for this product, 1862 Commissioning data, P&ID, internal reports, Prototype development, 1522 Handbook, Data sheets for resin, 1691 Trouble-shooting files, 2118 .

What is the current operation

Visit control room: control-room data: values now and from past records, Fill cycle, 2314 Cool cycle, 2892 Open cycle, 2629 Total cycle, 384 Feed temperature into mould, 721 Injection pressure, 904 Extruder: rear-barrel temperature, 1076 Extruder: head temperature, 1419 Shot to cylinder size, 1939 Concentration of foamer, 1622 Backpressure, 1980 Mold temperature, 2447 Process operators, 2007 Operating procedures, Shutdown procedures used, 2591 Cleaning procedures used, 2673 Were the feed materials from the same batches of resin, coloring agent and blowing agent as was used for the prototype?, 149 .

Check with colleagues about hypotheses, 344

On-site simple tests: Increase injection speed by 10%, 623 Decrease injection speed by 10%, 873 Change injection temperature, increase to 243 C and keep the cooling time the same, 990 Change feed temperature, increase to 243 C and increase the cooling time, 1022 Clean the mold and extruder before the run by sending through a charge of acrylic. Then use standard conditions, 1308 Increase the foamer from 0.5 to 2.1% and maintain the same conditions otherwise, 1479 Increase the foamer from 0.5 to 0.8% and maintain the same conditions otherwise, 1608 Change mold to 1 feed gate near the hub with a fill cycle of 8 s; all the rest the same, 1741 Increase the mold temperature to 65 C, 1965 Increase the mold temperature to 90 C, 2267 Increase the cold water flow to the mold, 2053 Decrease the cold water flow to the mold, 2693

8.2 Cases to Help you Polish Your Skill

Increase the backpressure from 400 to 500 kPa, 2556 Change the foamer from an exothermic to an endothermic type, 95 Increase rpm from 160 rpm to 190 rpm, 373 Use the same mold, same resin, same operator but move from a different molding machine, 782 Use the same mold, same resin, same machine, same conditions but different operator, 965 Use the same mold, same operator, same machine, same conditions but resin from a different supplier, 1159 Use the same mold, same resin, same machine, same conditions, same operator but delete the coloring agent, 1126 Separately dry the resin pellets just before using and maintain the same conditions, 1754 Gather data for key calculations, Calculations of the heat loss through the walls for various parts of the mold, 160 Redesign of product and mold, One feed gate instead of two, 653 Longer plastic neck around the axle from 2.5 to 3.5 cm, 807 Realignment of axle so that plastic over the axle is centered over the pressure on the pedal, 1388 Sensors: check response to change Temperature sensor at nozzle, 1049 Temperature sensor at first stage, 1726 Sensors: use of temporary instruments, Use a pyrometer or laser sensor to measure melt temperature and compare with temperature sensor, 1597 Call to vendors, licensee, or suppliers, Properties of feed polycarbonate ABS blend: 2249 Properties of foamer, 2784 Properties of coloring agent, 2172 Samples and measurements, Measure moisture in resin, 1783 More complicated tests, Change the mold from a aluminum to P-40, 1557 (courtesy W. F. Taylor, B. Eng. 1966, McMaster University) [9, vacuum distillation column plus auxiliaries; ammonia] The solvent sulfinol is stripped of unwanted degradation product heavies in a solvent recovery column. The solvent is then recycled for reuse. The column, 0.75 m diam 10.5 m high, treats a small sidestream of organic solvent to remove impurities as the bottoms. The feed flowrate is 0.06 to 1. L/s. Feed enters at tray ’5 and the recovered solvent is taken off at the tray above, ’4, and pumped, via a positive displacement pump, 114-J, to the sulfinol storage tank, about 15 m away. The top four trays act as a water wash to prevent overhead losses. The bottom ten trays do the separation. Live stripping steam is injected into the bottoms via valve HCV-13. Most of the steam eventually goes overhead. The wash water fed to tray ’1 Case ’50: The cleanup column

383

384

8 Prescription for Improvement: Put it all Together

is sufficient to condense all the organic product and some water; the product contamination is about 20 to 50% water. However, most of the overhead stream going to the vacuum equipment is water. The column is shown in Figure 8-29. 200# steam

CW

200# steam

CW

To atmos.

TR 100 P 100

PI 101

T 101 CW

Product

4 5

FIC-64 Reflux Water

14 FIC

103E

200# Steam 138-C

FEE D

Condensate

HCV-13

P 201

T 201

50# Sparge Steam

Drum Disposal of Bottoms 114-J

115-J

To Storage

Product to Storage

Figure 8-29

From Storage

The vacuum distillation unit for Cases ’50 and 51.

The vapor pressure of the steam is about 100 times that of the organic product at the tower conditions. The vapor pressure of the organic product, in turn, is about 10 times that of the bottoms. The tower operates under vacuum developed by a two-

8.2 Cases to Help you Polish Your Skill

stage steam ejector system. This particular column is operated for a day once a week so that production downtime costs are not a significant contribution to any problems on this plant. The problem: The tower has never worked properly. Overhead losses of organic into the hot well are probably high, but they have not been monitored. Operating conditions are rarely stable. No standard operating procedure has been developed. The bottoms product typically contains 10% organic solvent, whereas the column was designed to produce a bottoms concentration of < 2% sulfinol. But today is particularly frustrating. It was a cold and blustery night and product pump 114-J is just not delivering. The amount of “overhead product” being pumped to storage is almost zip! Case ’50: Cleanup column . .

. .

. .

MSDS, 719 Immediate action for safety and hazard elimination, Put on safe-park, 67 Safety interlock shut down, 2715 SIS plus evacuation, 2068 More about the process, 589 IS and IS NOT: (based on given problem statement), What, 2968 When, 986 Where, 574 Weather, Today and past, 753 Maintenance: turnaround, When and what done?, 157

Maintenance: routine, When and what done?, 1410 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Column, 1122 Feed pumps system, 1205 Product pump system, 114-J, 1305 Vacuum system, 1354 Vendor files: Pump 114-J, 1486 Commissioning data, P&ID, internal reports, 1964 Handbook, Sulfinol composition, 1900 Expected density of liquid product, 1750 Steam tables, 1650 Trouble-shooting files, 1540

385

386

8 Prescription for Improvement: Put it all Together

Calculations and estimations (that can be done in the office before special tests are done) Rate, Use DT to estimate whether boiling in reboiler is nucleate or film, 2908 Equipment performance, insufficient data available to do calculations .

.

What is the current operation

Visit control room: control-room data: values now and from past records, Pressure of sparge steam P/ 201, 2605 Temperature of sparge steam, T/201, 2230 Temperature of water in the barometric leg, 1850 Temperature of the overhead TR/ 100, 1950 Pressure at the top of the column PI/100, 1674 Pressure in the overhead line to the ejectors. PI/ 101, 847 Flowrate of feed to column PI/101, 946 Process operators, When the process was shut down the last time, was the pump 114-J drained to prevent possible freezing in the idle pump, 695 Please describe what has been done so far, 650 Contact with on-site specialists, Utilities: steam pressure and flow to unit stable, 132 Utilities: cooling water to reflux cooler and to the barometric condenser: temperature and flow, 419 Utilities: reflux water; temperature, flowrate and composition, 1973 Visit site, read present values, observe and sense. Is the steam tracing on, 2374 Compare speed of pump 114-J with specs, 2475 Compare pump stroke on 114-J with specs, 2868 Pressure drop across trays 1 to 14, 797 Pressure drop across trays 1 to 4, 2517 Pressure drop across trays 4 to 14, 1393 Flowrate of reflux water, FIC/64, 2030 Temperature at the bottom of the column, 2413 Temperature of the feed to the column, 1612 Listen to pump 114-J for sounds of cavitation, 1854 Read level glass on the sump at tray 4, 1525 Check diagram and P&ID versus what’s out on the plant, 1336 On-site simple tests: Close bypass valve on 114-J, 1443 Shut-off steam tracing, wait for h and note performance, 1039 Block off pump 114-J and open bypass direct to storage. Observe liquid level on tray 4 and the “bulls eye” on the line to product storage, 761 Open suction line on 114-J; drain, 524 Reduce the flowrate of reflux water, check bottoms temperature and sample/ analyze bottoms, 211

8.2 Cases to Help you Polish Your Skill

Install a pressure gauge on the suction of 114-J and note readings under present, usual operation and compare with past, 393 Pressure at top of the column; compare P/100 with PI/101, 282 Gather data for key calculations Pressure profile, Calculated NPSH supplied to pump 114 J, 2141 Pressure profile from tray 4 to storage, 2281 Mass balance, On sulfinol over the still, 2802 Samples and measurements, Measure flowrate of water from the hot well and compare with design value, 2895 Sample water from the hot well and analyze concentration of organic, 2691 More complicated tests, Stop feed to column. Maintain vacuum conditions. Unhook suction line from pump 114-J; hook up water line to the suction line and blow water backward up the suction line and watch the level via the sight glass on tray 4, 2785 Install a pressure gauge on the suction of 114-J, disconnect the dP and measure the pressure in the column at tray 4. Hook up a portable liquid pump; start pump, measure liquid flowrate and compare the Dp measured with an estimate Dp for flowrate, 2857 Gamma scan around tray ’4 to locate liquid level in on tray 4 and in the product sump, 1420 Open and inspect, Inspect foot valve on 114-J and compare valve size with the size of suction line. Correct as needed, 2117 Check for plugs, obstructions in the suction pipe from Tray 4 to pump 114-J, 2386 Check the strainer on the suction line to pump 114-J for plugs and obstructions, 1227 Open column and look for blockage in the nozzle leaving product sump and the condition of the vortex breaker, 221 .

Take “corrective” action, Replace two valves on the steam to both ejectors, 834 Replace bypass valve on 114-J, 1880 Replace faulty pressure relief valve on pump 114-J, 2177

(courtesy W. K. Taylor, B. Eng. 1966, McMaster University) [9, vacuum distillation column plus auxiliaries; ammonia] The diagram of this unit is given in Case ’50. The solvent sulfinol is stripped of unwanted degradation product heavies in a solvent-recovery column. The solvent is then recycled for reuse. The column, 0.75 m diam 10.5 m high, treats a small sidestream of organic solvent to remove impuriCase ’51: The cleanup column revisited

387

388

8 Prescription for Improvement: Put it all Together

ties as the bottoms. The feed flowrate is 0.06 to 1.0 L/s. Feed enters at tray ’5 and the recovered solvent is taken off at the tray above, ’4, and pumped, via a positive displacement pump, 114-J, to the sulfinol storage tank, about 15 m away. The top four trays act as a water wash to prevent overhead losses. The bottom ten trays do the separation. Live stripping steam is injected into the bottoms via valve HCV-13. Most of the steam eventually goes overhead. The wash water fed to tray ’1 is sufficient to condense all the organic product and some water; the product contamination is about 20 to 50% water. However, most of the overhead stream going to the vacuum equipment is water. The vapor pressure of the steam is about 100 times that of the organic product at the tower conditions. The vapor pressure of the organic product, in turn, is about 10 times that of the bottoms. The tower operates under vacuum developed by a twostage steam-ejector system. This particular column is operated for a day once a week so that production downtime costs are not a significant contribution to any problems on this plant. The problem: The tower has never worked properly. Overhead losses of organic into the hot well are probably high, but they have not been monitored. Operating conditions are rarely stable. No standard operating procedure has been developed. The bottoms product typically contains 10% organic solvent, whereas the column was designed to produce a bottoms concentration of < 2% sulfinol. Get the bottom composition within design specs. And, by the way, the product pump 114-J is sporadic; just not delivering the amount of product we expect, consistently. Case ’51: More about the cleanup column . .

. .

. .

MSDS, 719 Immediate action for safety and hazard elimination, Put on safe-park, 67 Safety interlock shut down, 2715 SIS plus evacuation, 2068 More about the process, 589 IS and IS NOT: (based on given problem statement), What, 2968 When, 986 Where, 574 Weather, Today and past, 2753 Maintenance: turnaround, When and what done?, 157

Maintenance: routine, When and what done?, 1410 .

What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties)

8.2 Cases to Help you Polish Your Skill

Column, 1122 Feed pumps system, 1205 Product pump system, 114-J, 1305 Vacuum system, 1354 Vendor files: Pump 114-J, 1486 Commissioning data, P&ID, internal reports, 1964 Handbook, Sulfinol composition, 1900 Expected density of liquid product, 1750 Steam tables, 1650 Trouble-shooting files, 1540 Calculations and estimations (that can be done in the office before special tests are done) Rate, Use DT to estimate whether boiling in reboiler is nucleate or film, 2908 Equipment performance, insufficient data available to do calculations. .

.

What is the current operation

Visit control room: control-room data: values now and from past records. Pressure of sparge steam P/201, 2605 Temperature of sparge steam, T/201, 2230 Temperature of water in the barometric leg, 1850 Temperature of the overhead TR/100, 1950 Pressure at the top of the column PI/100, 1674 Pressure in the overhead line to the ejectors. PI/101, 847 Flowrate of feed to column PI/101, 946 Process operators, Please describe what has been done so far, 2650 Contact with on-site specialists, Utilities: steam pressure and flow to unit stable, 132 Utilities: cooling water to reflux cooler and to the barometric condenser: temperature and flow, 2419 Utilities: reflux water; temperature, flowrate and composition, 1973 Visit site, read present values, observe and sense. Is the steam tracing on?, 1376 Compare speed of pump 114-J with specs, 2475 Compare pump stroke on 114-J with specs, 2868 Pressure drop across trays 1 to 14, 2644 Pressure drop across trays 1 to 4, 2517 Pressure drop across trays 4 to 14, 2133 Flowrate of reflux water, FIC/64, 2030 Temperature at the bottom of the column, 2413 Temperature of the feed to the column, 1612 Listen to pump 114-J for sounds of cavitation, 2854 Read level glass on the sump at tray 4, 2225 Check diagram and P&ID versus what’s out on the plant, 1336

389

390

8 Prescription for Improvement: Put it all Together

On-site simple tests: Close bypass valve on 114-J, 1443 Block off pump 114-J and open bypass direct to storage. Observe liquid level on tray 4 and the “bulls eye” on the line to product storage, 761 Open suction line on 114-J; drain, 524 Reduce the flowrate of reflux water, check bottoms temperature and sample/ analyze bottoms, 2221 Install a pressure gauge on the suction of 114-J and note readings under present, usual operation and compare with past, 393 Pressure at top of the column; compare P/100 with PI/101, 282 Gather data for key calculations Pressure profile, Calculated NPSH supplied to pump 114 J, 2141 Pressure profile from tray 4 to storage, 2281 Mass balance, On sulfinol over the still, 2802 Samples and measurements Measure flowrate of water from the hot well and compare with design value, 2895 Sample water from the hot well and analyze concentration of organic, 2691 More complicated tests, Stop feed and reflux to column. Maintain vacuum conditions. Unhook suction line from pump 114-J; hook up water line to the suction line and blow water backward up the suction line and watch the level via the sight glass on tray 4. When it reaches the top of the sight glass, stop the flow and observe the level over the next 10 minutes, 785 Gamma scan around tray ’4 and in the stripping section to locate liquid level in on tray 4 and possible tray collapsed, 404 Open and inspect, Inspect foot valve on 114-J and compare valve size with the size of suction line. Correct as needed, 2117 Check for plugs, obstructions in the suction pipe from Tray 4 to pump 114-J, 2386 Check the strainer on the suction line to pump 114-J for plugs and obstructions, 1227 Open column and look at the product sump and the trays in the stripping section, 1120 .

Take “corrective” action, Replace two valves on the steam to both ejectors, 834 Replace bypass valve on 114-J, 1880 Replace faulty pressure relief valve on pump 114-J, 2177

8.2 Cases to Help you Polish Your Skill

(courtesy of W. K. Taylor, B. Eng. McMaster, 1966) [9, reactor, compressor, separator; ammonia] Ammonia is produced on two interconnected reactor loops for the production of ammonia are given in Figure 8-11, the figure in Case ’29. Feed gas consists of hydrogen and nitrogen in the proper 3:1 ratio with about 1% methane as an inert. In this ammonia synthesis reaction about 10% conversion occurs per pass through the reactor. Feed gas is compressed to 34.5 MPa abs and fed to a common header that feeds two reactor loops. Liquid product is condensed and removed from the system; gas is recycled back to the loop via the recycle stage compression. The reactor operates at 500 C. There is an internal gas-gas heat exchanger within the reactor. The DT due to exothermic reaction is about 50 C. Each compressor is a multistage reciprocating constant-speed machine rated at about 3000 kW. Bypass valve B is operated to control the Dp across the recycle stage that must not exceed 3.5 MPa. Opening valve B lowers the Dp and the flow of recycle gas to the loop. The recycle flow is about five times the flow of fresh feed. Bypass valve A is operated to trim the loop: closing this valve forces more gas over to the reactor; opening the valve causes gas to bypass the loop. Valve A is used to control the reaction temperature. If too much gas is fed to the reactor and the catalyst is inactive, the high flow might extinguish the reaction. Similarly, if the flow to the reactor is too low, the reaction will go further because of the longer reaction time; the reactor will overhead because there is not enough flow to carry away the heat of reaction. Normally valve A is open slightly during plant operation. Methane is an inert coming in with the feed. The methane concentration is kept about 15% in the loop gas to the reactors by maintaining a small purge. The pressures, levels in the separators and the temperature profile in the reactor are shown in the control room. The design provides operating flexibility. If one compressor breaks down the other machine can feed both loops thus keeping the reactors at operating temperatures while repairs are done. This avoids costly startup expense. Isolation block valves on the compressors are not shown on the figure. Furthermore, both loops are equalized in pressure thus evening out any slight variations introduced by the compressors. Case ’52: The case of the swinging loops

The problem: The plant has been in operation with everything apparently running smoothly. Production rates, however, are only 80% of design. When the rates are increased by increasing the front-end feedrate and closing the compressor kickbacks, then the whole process becomes very hard to control. Operating logs report the following “... all the gas goes to the North loop for a while and then it swings and all the gas goes to the South loop ...“ “... the reactor is overheated ... and the loops flip flop again from South to North ... or sometimes from North to South. Usually it starts with the South”. You are brought in to sort out what the operators call “The Swinging Loops”. This costs us $40 000/ day.

391

392

8 Prescription for Improvement: Put it all Together

Case ’52: The swinging loops . .

. .

. . .

MSDS, 186 Immediate action for safety and hazard elimination, Put on safe-park, 2665 Safety interlock shut down, 2578 SIS plus evacuation, 2962 More about the process, 483 IS and IS NOT: (based on given problem statement), What, 1976 When, 1583 Who, 1704 Where, 1803 Weather, Today and past, 1264 Maintenance: turnaround, When and what done?, 1080 What should be happening

Design and simulation files (allowances made for fouling, overdesign and uncertainties) Reactor, 608 Internal heat exchanger, 903 Reciprocating compressor, 153 Refrigeration system, 335 Condensers: refrigerant, 1928 Condensers: water, 2929 Gas-liquid separators, 2278 Valves A and B, 1756 Vendor files: Reciprocating compressors, 1907 Refrigeration system, 1353 Commissioning data, P&ID, internal reports, 1158 Handbook, Thermal properties of hydrogen and ammonia, 1053 Trouble-shooting files, 1467 Calculations and estimations (that can be done in the office before special tests are done) Pressure profile, Direction of leak: condensers with CW, 107 Direction of leak: condensers with refrigerant, 64 Around the loop and fresh feed into the loop, 260 Mass balance, Bleed to maintain the inerts in the recycle at 15%, 406 Thermodynamics, 489 Equipment performance, Controlling any hot spots, 217 Maximum temperature before the catalyst is damaged, 1309 .

8.2 Cases to Help you Polish Your Skill .

What is the current operation

Visit control room: control-room data: values now and from past records. Valve A in South loop, 1024 Valve A in North loop, 1074 Valve B in South loop, 1113 Valve B in North loop, 1208 Temperatures in bed in South reactor, 1414 Temperatures in bed in North reactor, 1496 Production of liquid ammonia South loop, FR, 2417 Production of liquid ammonia North loop, FR, 2258 Dp across catalyst bed South reactor, 2113 Dp across catalyst bed North reactor, 2060 Hydrogen concentration in feed gas to the reactor: South loop, 2021 Hydrogen concentration in feed gas to the reactor: North loop, 2156 Temperature at the exit of the cooling water exchanger, South loop, 2804 Temperature at the exit of the cooling water exchanger, North loop, 2607 Temperature at the exit of the refrigeration exchanger, South loop, 2778 Temperature at the exit of the refrigeration exchanger, North loop, 2958 Calrod startup heaters to both South and North loops, 2333 Visit control room: control-room data: As soon as one of the loops starts to swing record the values and note the “usual values”. Here are the data when the South loop swings taken 4 min after the swinging was first noted by the operators, Valve A in South loop, 2354 Valve A in North loop, 1852 Valve B in South loop, 1982 Valve B in North loop, 1845 Temperatures in bed in South reactor, 1602 Temperatures in bed in North reactor, 920 Production of liquid ammonia South loop, FR, 683 Production of liquid ammonia North loop, FR, 556 Dp across catalyst bed South reactor, 131 Dp across catalyst bed North reactor, 369 Hydrogen concentration in feed gas to the reactor: South loop, 289 Hydrogen concentration in feed gas to the reactor: North loop, 1824 Calrod startup heaters to both South and North loops, 1884 Process operators, What is the frequency of the swinging loops, 2318 What is the cycle time of the swinging loops, 2392 What do you do when you first notice a hot spot, 2872 Can you predict when one loop will start swinging, 2903 What do you think is the most noteworthy observation when swinging is occurring, 2623

393

394

8 Prescription for Improvement: Put it all Together

Is it always the South loop that gets the hot spot that seems to trigger the swinging loops or does it happen with equal frequency for the North loop, 1620 Operating procedures, What to do if there is a hot spot, 1657 New procedure; if reactor “hot spot” is 10 to 20 C; immediately close valve A in that loop, 1181 Contact with on-site specialists, Supervisor for construction crew for the North loop and the tie-ins: precautions that were taken to ensure that lubricating oil and or water was not left in any of the lines especially the tie-in lines, 706 Any possible preservative oil or residual valve stem lubricant left on any of the tie-in lines/valves for the recent construction, 806 Visit site, read present values, observe and sense. Change set point on valve A, South and North loops, and observe whether the valve stem moves, 988 Change set point on valve B, South and North loops and observe whether the valve stem moves, 80 Observe the “kickback” behavior of the compressor on the South loop, 206 Observe the “kickback” behavior of the compressor on the North loop, 655 Observe-listen for surge in compressors on South loop, 887 Observe-listen for surge in compressors on North loop, 800 Check diagram and the P&ID versus what’s actually out on the plant, 1216 On-site simple tests: Record evidence and actions over a three hour period whenever the swinging loop behavior “just starts”, 1384 Consistency checks, For temperature sensors in catalyst bed; do they make sense relative to each other for Loop South reactor, 1145 For temperature sensors in catalyst bed; do they make sense relative to each other for Loop North reactor, 1725 Trend checks, Cycle time: duration and frequency, 1268 Sensors: check response to change Temperatures North and South loop, 720 Ammonia production flowrate: North and South loop, 580 Level on liquid receivers: North and South loop, 1059 Temperatures on exit of condensers: both water and refrigerated condensers. North and South loop, 1437 Hydrogen analyzer in loop circuit. North and South loop, 1714 Sensors: calibrate, Eight temperature sensors North loop reactor, 2244 Eight Temperature sensors South loop reactor, 2146 Ammonia production flowrate: North and South loop, 2444 Level on liquid receivers: North and South loop, 1686

8.2 Cases to Help you Polish Your Skill

Temperatures on exit of condensers: both water and refrigerated condensers. North and South loop, 1534 Hydrogen analyzer in loop circuit. North and South loop, 1239 Kickback relief on North loop compressor, 1979 Kickback relief on South loop compressor, 1678 Call to vendors, licensee, Supplier of the catalyst: anything to watch for that might poison the catalyst, 1873 Was the same batch of catalyst shipped for the new reactor in the North loop as was used in the previously commissioned South loop system, 2375 Samples and measurements Shut down the system; sample the “reduced” catalyst from both reactors and compare the activity of each, 1295 Sample the feed to South reactor when it has the hot spot; the samples are to be every 2 minutes for the first 10 minutes of the swinging loop cycle. Analyze for inerts, 1340 Sample the feed to North reactor when it has the hot spot; the samples are to be every 2 minutes for the first 10 minutes of the swinging loop cycle. Analyze for inerts, 633 Fresh feed to South loop; analyze for inerts. Sample once per hour for 24 hours and once every two minutes for ten minutes during the swinging loop, 940 Fresh feed to North loop; analyze for inerts. Sample once per hour for 24 hours and especially during the swinging loop, 2430 More complicated tests, Isolate the North loop and run commissioning tests the same way we had done before turnaround for the South loop alone, 2756 Open and inspect, Open the reactors and note the height of catalyst in the two reactors, 1710 Valve A; South loop and look for chatter anything that might vibrate or oscillate to cause fluctuations, 1949 Valve A; North loop and look for chatter anything that might vibrate or oscillate to cause fluctuations, 2949 Valve A from both North and South loops and pressure test for leak across the butterfly valves, 350 Valve B from both North and South loops and pressure test for leak across the butterfly valves, 100 .

Take “corrective” action, Increase the bleed rate in the South loop to decrease the inert concentration in the recycle from 15% to 14%, 867 Replace valve on purge line in the South loop, 1867 Replace the butterfly valves (used for valves A and B on both North and South loops) with full size globe valves, 2865 Replace the temperature sensors in the South loop reactor, 2737 Replace kickback valve system on the South loop reactor, 763

395

396

8 Prescription for Improvement: Put it all Together

8.3

Summary

I hope you enjoyed your journey to improve your skill and confidence as a trouble shooter. I hope you had a chance to use the triad method; this provides a rich experience for you to see how others handle a situation, to spend the time to really understand a process so that you can respond quickly, as the expert system, to any questions the trouble shooter might pose. Working the cases on your own is a great experience as well. I trust that the range of cases gave you enriched insights about processes and trouble shooting. I welcome your feedback on how to improve the cases and, of course, cases and answers of cases you solved.

397

9

What Next? Here we summarize the highlights of this book so that we can define where we are now. Then we emphasize why reflection and self-assessment are key – yet time-consuming, challenging and frustrating – activities to develop confidence. Ideas are given in Section 9.3 on how to use goal setting as a neat way to develop skill beyond that discussed in this book. Sections 9.4 and 9.5 focus on how to go beyond this book for knowledge about process equipment (as given in Chapter 3) and for interesting and challenging trouble-shooting cases (as given mainly in Chapter 8).

9.1

Summary of Highlights

This book summarizes research about what is known about trouble shooting and what you can do to improve your skills. This is not a series of anecdotes of my trouble-shooting experience or how I personally solve problems. This book is flexibly designed to address your interests and needs. This book is filled with activities! .

Five key skill areas needed by trouble shooters: The five areas of skills needed in trouble shooting are 1) problem solving, 2) knowledge of process equipment, 3) process safety and properties of materials, 4) “systems thinking”, and 5) people skills. Although research has shown that knowledge of process equipment (common faults, typical symptoms) is of vital importance in determining success in trouble shooting, equal emphasis is given in this book to develop your skill in the other areas and to give you practice in using all the skills to trouble shoot. Section 1.3 gives an inventory for you to reflect on and rate your skills in these five areas. Five cases, Cases ’3–7, follow in Section 1.6 on which you can try out your skills. Where were your five strengths? What were the areas you wanted to work on? How did you rate your skill after you had worked on some of the five cases in Section 1.6?

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

398

9 What Next? .

In problem solving: the four key components of the process are: 1) the overall problem-solving process as characterized by frequent monitoring, emphasis on checking and double checking, being organized and systematic and keeping the problem in perspective. 2) the elements of data handling and critical thinking: data gathering and resolution, based on fundamentals, with valid reasoning and being complete. 3) how well we synthesize and put all the ideas together: having five to seven working hypotheses and being flexible. 4) decision-making: based on criteria, priorities and avoiding bias. This provides, in Table 2-1, target skills for us to emulate. To help quantitatively see progress in skill development an evaluation form (Worksheet 2-2) and a Trouble-Shooter’s Worksheet (Worksheets 2-1 and 2-4) are given. The example use of the Trouble-Shooter’s Worksheet was illustrated for Case ’8: “the depropanizer: the temperatures go crazy”. The target skills are described in Table 2-1. Reconsider your rating in Chapter 1 based on your better understanding of these problem-solving skills. Did the Trouble-Shooter’s Worksheet help you? If not, develop your own guide to help you to be systematic.

.

Knowledge of process equipment: symptoms–probable causes. Scattered through the literature and one’s experience are details of symptoms (and cause) from malfunctioning processes. Sometimes the information is confounded because the cause is not the root cause. Chapter 3 attempts to organize this information for ease in use with cross referencing for systems of equipment. Consistent SI units of measurement, terminology and organization are used, following the book Process Design and Engineering Practice. How does this collection of practical details match yours? How might you combine this collection with yours? How might you expand and keep it up-to-date?

.

Want feedback about trouble shooting? ... then Chapter 4 gives story time describing the adventures of five engineers as they trouble shoot problems that range in complexity. But these aren’t your usual stories. The story is interrupted. Each of the five scripts consists of about three parts with each part concluding with a few questions for you to consider. This question break was introduced to give you a chance to reflect on how you would have handled the case, and to decide what you should do next. At the end of each case an assessment is given of the problem-solving processes used by each of the trouble shooters. How was your approach similar to, different from Michelle, Pierre, Dave, Saadia or Frank? How does your rating of their approach agree with my rating of their approach? How well could you describe the approach you took? How would you rate your approach?

9.1 Summary of Highlights .

If you want to improve your confidence and skill in problem solving ... then Chapter 5 provides activities, with feedback, to develop skills and confidence in – awareness or the ability to describe your problem-solving processes; – strategies or the ability to see patterns in the process; – exploring the context using the Why? Why? Why? process; – being creative; – being skilled in performance assessment. For each, target skills were listed, some examples were provided, activities to develop the skills were described and forms of evidence were given. Cases ’9 and ’10 (the bleaching plant and to dry and not to dry) were introduced. What problem solving strategy do you use? How does it compare with the MPS 6step? Is the Why? Why? Why? approach effective for you? When are the best times to apply it? For creativity, are you willing to spend the extra time to write down a wide range of ridiculous ideas? For creativity, what triggers work best for you?

.

If you want to improve your confidence and skill in data collection and critical thinking ... then Chapter 6 provides activities, with feedback, to develop skills and confidence in – gathering information, selecting diagnostic actions, testing hypotheses/ causes. Criteria are given for selecting tests. Biases and mistakes made in gathering and interpreting data are confirmation bias, over-interpretation, under-interpretation and mis-interpretation, availability bias, premature closure, anchoring. The Jungian typology dimension P-J provides a useful indicator of one characteristic of your personal style for gathering and interpreting data. – being consistent in our use of words, and our use of knowledge about processes and process equipment. Specifically the focus was on identifying fact versus opinion, gathering accurate cause–symptom data and astutely reversing the connection to link symptom–cause. Our usage should be consistent with the rules of English, mathematics, the fundamentals of science and engineering and practical experience. – classification; the process of dividing large sets of information into meaningful parts. In making the division we should use a single basis/criteria per level, use no single entries and avoid faulty coordination or subordination. This was illustrated for the task of classifying the starting information into “symptoms” and “triggering” events. – identifying patterns. – reasoning. A systematic 9-step process is suggested and illustrated in the context of part of Michelle’s reasoning in Case ’3. Case ’11 (the lazy twin) was introduced.

399

400

9 What Next?

Did the list of actions given in Section 6.1 agree with yours? What types of actions do you prefer to select when you trouble shoot? What is your style in matching hypotheses with symptoms? What methods do you use to select tests that will correctly test hypotheses? Did the two inventories about bias provide useful insight? If not, what else might you do to identify your preference and style? How easy was it for you to separate facts from opinions? How easy was it matching hypotheses consistent with symptoms? What special techniques do you use to identify “symptoms”? What methods do you use to spot patterns in data? Did the diagramming of an argument help you? .

If you want to improve your confidence and skills in interpersonal skills ... then visit Chapter 7 that provides activities, with feedback, to develop skills and confidence in: – communicating, – listening, – applying the basic fundamental principles of interpersonal relationships, – building trust and – understanding our own uniqueness and the uniqueness of others. – Factors that affect our performance and the performance of others include: – unwillingness to admit having made a mistake, – stress, – level of frustration and lack of motivation. – preference to follow their own approach even though it may result in incorrect and even unsafe operation of the process. – preference to infer and interpret what we see or hear. A questionnaire is included to give us a chance to evaluate the environment in which we trouble shoot. Cases ’12–14 (the drop boxes, the lousy control system, the condenser that was just too big!) were introduced. Cases ’15–18 were included as exercises. Which activities were most useful to you? How strong is your tendency to infer? How does your environment rate? What was the best new idea you learned from this Chapter?

.

If you want to improve your skill and confidence in trouble shooting ... then work the cases in Chapter 8, reflect on the process you used, check on the number of target skills you displayed and set goals for improvement. You can experience the cases as triads or as an individual. If you get the chance select the triad experience! The cases have been carefully selected to provide varying levels of difficulty, different types of situations (from startup to usual operations) and different varieties of equipment. Start looking at cases at Level 3 and 4. The lower ratings start with Case ’19.

9.2 Reflection and Self-Assessment are Vital for the Development of Confidence

Which case did you enjoy the most? From which case did you learn the most? When you experienced the triad activity, which role did you prefer? What was the biggest challenge in reflecting on and evaluating the approach you used? Did you complete a Trouble-Shooter’s Worksheet when you started? If not, why not? This book was designed as a challenging and enjoyable educational experience.

9.2

Reflection and Self-Assessment are Vital for the Development of Confidence

Research has shown that the quality of the answer to a problem and the problemsolving skill and confidence improves if we: .

.

pause during the process and write down reflections of what you have done so far and where you are going next. have clear goals describing the skill, have measurable criteria about how you will know when you have the skill and have opportunities to collect evidence.

The design for self-assessment and confidence building is illustrated in Chapter 5. .

.

.

The target skills are listed. These are the proven behaviors of successful trouble shooters. Write these in terms that can be observed. Try to remove any ambiguity. Include some quantitative criterion to measure achievement of the target skill. This was done, for example in Sections 5.1.1, 5.2.1, 5.4.1. In the context of the overall process of trouble shooting, this was presented in Table 2.1. An activity is posed. The activity gives you a chance to display the behavior. In Chapter 5 the tasks were described in Sections 5.1.2, 5.2.2, with the materials to be used given in Sections 5.1.3, 5.2.3. For the overall process of trouble shooting, the overall process was illustrated first through the completed TroubleShooter’s Worksheets (for Case ’8 in Chapter 2, and the Cases ’2–4 in Appendix C) and through the examples given in Chapter 4 of Michelle, Pierre, Dave, Saadia and Frank. For your approach, two options were presented in Section 8.1: the triad and the individual activity. Cases were scattered through the various chapters with a list of options given in Section 8.1. Feedback forms and forms of evidence are created. In Chapter 5 the forms of evidence were listed in Sections 5.1.4, 5.2.4. For the overall trouble-shooting process, the forms of evidence will include: the problem statement (and the marks, underlines and notations you made directly on this); your trouble shooter’s notebook or worksheets that you used; the pauses and reflections you wrote as you worked through the case, the Trouble-Shooter’s Worksheet (and especially the hypotheses–symptoms chart), the written requests for information (if you worked in triads) or the sequence of codes for the ques-

401

402

9 What Next?

tions (if you worked as an individual). The most important form of evidence is Worksheet 2-2. This is significant because it relates directly to the target behaviors of successful trouble shooters, noted in Table 2.1. Use this to give yourself-feedback about the process you used.

9.3

Going Beyond this Book: Setting Goals for Improvement

Prepare yourself for success and use reflection and self-assessment effectively. 9.3.1

Prepare Yourself for Success .

.

.

.

Continually update your knowledge of process equipment. Research has shown that the key is a broad knowledge of process equipment. Continually draw on your experience and the experience of colleagues to enrich that experience. Praise yourself for where you are now. Too often we focus only on the negative and things we can’t do well. Throughout this book you have noted that the feedback was always five strengths and two areas to work on. At this time, write down your five strengths 1. ______________________________________________________________ 2. ______________________________________________________________ 3. ______________________________________________________________ 4. ______________________________________________________________ 5. ______________________________________________________________ Set achievable goals. These should be expressed in terms of the target skills of successful trouble shooters, described in Table 2.1. These should be consistent with the two areas you want to work on. You might identify your goals as: 1. to maintain the five strengths noted above and to shift one of the following “areas to work on” to a strength within the next year. My two areas to work on are: 1. ______________________________________________________________ 2. ______________________________________________________________ Arm yourself for success. In Section 6.1.2b I listed the stuff I had available in my office related to trouble shooting. A notebook, paper or hand-held electronic, is essential. Gloves, tape measure and string and always good additions. The first time you appear with a stethoscope will require courage! But after you have solved a bunch of tricky problems because of the stethoscope, others will want one too.

9.5 Beyond this Book: Sources of Other Cases

9.3.2

Use Reflection and Self-Assessment Effectively

Follow the principles of self-assessment, given in Sections 5.5 and 9.2. Assessment or performance review is not a dirty word! Conscientious self-assessment is the guide to growth.

9.4

Going Beyond this Book: Updating your Rules of Thumb and Symptom ‹ Cause Data for Process Equipment

Rules of thumb are generalized, usually numerical, values of “usual” practice. Although we all generate such “experience” numbers intuitively, it helps to organize and write these down. You might create files for different equipment, following, for example, the titles used in Chapter 3. As you read articles in such journals as Chemical Engineering Progress, Chemical Engineering and Hydrocarbon Processing record the rules of thumb in your paper or electronic files. In creating the files, decide on a system of units that you will consistently use. Take the time to rework information from other systems of units. Create your own set of rules of thumb for the processes and unit operations with which you work. Extend the rules of thumb to include cause ﬁ effect data. Check through the vendor files on the web or that are in the equipment files. Some sources in the literature include Bloch and Geitner (1983), and McNally Institute at http://www.mcnallyinstitute.com (2003).

9.5

Beyond this Book: Sources of Other Cases

Some computer simulation games for trouble shooting have been developed by Doig and colleagues for the SYSCHEM process (1977, 1980). Trouble-shooting cases reported in the literature tend to describe the problem, the process used to discover the fault and the corrective action taken. Although this format is not easy to use to help polish your skill, these cases broaden our perspective, provide more cause ﬁ effect data and can be converted into the format used in this book to aid in skill development. Some sources include Liberman (1985), Saletan (1994), the Riance series (1983 ff), the articles by Henry Kister that appear in Hydrocarbon Processing or at the AIChE conferences and the Marmaduke series in Power Magazine 1950 ff with some reprinted by Elonka (1979).

403

405

Literature References Preface Branan, C (1998) “Rules of Thumb for Chemical Engineers,” 2nd edition, Gulf Publishing Co. Gans, M. et al., (1983) “Plant Start-up step by step by step,” Chem. Eng. 90, 20, Oct 3 p 74. Kister, H.Z, series of articles in Hydrocarbon Processing 1979 ff. Lieberman, N.P. (1985) “Trouble-shooting process Operations” 2nd edition, PennWell Books, Tulsa OK. Saletan, D, (1994) “Creative Trouble Shooting in the Chemical Process Industries,” Chapman and Hall, London.

Chapter 1 Example articles describing a personal approach to trouble shooting in engineering: Gans, M. et al., (1983) “Plant Start-up step by step by step,” Chem. Eng. 90, 20, Oct 3 p 74. Laird, D, B. Albert, C. Steiner and D. Little (2002) “Take a Hands-on approach to refinery troubleshooting,” Chem. Eng. Prog, June, 68–73. Smith, K. (2002) “Refine your approach to process trouble shooting and optimization,” Hydrocarbon Processing, June, 63–66. Taylor, W.K. (1980) “Trouble Shooting at Canadian Industries Limited” Chemical Engineering Education, Spring, p 88–89. .

Books to help improve knowledge about safety and hazards.

Kletz, T.A. (1985) “What went Wrong?” Gulf Publishing Co, Houston, TX.

Kletz, T.A. (1983) “Hazop and Hazan–Notes on the Identification and Assessment of Hazards,” Institution of Chemical Engineers, London, UK. Woods, D.R. (1995) “Data for Process Design and Engineering Practice,” Prentice Hall. .

Example references summarizing research about trouble shooting:

Dubeau, C.E. et al. (1986) “Premature Conclusions in the Diagnosis of Iron-deficiency Anaemia: cause and effect”, Medical Decision Making, 6, 3, 169–173. Elstein, A.S, L.S. Shulman and S.A. Sprafka (1978) “Medical Problem Solving: an analysis of clinical reasoning”, Harvard University Press, Cambridge MA. Groen, G.J. and V.L. Patel (1985) “Medical Problem Solving: some questionable assumptions” Medical Education 19, 95–100. Johnson-Laird, P.N. and W.C. Wasan, eds, (1977) “Thinking: Readings in Cognitive Science”, Cambridge University Press, Cambridge, MA. Kassirer, J.P, and G.A, Gorry (1978) “Clinical Problem Solving: A Behavioral Analysis”, Ann. Int. Medicine, 89 245–255. Kern, L, and M. E. Doherty (1982) “Pseudodiagnosticity” in an Idealized Medical Problem Solving Environment”, J. Medical Education, 57 100–104. McGuire (1985) “Medical Problem Solving: A Critique of the Literature”, J. Med. Education, 60, 587–595. Nisbett, R. and L. Ross (1980) “Human Inference: strategies and shortcomings of social judgement”, Prentice Hall, Englewood Cliffs, NJ.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

406

Literature References Spencer, Joanne, (1988) “ Critique of Problem solving and trouble shooting approaches”, Department of Chemical Engineering, McMaster University, NSERC summer student report. Tversky, A, and D. Kahneman (1974) “Judgement and uncertainty: heuristics and biases”, Science, 185, 1124–1131. Voltovich, A.E. et al. (1985) “Premature conclusions in diagnostic reasoning”, J. Med. Education, 60, 302–307. Whitman, N. et al. (1986) “Problem Solving in Medical Education: Can it be Taught?”, Current Surgery, 43 453–4. Wolf, F.M, L.D. Grappen and J.E. Billi (1985) “Differential Diagnosis and the Competing-Hypothesis Heuristic”, J. AMA, 253, 19, 2858–2861.

Chapter 2 Covey, S.R. (1990) “Seven Habits of Highly Effective People,” Fireside Book, Simon Schuster. Holmes, T.H. and R.H. Rahe (1967) “The Social Readjustment Rating Scale,” J of Psychosomatic Research, Aug, 213–218. Kepner, C.H. and B.B. Tregoe (1985) “The New Rational Manager,” McGraw Hill, New York. Woods, D. R. (1994) “Problem-solving skills”, Chapter 3 in “Problem Based Learning: How to Gain the Most from PBL” Woods, Waterdown, ON, Canada. Woods, D.R. (2000) “An Evidence-based strategy for problem solving,” J. of Engineering Education, Oct, 443–459. Woods, D.R. et al, (1997) “Developing Problem-solving skills: The McMaster Problem Solving Program,” J of Engineering Education, 86, 2, 75–91 and http://www.chemeng.mcmaster.ca/innov1.htm and click on MPS. Woods, D.R. (1988) “Novice versus Expert Research suggests ideas for implementation,” J College Science Teaching, Sept, p 77–79, 66–67; Nov, p 138–141; Dec, p 193–195.

Chapter 3 Gans, M. et al., (1983) “Plant Start-up step by step by step,” Chem. Eng. 90, 20, Oct 3 p 74.

Griff, A. (1968) “Plastics Extrusion Technology”, Reinhold Book Co. Lieberman, N.P. (1985) “Trouble-shooting Process Operations” 2nd edition, PennWell Books, Tulsa OK. Rauwendall, C. (1986) “Polymer Extrusion”, Hanser Publishers, Munich. Vlachopoulos, J. et al, (2001) “The SPE Guide on extrusion technology and troubleshooting,” The Society of Plastics Engineers, Brookfield, CT. Woods, D.R. (1995) “Process Design and Engineering Practice, Prentice Hall. Francis, D, and D. Young, “Improving Work Groups: A Practical Manual for Team Building,” University Associates, San Diego, CA. (1979). Woods, D.R, “Group skills,” Chapter 5 in “Problem-based Leaning: how to gain the most from PBL,” Woods Publisher, Waterdown ON Canada distributed by McMaster University Bookstore, Hamilton, ON (1994). Fisher, K, et al. (1995) “Tips for Teams,” McGraw Hill, New York. Kirton, M.J. (1976) “Adaptors and innovators: a description and measure,” J. Applied Psychology, 61, 622–629. Keirsey, D. and M. Bates, “Please understand me: character and temperament types,” Gnosology Books, Del Mar, Ca (1984) and http://www.keirsey.com. Schutz, W.C, “FIRO: a three-dimensional theory of interpersonal behavior,” Holt Rinehart and Winston, New York, NY, 1958, with the instrument and scoring available from Whetton, D.A, and K.S. Cameron, “Developing Management Skills,” Scott Forseman, Glenview, IL,1984, p. 80. Tannen, D, “You Just Don’t Understand: Women and Men in Conversation,” Ballantine, New York, 1990. Johnson, D.W, “Reaching Out,” Prentice Hall, Englewood Cliffs, NJ, 1986.

Chapter 4 Lieberman, N.P. (1985) “Trouble-shooting process Operations” 2nd edition, PennWell Books, Tulsa OK. Kister, H.Z., Series of articles in Hydrocarbon Processing 1979 ff.

Literature References Gans, M. et al., (1983) “Plant Start-up step by step by step,” Chem. Eng. 90, 20, Oct 3 p 74.

Chapter 5 Basadur, M. (1995) “Simplex: a flight to creativity”, Creative Education Foundation, Buffalo, N. Y. Gadsby, R.E. and J.G. Livingstone (1977) “Catalysts and some incidents they have survived,” CEP Ammonia Plant Safety, 20, p 70 ff. Kimbell, R. et al. 1991 “Assessment of Performance in Design and Technology,” SEAC report, UK. Krishnaswamy, R. and N.H. Parker (1984) Chemical Engineering, April 16, p 93–98. Leifer, L, (1997) “Design team performance: metrics and the impact of technology,” in “Evaluating organizational training: models and issues,” S.M. Brown and C. Seidner, eds, Kluwer Academic publishers. Lombard, J.F. and R.A. Culberson (1972) “Defining Reformer Performance,” CEP Ammonia Plant Safety, 15, p 29–35. Woods, D.R. (1988) “Novice versus Expert Research suggests ideas for implementation,” J College Science Teaching, Sept, p 77–79, 66–67; Nov, p 138–141; Dec, p 193–195.

Prometheus Books, Del Mar CA and http://www.keirsey.com. Kirton KAI see Kirton, M. (1976) “Adaptors and Innovators: a description and measure,” J. Applied Psychology, 61, no. 5, 622–629. Kletz, T.A. (1986) “What went Wrong? Case histories of Process Plant Disasters,” Gulf Publishing Co, Houston, TX. Powers, G.J. and S.A. Lapp (1983) “Fault Tree Analysis,” Carnegie Mellon University Short Course, Pittsburgh. Woods, D.R. (1994) Chapter 5 in “Problem based learning: how to gain the most from PBL,” Woods, Waterdown, ON, Canada.

Chapter 8 Barton, J. and R. Rogers (1997) “Chemical Reaction Hazards”, 2nd edition, Gulf Publishing Co, Houston TX. Krishnaswamy, R. and N.H. Parker (1984) “Corrective Maintenance and Performance Optimization”, Chemical Engineering, April 16, p 93–98. Yokell, S. (1983) “Trouble shooting shell and tube heat exchangers”, Chem. Eng. 90 no 15, p 57–75.

Chapter 9 Sources of symptom-cause information:

Chapter 6

.

Branan, C. (1998) “Rules of Thumb for Chemical Engineers,” 2nd edition, Gulf Publishing Co. Halpern, Diane (1996) “Thought and Knowledge; an introduction to critical thinking” 3rd edition, Lawrence Erlbaum. Scriven, M. (1976) “Reasoning”, McGraw Hill. Walas, S.M (1988) “Chemical Process Equipment,” Butterworths. Woods, D.R. (2001–3) “Rules of Thumb” John Wiley and Sons, forthcoming. Woods, D.R. (1995) “Process Design and Engineering Practice,” Prentice Hall.

Bloch, H.P. and F.K. Geitner (1983) “Practical Machinery Management for Process Plants, part 2: Machinery Failure analysis and troubleshooting,” Gulf Publishing. .

Simulation:

Doig I.D. (1977) “Training of Process Plant Malfunction Analysis,” Chemeca 77, Canberra, 14–16 Sept, p 144–148 and (1980) “Trouble-shooting systems and experiences at New South Wales,” Chemical Engineering Education, Summer 1980, p 130. Sources of other cases:

Chapter 7

.

Johnson, D.W. and F.P. Johnson (1986) “Joining Together,” Prentice Hall, Englewood Cliffs, NJ. Jungian typology or MBTI: see Keirsey, D. and M. Bates (1984) “Please Understand Me”

Drew, J.W. (1983) Distillation Column Startup,” Chem. Eng. Nov 14, p 221. Elonka, S.M. (1979) “Marmaduke Surfaceblow’s Salty Technical Romances” R. Krei-

407

408

Literature References ger Publishing, New York. Collection of articles from Power Magazine. Karassik, I.J. (1981) “Centrifugal Pump Clinic,” Marcel Dekker Inc. New York. Kister, H.Z., Series of articles in Hydrocarbon Processing 1979 ff. Lieberman, N.P. (1985) Trouble-shooting process Operations, 2nd edition, PennWell Books, Tulsa, OK. Riance, X.P. (1983 ff) “Learning–the hard way” series of articles in Chemical Engineering Magazine, in 1985: Oct 3, Oct 31, In 1985: May 13, June 10; In 1986, Feb 17. In 1987, Jan19. In 1988, Feb 15 1986. Saletan, D, (1994) “Creative Trouble Shooting in the Chemical Process Industries,” Chapman and Hall, London. Shah, G.C. (1979) Trouble shooting reboiler systems,” Chem. Eng. Prog., July 53–58. Yokell, S. (1983) “Trouble shooting shell and tube heat exchangers”, Chem. Eng. 90 no 15, p 57–75. .

Other articles for rules of thumb and trouble shooting:

Eckert, J.S. (1979) “Design of Packed Distillation Columns” Section 1.7 in “Handbook of Separation Techniques for Chemical Engineers”, P.A, Schweitzer, ed, McGraw Hill, New York. p 1.221–1.240. Ellerbe, R.W. (1979) “Batch Distillation” Section 1.3 in “Handbook of Separation Techniques for Chemical Engineers”, P.A, Schweitzer, ed, McGraw Hill, New York. p. 1.164–1.167. Farminer, K.W. (1988) “Defoaming” in “Encyclopedia of Chemical Technology and Design”, J McKetta, ed, Marcel Dekker, NY. Gans, M. and Fitzgerald, F.A. (1966) “Plant Startup” Chapter 12 in “The Chemical Plant” R. Landau, ed. Reinhold Publishing Co. New York. Gardner, K.A. (1974) “Anticipation of Operating Problems in the Design of Heat Transfer Equipment” in “Heat Exchangers: Design and Theory Sourcebook” N. Afgan and E.U. Schlunder, eds, Scripta Book Co, McGraw Hill, New York. Godard, K.E. (1973) “Gas Plant Startup Problems” Hydrocarbon Process. Sept. p 151– 155.

Griffith, S. and Keister, R.G. (1970) “This Butadiene Unit Exploded” Hydrocarbon Process, 49, no. 9, p 323. Kister, H.Z. (1979) “When Tower Startup has Problems” Hydrocarbon Process Feb p 89–94. Kletz, T. (1979) “Learn from these HPI fires” Hydrocarbon Processing Jan p 243–250. Kletz, T.A. (1986) “Hazop and Hazan” 2nd edition, The Institution of Chemical Engineers, London, UK. Kletz, T.A. (1985) “What Went Wrong? Case Histories of Process Plant Disasters” Gulf Publishing Co, Houston, TX. Lapp, S.A. and Powers, G.J. (1977) “Computer Aided Synthesis of Fault-trees” IEEE Transactions on Reliability, April p 2–13. Lefevre, L.J. (1986) “Ion Exchange: Problems and Troubleshooting” Chem Eng, July 7, p 73–75. Lieberman, N.P. (1983) “Process Design for Reliable Operations” Gulf Publishing Co. Houston, TX. Lieberman, N.P. (1985) “Trouble-shooting process Operations” 2nd edition, PennWell Books, Tulsa OK . McLaren, D.B. and Upchurch, J.C. (1970) “Guide to Trouble-free Distillation” Chem. Eng. 77, No 12, June 1, p 139–152. Penny, W.R. (1970) “Guide to Trouble-free Mixers”, Chem. Eng. 77, No 12, June 1, p 171–180. Powers, G. J. and Lapp, S.A. (1983) “Fault Tree Analysis” Carnegie Mellon University Short Course, Pittsburgh. Powers, G.J. and Lapp, S.A. (1981) “A Short Course on Risk and Reliability Assessment by Fault tree Analysis” Carnegie Mellon University. Powers, G.J. and Lapp, S.A. (1976) “Computer-aided Fault Tree Synthesis” Chem. Eng. Prog. April p 89–93. Reason, J. and Mycielska, K. (1982) “Absent Minded? The Psychology of Mental Lapses and Everyday Errors” Prentice Hall, Englewood Cliffs, NJ. Reginald, S. and Gupta, J.P. (1986) “A Note on 1.2 Heat Exchanger Trouble Shooting” Int. Comm. Heat Mass Transfer, 13, p 235–243. Riance, X.P. (1983) “Learning- the Hard Way” Chem. Eng. Oct 3, p115–116.

Literature References Riance, X.P. (1987) “More Learning–the Hard Way: part 4, Chem. Eng. Jan 19, p 131– 133. Shah, G.C. (1978) “Trouble Shooting Distillation Columns” Chem. Eng. July 31 p 70– 78. Swartz, A. (1988) “Evaporator Operation: Trouble shooting” in Encyclopedia of Chemical Technology and Design”, J McKetta, ed, Marcel Dekker, NY. Talley, D.L. (1976) “Startup of a Sour Gas Plant” Hydrocarbon Process. April p 90–92. Troyan, J.E. (1960) “Trouble Shooting New Processes” Chem. Eng. Nov 14, p 223–226. Troyan, J.E. (1961) “Trouble Shooting New Equipment” Chem. Eng. March 20, p 147–150.

Troyan, J.E. (1961) “More on Trouble Shooting New Equipment: Pumps, compressors and agitators” Chem. Eng. May 1, p 91–94. Wetherhorn, D. (1970) “Guide–Trouble-free Evaporators” Chem. Eng. 77, No. 12, June 1, p 187–192.

Appendix A Chin, T.G. (1979) “Guide to distillation pressure control methods”, Hydrocarbon Processing, October, p 145–153. Goyal, O.P. (2000) “Evaluating troubleshooting skills”, Hydrocarbon Processing, October, p 100C.

409

411

Appendix A Feedback about Experience with Process Equipment (from Goyal (2000) and reproduced with permission from Hydrocarbon Processing) For each of the ten questions, select the best answer. Then score your answers and obtain a total score. Although Goyal’s published Test about Trouble Shooting has 40 questions that consider human skills, problem solving or conceptual skills and technical knowledge, here we include ten of the questions about technical knowledge. see O.P Goyal “Evaluating troubleshooting skills”, Hydrocarbon Processing, Oct., 2000, p 100-C to 100-N. 1.

Suddenly, emergency situations have occurred simultaneously on many fronts: a) Control valve on fuel to furnace is stuck open. b) Side-stream pump-around on the distillation column is showing dark color; crude leak suspected. c) Alarm sounding loud – high level in the steam drum. d) Flare pilots are extinguished. e) Safety valve leaking at the top of the fractionating tower. f) Flames impinging on tubes in the crude heater. g) Chlorine leakage in the utilities building. h) Rupture of fire/water main at offsite area. i) Steam trap blowing live steam at 0.5 kg/s. j) Vacuum constantly falling in a vacuum distillation column. k) Instrument air supply acting erratically. l) Rundown temperature of gasoline running 5 C higher than maximum allowed.

Assuming that all these cannot be simultaneously attended to, indicate the top five, most sensitive operating problems that require immediate attention by selecting any one of the following combinations (i) a, g, h, k, l (ii) a, c, f, g, h (iii) b, e, j, k, l (iv) b, d, e, i, j 2.

A two-stage reciprocating compressor is operating at 70 kPa abs inlet and 620 kPa abs outlet. If the pressure drop across the intercooler is 14 kPa, the pressure at the inlet to the second stage will be approximately: a) 340 kPa abs. b) 270 kPa abs. c) 200 kPa abs.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

412

Appendix A Feedback about Experience with Process Equipment

3.

4.

5.

6.

7.

8.

9.

For an exchanger with no phase change, given that T1 = 213 C; T2 = 57 C; t1 = 29 C and t2 = 41 C. If flowrates, area and configuration are kept constant, if T1 is reduced to 201 C and t1 is increased to 35 C; then (a) the new temperatures are T2 = 60 C; and t2 = 45.8 C; (b) insufficient data are available to make a calculation; (c) data are sufficient but I don’t know how to calculate; (d) the new temperatures are T2 = 60 C; and t2 will remain= 41 C; (e) the new temperatures are T2 will remain= 57 C; and t2 = 45.8 C. For most distillation cases, an increase in pressure means: a) decrease or b) increase in column capacity and (i) better or (ii) worse fractionation. Select (a) or (b) and (i) or (ii). If the circulation rate of a cooling tower is substantially decreased by pinching water to the condensers/coolers, the tower water-inlet temperature will: (a) go up; (b) come down; (c) remain nearly the same. The tower water-outlet temperature will (i) go up; (ii) come down; (iii) remain nearly the same. Select (a), (b) or (c) and (i), (ii) or (iii). A cylindrical furnace has been insulated with two types of refractory material of equal thickness, such that the low thermal conductivity material is on the furnace side and the higher conductivity is on the outside. (a) what has been done is correct to minimize heat loss; (b) since these are in series, it makes no difference which one is next to the furnace, c) what has been done is incorrect; the opposite sequence should have been used. A shell and tube, horizontally installed heat exchanger with two tube passes has pressure gauges (P1 and P2) installed, respectively, on the inlet pipe at the lower nozzle and on the outlet pipe from the upper nozzle of the channel. The pressure gauges are approximately 1.4 m apart. When water flows in the tubes the difference in readings of P1 inlet and P2 outlet is 41 kPa. Each nozzle takes 7 kPa pressure drop. Estimate the difference in pressure gauge reading as P1–P2 for the following cases: A) Tube side passes changed from 2 to 4. Answer a) 135 kPa; b) 82 kPa; or c) 40 kPa. B) Tube-side flowrate is doubled. Answer a) 124 kPa; b) 82 kPa; or c) 40 kPa C) Tube-side flow direction is reversed. Answer a) –14 kPa; b) 0 kPa; or c) 14 kPa. For a centrifugal pump: a) flow is directly proportional to the impeller diameter and the power to its square. b) head is directly proportional to the square of the rpm and the flow to the cube. c) flow is directly proportional to the rpm and the impeller diameter and the power to the cube of each of these. At the inlet of a centrifugal compressor, if the pressure is throttled maintaining the actual volumetric flowrate constant at the inlet, A) the net mass flowrate will a) decrease; b) increase B) the net standard volumetric flowrate will a) decrease; b) increase

Appendix A Feedback about Experience with Process Equipment

C) keeping other parameters the same, an increase in the specific heat ratio will: a) decrease; b) increase the discharge temperature significantly. 10. For steam ejectors: A) The higher the design steam pressure for an ejector, the a) higher; b) lower the steam consumption, particularly for one- or two-stage ejectors. B) Running a steam ejector at steam pressures slightly above design a) does; or b) does not increase suction capacity. C) At steam pressure > 25% above design, the ejector capacity actually starts to a) decline; or b) increase. D) At steam pressures even a few kPa below design pressure, significant reductions may be expected in a) both suction capacity and compression ratio; or b) suction capacity alone; or c) compression ratio alone. Answers and scoring: 1. (i) +1; (ii) +2; (iii) 0; (iv) –1. 2. (a) –2; (b) –1; (c) +3 [answer: 200kPa abs] 3. (a) +3; (b) –1; (c) 0; (d) –1; (e) –1 The heat lost in the hot stream must = heat gained in cold stream. 4. (a) –1; (b) +1; (i) –1; (ii) +1 Except in rare cases, increasing the pressure will increase the capacity but reduce the fractionation. 5. (a) +1; (b) –1; (c) 0; (i) 0; (ii) +1; (iii) +1. The water exiting the condensers will go up so the water temperature entering the tower will go up. The exit water temperature from the cooling tower will not go up; the temperature will likely be reduced or it may remain the same depending on other parameters. 6. (a) +2; (b) 0; (c) –1. 7. A: a) +1; b) 0; c) –1. The static head is 13.5 kPa. Calculation: 8 (41–7–7–13.5) + 14 + 13.5 = 136.5 kPa. B: a) +1; b) 0; c) –1. Calculation: 4 (41–7–7–13.5) + 4 (14) + 13.5 = 123.5 kPa. C: a) +1; b) 0; c) –1. Calculation: 13.5–7–7–(41–7–7–13.5) = –14 kPa. 8. a) 0; b) –1; c) +2. 9. A: a) +1; b) –12 When the pressure at the suction decreases, the actual volumetric flowrate is maintained resulting in a decrease in the mass flowrate and also a decrease in the standard volumetric flowrate. B: a) +1; b) –12 C: a) –1; b) +1. Temperature increases significantly. 10. Steam ejectors designed for high inlet steam pressure consume less steam compared to those designed for low inlet-pressure steam. Operating an ejector at the design pressure is very important. A) operating at a steam pressure higher than design does not increase the suction capacity. a) 0; b) + 12. B) operating at a steam pressure > 25% more than design makes the capacity decline. a) –12 ; b) +12. C) operating at a pressure slightly less than design pressure causes both suction capacity and compression ratio to suffer significantly. a) + 12; b) –12. D) a) +12; b) 0; c) 0. Minimum score: –15 Maximum score: 24

413

414

Appendix A Feedback about Experience with Process Equipment

General guidelines: > 22

Outstanding experience with process equipment; you will probably add your insight to the ideas in Chapter 3. 17–22 Very good understanding of process equipment; please add your ideas to Chapter 3. 10–17 Average understanding of process equipment; you might want to check ideas in Chapter 3 as you trouble shoot. < 10 Beginning understanding of process equipment; use ideas in Chapter 3 to help you trouble shoot.

415

Appendix B Improving “Systems Thinking” In Section B.1 are summarized some symbols commonly used on Process and Instrumentation Diagrams, P&IDs. Section B.2 gives you an opportunity to practice analyzing a P&ID.

B.1

Symbols for P&ID

SYMBOL KEY FOR PROCESS SKETCHES

XXX 100

XXX 100

CONTROL ROOM BOARD MOUNTED INSTRUMENT LOCAL BOARD MOUNTED INSTRUMENT

}

MEASURED VARIABLE: L: LEVEL P: PRESSURE A: ANALYZER F: FLOW T: TEMPERATURE

FUNCTIONS: I: INDICATOR C: CONTROLLER R: RECORDER S: SENSOR A: ALARM (H =HIGH, L=LOW)

EXAMPLES: PRC 100

FI

PRESSURE RECORDER CONTROLLER: I.D. NUMBER 100, MOUNTED IN CENTRAL CONTROL ROOM.

TAH

LOCAL FLOW INDICATOR

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

TEMPERATURE ALARM: ACTIVATED AT HIGH TEMPERATURE.

416

Appendix B Improving “Systems Thinking” VALVE DIAPHRAGM VALVE: CONTROLLED BY AIR LINE CHECK VALVE

THREE-WAY VALVE SPRING LOADED SAFETY VALVE AIR LINE ELECTRICAL SIGNAL

P

PACKED TOWER OR PROCESS VESSEL (SUCH AS A REACTOR). SIZE VARIES.

GENERIC TOWER OR PROCESS VESSEL (SUCH AS A REACTOR). SIZE VARIES.

INSTRUMENT SIGNAL

BURST DIAPHRAGM

ELECTRICAL TO PNEUMATIC SIGNAL CONVERTER

FURNACE

HEAT EXCHANGER, COUNTERCURRENT (BOX DIRECTLY AFTER INDICATES STEAM TRAP)

Figure B1

CENTRIFUGAL PUMP

SOLENOID VALVE CONTROLLED BY ELECTRICAL SIGNAL

S

I

HEAT EXCHANGER, COCURRENT (BOX DIRECTLY AFTER INDICATES STEAM TRAP)

AXIAL COMPRESSOR

B.1 Symbols for P&ID

SYMBOL KEY FOR PROCESS SKETCHES

ELECTRICAL GENERATOR

FILTER AND DRAIN CYCLONE

HEADER (STEAM OR CONDENSATE)

T M

STEAM TRAP

DAMPER OR BUTTERFLY VALVE

STEAM EJECTOR MOTOR

DRAIN TO SEWER FLARE STACK POSITIVE DISPLACEMENT PUMP

CONDENSER

FLOW METER

ORIFICE PLATE

IMPELLER

VAPORIZER OR REBOILER

Figure B2

SHELL AND TUBE HEAT EXCHANGER

FAN OR BLOWER

FAN

TURBINE USED TO PROVIDE WORK TO DRIVE COMPRESSOR

417

418

Appendix B Improving “Systems Thinking”

B.2

Systems Thinking

For the P&ID given in Figure 2-4 of the depropanizer, Section 2.4 Case ’8, systematically work your way through the following questions. For more see http:// www.chemeng.mcmaster.ca/courses/che4n4/Process operability/ 1.

What’s going on? For column C-8; how can you tell?

2.

What’s unique? draw simplified sketch; compare C-8 with C-9.

.

So what?

3.

Setting temp and pressure: what is the pressure at top of C-8? How can you tell?

We try to operate columns at atmospheric pressure. . . .

Why use this pressure? Cox chart. What does this tell us about the species? Cox chart Pressure at the bottom of the column. Bigger than top? smaller than top? same as top? Why?

Estimate: . .

Why is this a good pressure for the bottom? What’s the species at the bottom?

Quick check on the debutanizer. 4.

Controlling pressure and temperature

Pressure: For deprop C-8; how control? What if the sequence of the control valve and sensor is reversed? Control (see in control room?) Alarm SIS: fail open? Relief For debutanizer C-9, how control? .

What if it is a vacuum?

Use vacuum for those species with a high decomposition index: foodstuffs and “higher boiling pharmaceuticals” Temperature: For Deprop C-8, how control: top? bottom? . . . .

Why sensor location? What signals: electrical or pneumatic? how can you tell? At the bottom, what type of reboiler? Is it an equilibrium stage?

B.2 Systems Thinking . . . .

.

Is it nucleate boiling? Is steam superheated? H-S diagram. Estimate the flowrate of steam What type of steam trap? Does it make any difference which type of trap is used? Compare with Debut C-9

Level:

.

Is the level controlled in bottoms? in overhead drum V-30? in feed drum?

5.

Checking the fluid mechanics

. .

Why do fluids move? Consider Bernoulli’s D2 Dp L 2 þ þ Dz ¼ power needed 4f D 2 2 and the microscopic equation of motion: Dv ¼ rp Dt inertia pressure

2

þ lr v friction

þ g body forces

to see the fundamentals of why fluids move. Fluids move because of a pressure difference; or because they are dragged along by something moving (eg a steam ejector), because of momentum and because of body forces acting on the mass. Pressures occur because 1. pump or blower, 2. suck or pull vacuum, 3. heat up liquid in closed vessel 4. density difference on two ends of a tube. Pressure Profiles: Estimate: Tabulate Dp across the following equipment: Packed beds: Fittings: Trays and packing: Turbulent flow of gas flow and liquid: Major Dp across control valves: Create a summary page of rules of thumb Dp. .

.

.

Find the thermosyphon reboiler on the bottom of tower C-8. Track the pressure. Use 100 kPa approach by arbitrarily setting the pressure = 100 kPa at one location. Track the pressure as you go from the bottom of C-8 up the column and around to the vessel V-30. Consider the reflux drum V-30 and the flow of reflux back into column C-8. How could we make the reflux flow by gravity? Self-standing column. skirt and configuration.

419

420

Appendix B Improving “Systems Thinking”

Vacuum: Steam ejectors; work on principle of constant S and thus have available the enthalpy difference per kg of steam but 15% efficiency. Pumping liquids: express the pressure losses in terms of “head”; the head loss diagram is independent of the liquid. . .

Do a mass balance around column C-8 on PID-2A. What is the reflux ratio?

Pump F-26 Estimate the pressure drop from pump F-26 into depropanizer C-8. For a centrifugal pump 1. 2. 3. 4.

5. 6.

.

.

head: independent of fluid if it is liquid. efficiency: where it is located. Sometimes shown as “concentric rings” on the diagram. power and how we can estimate the efficiency from the head capacity and the power: 60%. can shut the exit valve off completely and the pump is still OK and gives the max. head. This is a handy way to help you identify which pump is installed in a circuit if the ID plate is removed or if you are troubleshooting and suspect it is not delivering what it should. Note that the pressure gauge must be installed upstream of the valve. pumps usually run at 1800 rpm or 3600 rpm. At the former, the heads are usually about 10 m; at the latter about 40 m. Key idea, the pump operates along the operating curve! You change the Dp in the system and the pump moves to a new operating location on the operating curve. Select, from a manufacturer’s catalog, pump F-26, on the feed to the depropanizer. What changes if we use this pump, F-26, from the Deprop circuit to operate with a) benzene. b) water and c) with mercury.

Net positive suction head required: “head” needed to keep liquid from boiling in the eye of the impeller. Data must be supplied from manufacturer (for water at 20 C). Must supply “more than this”. Options for control: 1. 2. 3. 4.

large suction line, negligible stuff to cause Dp on the suction line, have a large height for static head, use vortex breakers and submersion.

.

Spot places where we could have trouble on this plant from NPSH.

B.2 Systems Thinking

Control configurations .

.

. .

Note the bypass loop around the control valve so that we can operate temporarily on bypass. Note change in diameter of the control valve compared with the diameter of the line in and out. Why? Note drain valve so that we can take out the valve safely.

421

423

Appendix C Feedback on the Cases in Chapters 1, 2 and 7 Here are some example Trouble-Shooter’s Worksheets for the Cases posed in Chapter 1. Trouble-Shooter’s Worksheet

copyright 2003 Donald R. Woods and

Thomas E. Marlin Case ’3: The case of the cycling column 1. Engage: Write down what is said; what you sense, smell, hear. If someone is telling you, then use skilled reflective statements to ensure you accurately obtain the information. .

.

.

Emergency priority: Safety? Hazard? Equipment damage? shut down &; safe-park &. If not, & then: Draw a sketch of the process and mark on values. Provide a description in words of what is going on. This is a distillation column. The bottoms is heated via a thermosyphon reboiler. The liquid handled is a hydrocarbon, iC4. Liquid from the bottom of the column fills the tubes in the vertical shell and tube reboiler. Steam is applied to the shell side. Some of the hydrocarbon in the tubes boils. The resulting density difference causes fresh bottoms liquid to enter the tubes and the evaporation continues. Usually 5 to 25% is evaporated per pass. This is rarely used in vacuum or high-pressure service. The steam comes off the top of the header, as it should. The condensate from the bucket trap goes into the top of the condensate header, as it should. The flow of steam is controlled by the temperature. Symptoms: a) The level in the sight glass, hopefully representing the level in the bottom of the column, rises slowly about two feet above the normal operating level and then quickly drops to about two feet below normal. The operators say “the column cycles madly”. b) this is just after annual maintenance.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

424

Appendix C Feedback on the Cases in Chapters 1, 2 and 7 .

.

Manage any panic you might feel by saying “I want to and I can. I have a strategy that works. Let’s systematically follow it.” Yes! I can do my best! Monitor: Have you finished this stage? Can you check? What next? I have been cautious. I restated the words from the statement.

2. Define the stated problem: Systematically classify the given information using IS and IS NOT. If the information is not known at the stage check ?& to remind you to gather this information IS

IS NOT

(should be happening but it’s not) Level should be “level” Performance is not controlled by level.

WHEN WHERE

Level in the bottoms is cycling with a four-foot variation in level. Performance is controlled by temperature. ?& just after turnaround maintenance ?& bottoms of column

WHO

?& current shift operators

WHAT

?& didn’t cycle before turnaround ?& no cycling reported in feed, feed concentration. ?& this is first shift to start this plant

– startup new process & suggest use Basics – startup after maintenance or change & suggest use Change – usual operation but changes made in operation but not in equipment & suggest use Basics – usual operation & suggest use Basics. .

Monitor: Have you finished this stage? Can you check? What next? I have three that need to be answered. This sounds like a change problem.

3. Explore: Gather information to be gathered ?& in Define stage &. Done and recorded in chart. Exercise? & or a problem? & . Strategy: change & or basics &. Perspectives. Why? Why? Why?

Why? Why? Why? Why? Why?

__________________________________________________________ › __________________________________________________________ › 4. so product can be sold › 3. get on-spec tops and bottoms › 2. level out performance of column ›

Start ﬁ 1. stop great variation in level and in steam flow

Appendix C Feedback on the Cases in Chapters 1, 2 and 7

Main goal is to “get on-spec tops and bottoms” . .

Prioritize: product quality & ; production rate &; profit &. Goal: safe-park? & : short term now with long term later &; long term now &

Action to be achieved: Specific terms and Measurable get on-spec tops and bottoms Attainable? could put it on manual control or on safe-park Reliable? Timely? Safe? $ .

.

.

Check consistency of data/symptoms: interdata consistency? OK & no & data consistent with fundamentals? OK & no & Type of problem: startup new process & maybe mechanical/electrical failure usual operation &: ambient temp? & maybe fluids problems; high temperature? & then maybe materials problems System? failure of heat exchanger & > rotating equipment & > vessels & > towers & Identify key and What if?

What if? _______________________ then _____________________________ What if? _______________________ then _____________________________ What if? _______________________ then _____________________________ .

List changes made & and/or List trouble-shooting causes based on symptom & . List both.

Changes made: During the shutdown 1) The condensate from this inverted bucket trap, that previously discharged to atmosphere, was repiped to discharge into a 200-kPa condensate header. The condensate discharged into the top of the header. Steam to the reboiler was “saturated” at 1.7 MPa. 2) For this unit the only other maintenance was instrument checks. No faults were found in the instruments and no changes were made to settings in controllers. 3) And visual inspection of the trays. The visual inspection involved opening the access holes. Trouble shooting background. From Chapter 3 and from Michelle’s account in Chapter 4. Not reproduced here.

425

426

Appendix C Feedback on the Cases in Chapters 1, 2 and 7 .

Brainstorm root causes: collapsed tray, change in downstream pressure affecting bucket trap, instruments wrong; liquid level is not cycling, control system, restriction in the vapor line, cycling change in concentration of high boilers in the feed, temperature sensor in the feed zone.

In prioritizing these, focus on the change: change the exit pressure of the bucket trap. Hypotheses: list in Chart; Symptoms: code and list in chart; Analyze with S supports; D disproves and N neutral or can’t tell. Symptom a. liquid level increases by 2 ft and then decrease by 2 ft b. occurs after shutdown maintenance c. d. e. Working Hypotheses

1. change in discharge pressure to bucket trap 2. instrument reading wrong 3. cycling concentration of heavies in feed 4. control system 5. restriction in vapor line 6. temperature sensor in feed zone 7.

Initial Evidence a S S S S S

b S

c

e

A

N 4 S N

Diagnostic actions: A. B. C. D.

d

Diagnostic Actions

control system on manual open trap bypass discharge condensate to atmosphere check out sizing calculations for orifice in trap

B 4

C 4

D 4

Appendix C Feedback on the Cases in Chapters 1, 2 and 7

Trouble-Shooter’s Worksheet

copyright 2003 Donald R. Woods and

Thomas E. Marlin Case ’4: The platformer fires 1. Engage: Write down what is said; what you sense, smell, hear. If someone is telling you, then use skilled reflective statements to ensure you accurately obtain the information. .

.

Emergency priority: Safety? Hazard? Equipment damage? shut down &; safe-park &. If not & , then: Draw a sketch of the process and mark on values. Provide a description in words of what is going on. Liquid naphtha is preheated in a shell and tube exchanger and enters the packed bed, reactor where the naphtha is converted to platformate. The products from the reactor are high octane gasoline plus hydrogen-rich gas at 4.8 MPa g and 500 C. These hot gases are cooled in the feed preheater. The preheater shell is 1 m diameter. Stainless steel.

Three weeks since startup. Four flash fires along the flanges. Maintenance has tried unsuccessfully to tighten the flange to prevent leaks once the reactor heats up. .

.

Manage any panic you might feel by saying “I want to and I can. I have a strategy that works. Let’s systematically follow it.” This one sounds challenging but I’ll do my best. Monitor: Have you finished this stage? Can you check? What next? Yes.

2. Define the stated problem: Systematically classify the given information using IS and IS NOT. If the information is not known at the stage check ?& to remind you to gather this information

WHAT

WHEN WHERE WHO

IS

IS NOT

Flash fires along flanges. Six bolts broken trying to tighten the flange ?& During three weeks since startup

(should be happening but it’s not) No flash fires. A flange that is easy to seal mechanically. ?& not operated before so have no information ?& in the platformer reactor or elsewhere ?&

?& Along the flanges of the preheatereffluent cooler ?& no evidence that it occurs only on one shift

– startup new process & suggest use Basics – startup after maintenance or change & suggest use Change

427

428

Appendix C Feedback on the Cases in Chapters 1, 2 and 7

– usual operation but changes made in operation but not in equipment & suggest use Basics – usual operation & suggest use Basics. .

Monitor: Have you finished this stage? Can you check? What next? Think I’ve finished

3. Explore: Gather information to be gathered ?& in Define stage &. Exercise? & or a problem? & . Never seen anything like this before. Strategy: change & or basics & Perspectives. Why? Why? Why?

Why? Why? Why? Why? Why?

6. so that I can be paid › 5. so we can sell platformate and make profit › 4. so the whole process can operate › 3. make it safe; prevent flames to other parts of the process › 2. prevent fires on the effluent exchanger ›

Start ﬁ 1. prevent hydrogen-rich gas from leaking out the flange . .

Prioritize: product quality &; production rate &; profit &; safety & Goal: safe-park?&: short term now with long term later & ; long term now &

Action to be achieved: Specific terms and Measurable: level 2: prevent fires Attainable? Fires are hydrogen + spark + oxygen Reliable? Timely? Safe? $ .

.

Check consistency of data/symptoms: interdata consistency? OK & no &. Fires are hydrogen + spark + oxygen data consistent with fundamentals? OK & no &. Hydrogen diffuses very easily through small crevices. These are very high temperatures and pressures. The thermal expansion during startup will be extensive. Type of problem: startup new process & ; maybe mechanical electrical failure

Appendix C Feedback on the Cases in Chapters 1, 2 and 7

usual operation &: ambient temp? & maybe fluids problems; high temperature? & then maybe materials problems System? failure of heat exchanger & > rotating equipment & > vessels & > towers & .

Identify key and What if?

What if?

What if? What if? .

heated the outside of the exchanger then negligible differential thermal expansion Put exchanger in a furnace? let the hydrogen leak then focus on preventing oxygen from mixing with hydrogen _________________________ then ________________________

List changes made & and/or trouble-shooting causes based on symptom & fire: combustible like hydrogen + oxygen/air + spark/

Identification of cause is not the issue; prevention is .

Brainstorm root causes: possible ways to keep oxygen from mixing with the hydrogen. nitrogen, steam, not air ideas to contain the hydrogen once it leaks: a blanket or outside shell that contains the hydrogen and from which to extract it elsewhere; packed hydrogen adsorbent; spray liquid over exchanger that rapidly adsorbs hydrogen; react hydrogen.

.

Hypotheses: list in Chart; Symptoms: code and list in chart; Analyze with S supports; D disproves and N neutral or can’t tell.

Symptom a. four fires along flanges b. break six bolts trying to tighten c. during first three weeks d. e. Working Hypotheses

1. hydrogen leaks out of the flanges 2. cannot tighten the flanges any tighter to prevent the leak

Initial Evidence a S S

b N S

c N

d

Diagnostic Actions e

A

B

C

D

429

430

Appendix C Feedback on the Cases in Chapters 1, 2 and 7

Trouble-Shooter’s Worksheet

copyright 2003 Donald R. Woods and

Thomas E. Marlin Sulfuric acid pump problem 1. Engage: Write down what is said; what you sense, smell, hear. If someone is telling you, then use skilled reflective statements to ensure you accurately obtain the information.

Case ’5:

.

.

.

.

Emergency priority: Safety? Hazard? Equipment damage? shut down &; safe park &. If not &, then: Draw a sketch of the process and mark on values. Provide a description in words of what is going on. Acid comes from different parts of the plant and is collected in a common storage tank. A pump, located above the tank, pumps the acid to an elevated tank. Vertical dimensions are given. Manage any panic you might feel by saying “I want to and I can. I have a strategy that works. Let’s systematically follow it.” Monitor: Have you finished this stage? Can you check? What next?

2. Define the stated problem: Systematically classify the given information using IS and IS NOT. If the information is not known at the stage check ?& to remind you to gather this information IS

IS NOT

WHAT

When the level of acid in the storage tank drops to 0.7 m in the bottom of the storage tank, the operator says the pump makes a “crackling” noise that sounds like cavitation.

(should be happening but it’s not) Pump cavitation should not happen

WHEN

?& When the level of acid in the storage tank drops to 0.7 m in the bottom of the storage tank. ?& In the acid pump ?&

?& When the level of acid in the storage tank is above 0.7 m in the bottom of the storage tank. ?& In other pumps ?&

WHERE WHO

– startup new process & suggest use Basics – startup after maintenance or change & suggest use Change – usual operation but changes made in operation but not in equipment & suggest use Basics – usual operation & suggest use Basics. .

Monitor: Have you finished this stage? Can you check? What next?

Appendix C Feedback on the Cases in Chapters 1, 2 and 7

3. Explore: Gather information to be gathered ? & in Define stage &. Done. Exercise? & or a problem? &. Sounds like a cavitation problem and NPSH supplied & NPSH required Strategy: change & or basics & . Perspectives. Why? Why? Why?

Why? Why? Why? Why? Why?

6. Personal Happiness and Bliss › 5. pay my salary plus bonus › 4. improve profits › 3. keep costs down › 2. minimize erosion and allow longer pump cycle ›

Start ﬁ 1. stop pump cavitation . .

Prioritize: product quality & ; production rate & ; profit & Goal: safe-park?&: short term now with long term later & ; long term now &

Action to be achieved: Specific terms and Measurable: stop crackling noise when level drops to 0.7 m. Attainable? probably Reliable? temporary should work Timely? Safe? $ .

Check consistency of data/symptoms: inter data consistency? OK & no &. Onset of “crackling noise” should coincide with reduced pumping flowrate. data consistent with fundamentals? OK & no &. “Crackling noise” consistent with cavitation consistent with insufficient NPSH that is consistent with excessive suction lift that is consistent with level in tank too low.

.

Type of problem: startup new process & maybe mechanical electrical failure usual operation & : ambient temp? & maybe fluids problems; high temperature? & then maybe materials problems System? failure of heat exchanger &> rotating equipment & > vessels & > towers &

431

432

Appendix C Feedback on the Cases in Chapters 1, 2 and 7 .

Identify key and What if?

What if? What if? What if? .

acid more dense then more likely to cavitate earlier or > 0.7 m atmospheric pressure drops then more likely to cavitate earlier or > 0.7 m vent plugged then non flow

List changes made & and/or trouble-shooting causes based on symptom &

See Chapter 3 or list Dave generated. .

Brainstorm root causes: liquid too hot, non-condensibles, air leakage into suction, vortex entraining gas, density change in liquid, excessive Dp on suction side, suction velocity too high, increased rpm, increased liquid capacity.

The entries in italics are symptoms and not a root causes. Many of these do not relate to suction-lift situations. .

Hypotheses: list in Chart; Symptoms: code and list in chart; Analyze with S supports; D disproves and N neutral or can’t tell.

Symptom a. crackling noise related to cavitation b. c. d. e. Working Hypotheses

1. NPSH supplied is less because atmospheric pressure low 2. NPSH supplied is less because of entrained gas because of vortex 3. NPSH required is higher than expected a) because higher capacity because of less system resistance 4. NPSH supplied less because liquid density too high 5. 6. 7.

Initial Evidence a S S

b

c

d

Diagnostic Actions e

A 4

B

C

4

S

S

4

D

Appendix C Feedback on the Cases in Chapters 1, 2 and 7

Diagnostic actions: A. increase pressure on top of storage tank B. reduce liquid pumping capacity by shutting down on discharge valve (Will reduce the NPSH required and should reduce tendency to form vortex) C. sample acid and measure density D.

Cases ’15 to 18 are a few of the cases that are related directly to human error and issues raised in Chapter 7. There may be other problems given in Chapter 8 where the cause is human error. In this Appendix we simply provide the cause of the problem for Cases ’15 to 18. Case ’15: The flooded boot (from Bill Taylor) This is a case of initial construction error. Workers left a block of wood in the 15-cm diameter (6†) line between the ion exchanger and the pump. Material left in lines is common. For example, a line rigger temporarily places bolts in a pipe while taking a break. A colleague arrives back from break, continues assembling the line, although he cannot find the bolts so gets new bolts. The plant starts up and the bolts are carried into the blades of a downstream centrifugal compressor. Six months delay occurs in getting the repaired compressor on-line! Sandwiches, rags, wrenches and stuff ... don’t be surprised by what might be left in lines. Of course, after commissioning runs with water and air, expect pockets of these to remain in low places and unvented locations. Case ’16: The case of the dirty vacuum gas oil (From Bob Farrell) The problem is that crude oil is leaking into the vacuum oil. In an effort in increase capacity beyond the usual, the production manager increased the pressure in the crude line such that it exceeded the rating of the exchangers. A leak resulted. This is a case of operator/supervisor error Case ’17: Is it hot or is it not? (case from Jonathan Yip) This is a case of an operator ignoring the readings of the instruments. Within a short time after the operator had manually suppressed the purge valve, thinking it was a false temperature indication, a field operator felt something was wrong when he heard unusual cracking noise from the column. Looking at the DCS screen again, he realized that a section of the column had temperatures above the range. A fire had broken out in the column. He immediately activated the nitrogen purge and started to cool down the column with water spray. Investigations afterwards showed that the pump isolation valve was leaking due to a worn seal. This allowed air to migrate into the column. Since the fatty acids were above their autoignition temperature, the acids ignited in the presence of air. A

433

434

Appendix C Feedback on the Cases in Chapters 1, 2 and 7

further survey of the data showed that the column pressure had been gradually increasing or losing its vacuum over the same period suggesting a air leak. The internal damage to the column from the fire was extensive. The entire column had to be replaced. Case ’18: The streptomycin dilemma

This is a case of an operator ignoring operational procedure and substituting his own to make life easier. The expected procedure was: 1. brush out the tubes; 2. rinse. 3. fill the tubes with water and 4. add steam to the shell and boil the water. What the operators did was 1. brush out the tubes; 2. rinse and then 3. use a steam hose, with a long pipe attached to the end of it, to blow live steam into the tubes individually. They reasoned that the live steam “would do a better job of sterilizing” than the old slower method. They did not understand the impact of the high temperatures on the thermal expansion of the tubes. When asked if they followed the expected procedures, all operators said yes. None admitted to using live steam. This was only discovered when engineers unexpectedly visited the site when the tubes were being “sterilized”. In this case, the environment did not allow for admission of error; the rationale for the sterilization procedure had not been given. The consequences of using live steam had not been spelled out.

435

Appendix D Coded Answers for the Questions Posed to Solve the Cases Consult the code corresponding to the question or activity you want. Keep a record of the sequence in which you pose each question. You can later compare the sequence of questions you asked with those of an experienced trouble shooter. Suggestion, work question by question instead of collecting four or five questions and obtaining answers to them all at once. 1 2 3 4 5

6 7 8 9 10 11 12

13 14

15 16

Not needed. $650 This is the commissioning startup of a new unit. $50 Responds to change. $200 No. $30 Hot 30 C and humid: 80% relative humidity. This week started out with moderate temperatures but the temperature and humidity have climbed to today’s high. Not needed. $50 Estimate seems OK. No major blockage. $100 Not needed. $150 Water is used in the seal pot in the summer; kerosene is to be used in the winter. Kerosene is obtained from the crude unit. $500 Should work well. Baffles were in vertical. Vessel has air bleeds. Before you started up you would have opened the bleed to ensure all the air was out. $200 Responds to change. $200 Usual cycle is to power the heaters, add feed to hopper, start mixer, mix 35 s, stop rpm and screw advances to push melt into the mold, screw reverses, feed drops from hopper into mixer and so on. The resin is predried for 3 hours by direct contact with hot air at 93 C. $100 Fully open. $300 IS: on this reactor. IS NOT: elsewhere in the plant. $50. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 Perhaps shut down, depending on the length of time of operation since the last attempted flow increase. Allowance of 140 kPa across flow control valve FV-1 NPSH required = 1.8 m water. $150

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

436

Appendix D Coded Answers for the Questions Posed to Solve the Cases

17 18 19 20 21 22 23 24 25 26

27 28 29 30 31 32 33 34

37 38 39 40 41 42 43 44 45 46

47 48

IS: This shift. $3000 Not apparent. $2000 Not for 8 months. $50 Should do the job. $200 Usual design with direct water contact. Use river water. $120 No change. Flowrate and temperature should be the usual values. The temperature should be about 20 C. $200. Cold –1 C; cloudy, snow predicted. Previous week, cold with some snow. $20 Looked fine. $200 Equipment should do the job. Check on design calculations show liberal allowance for the pressure drop across the control valve so that it operates mid-range. Pump should have no trouble supplying the design reflux. $100 IS: Just after change to new catalyst. IS NOT: problem with old catalyst. $150 OK, recalibration not necessary. $700 Clear, not clogged. All clearances are OK. $2000 Raining. Yesterday: hot, dry, 30 C. Showers three days ago. Cleared up and has been hot ever since until today’s rain. $10 Clean. It just had new bags installed. Reverse jets had been serviced. Not needed. $30 000 No problems. $100 Ethylene NFPA: 1, 4, 2; n-butane: 1, 4, 0. Butane skin-contact with gas may cause frostbite (liquefied gas). May cause slight eye irritation, inhalation causes drowsiness, excitation or unconsciousness due to anesthetic and asphyxiation properties of this gas. $50 Typically rule-of-thumb value is 5 kPa; instruments read 50 kPa. Samples show a decrease in number of particles with time. No data available for previous operation. $2100 Yes; everything seems to be OK on our unit. $200 OK. Recalibration not necessary. 3 days ago completed the annual work on the baghouse: new bags installed. Reverse jets had been serviced. Not needed. $1000 Not needed Fluctuates. $200 Selected for flow and pressure drop around the hydrocyclone. $180 Well-designed for the design capacity. Since acetic acid vapor partially dimerizes (so that the molar mass varies between 60 and 120) a molar mass of 102 was used for all calculations involving acetic acid vapor. Heat transfer coefficient 20 to 40 W/m2 K. Triethyl phosphate added as catalyst about 0.3% w/w in the acetic acid feed. About 400 ppm for each 24 h composite for each of the seven days. $70 Not needed. $3000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

49

50 51 52 53 54

55 56 57 58 59 60 61 62 63 64 65 66 67 69 70

71 73 74 75 76

Responds to change. $300. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. Responds to change. $2000 Hot liquid fatty acids will oxidize and cause spontaneous combustion. $200 Crystals and liquids do not pose a hazard. $30 Responds to change. Crystals in compact layer instead of being movable crystals. Predict that this would cause a decrease in filtration rate compared with design and higher exit moisture content in crystals peeled from the centrifuge. $1400 Ethylene 99.99%. no butane. $5000 Your unit receives the first cut because this is so critical for you. Flowrate is steady. Temperatures are steady and the usual values for this time of year. Everything OK. $6500 Not needed. $20 000 Well designed; should do the job. $2000 From the package: strengths of cast samples well below specs. Particle-size distribution missing the very fine particles < 10 mm. None available. $50 Amps as expected except when low flowrates of acid are required. $1000 Should be OK even with increase in column pressure by 100 kPa; allowed for in the Dp across control valve FV-1. $100 Process stream into the refrigerant. $300 Reads as expected for the naphtha flowrate giving the same space velocity for the new conditions as the old conditions. $150 Estimate seems OK. No major blockage. $1200 Not needed. Level above level of tubes; design level. $50 The diagram is a reasonable schematic. Missing are the isolation block valves for the flare-gas compressor; the isolation block valves, drain and bypass with valve around the kickback control valve. The pressure gauges have pigtails with a shutoff valve with a tee nipple and valve installed between the gauge and the pigtail to allow backflushing of pigtail. The knockout pot has a demister and a level indicator; the flare pot has isolation block valves. The flare is an elevated flare with a molecular seal and an elevated flare burner. There is a flow meter on the gas to the flare that sends a signal to a ratio controller controlling the steam flow to the flare. Liquid from the knockout pot is pumped to oil recovery or slops. A gas purge line goes to the molecular seal. Moisture content was within the design range although slightly higher than before the change to the washing cycle was made. $60 Continuous. No break or short circuit in heater. Heater should work. $23 000 Not needed. $50 000 Design specs. seem OK. Centrifugal compressor; designed to operate without surge based on typical gas composition. $500 Responds to change.

437

438

Appendix D Coded Answers for the Questions Posed to Solve the Cases

77 78 79 80 82 84 85

86 87

88 89 90

91 92 93 94 95 96 97

98 99 100 101

102 103 104

On spec. No complaints from other customers for that batch. $350 DPI decreasing. $50 No change; still cycles. $1200 Valve stem responds to change. $600 Closes within –18%. $600 Not needed. $150 See Chapter 3: Distillation Section 3.4.2, perhaps pertinent condensers and reboilers, Section 3.3.3; pumps, Section 3.2.3 and controllers Section 3.1.1, $50 No change in feed sent to your unit $1200 Allowance for fouling on tube side, water = 0.0002 m2 K/W; shell-side stream= 0.0005. Care to prevent temperature cross-over. Simulation shows that unit was operating very close to design values for usual range of feedstocks and conditions. $400 Not needed., $2000 Vent bleed installed to break any syphon. The elevation of the header is such that the tubes at all levels should be flooded. $150 Instrument error in flowmeter/ motor fails to start when remote start button is pushed/ pump A has worn wear rings causing internal flow circulation/ on pump A the motor is turning backwards so that the impeller is turning in the wrong direction/ air lock in pump/ debris or stuff plugging the suction line of pump A/ reverse flow through check valve on pump B/ faulty block valves. $40 No change in operation. Pressures and temperatures still increasing. $250 25 C. $150 Same as usual concentration. $2500 No improvement; indeed, when level drops below complete coverage of the bundle, T/6 reads 15 C and rising. $2000 Product cracked coming out of the mold. $900 Agrees with head capacity. $200 Sensor fault/ poorly tuned control system/ sticky control valve/ control-valve hysteresis/ caustic waste flows backwards into the pump/ electrical interference with the control system/ cavitating pump/no vent on the storage tank; vacuum created in the storage tank/ density of acid > expected and motor overload. $100 Not needed. $50 000 Steady. $120 Both valves leak when fully closed. % leakage varies from valve to valve and ranges from 1 to 3%. $40 000 Previous flows of air and fuel gave 10% excess air for stoichiometric combustion. If didn’t put on safe-park as first step, then dangerous potential fire/ explosion conditions are created while you experiment. $500 000 Flow is the control value of 14 L/s. $50 Steady and usual value. $50 Difficult because the steam flowrate is not recorded. $50

Appendix D Coded Answers for the Questions Posed to Solve the Cases

105 106 107 108 110

111 112 113

115 116

Responds to change. $200 Well within design specs even though this is a hot and humid day. $3200 Process stream into the cooling water. $300 Yes, within –15%. $1000 Sodium hydroxide: Health Rating: 3 – Severe (Poison) Flammability Rating: 0 – None. Reactivity Rating: 2 – Moderate. Contact Rating: 4 – Extreme (Corrosive); Poison! Danger! Corrosive. May be fatal if swallowed. Harmful if inhaled. Causes burns to any area of contact. Reacts with water, acids and other materials. Lab Protective Inhalation: Severe irritant. Effects from inhalation of dust or mist vary from mild irritation to serious damage of the upper respiratory tract, depending on severity of exposure. Symptoms may include sneezing, sore throat or runny nose. Severe pneumonitis may occur. Ingestion: Corrosive! Swallowing may cause severe burns of mouth, throat, and stomach. Severe scarring of tissue and death may result. Symptoms may include bleeding, vomiting, diarrhea, fall in blood pressure. Damage may appear days after exposure. Skin Contact: Corrosive! Contact with skin can cause irritation or severe burns and scarring with greater exposures. Eye Contact: Corrosive! Causes irritation of eyes, and with greater exposures it can cause burns that may result in permanent impairment of vision, even blindness. Chronic Exposure: Prolonged contact with dilute solutions or dust has a destructive effect upon tissue. Aggravation of Pre-existing Conditions: Persons with pre-existing skin disorders or eye problems or impaired respiratory function may be more susceptible to the effects of the substance. Airborne Exposure Limits: – OSHA Permissible Exposure Limit (PEL): 2 mg/m3 Ceiling – ACGIH Threshold Limit Value (TLV): 2 mg/m3 Ceiling. Chlorine: NFPA: 4,0,1. Corrosive and poisonous gas. Contact may cause severe irritation or corrosive burns to the eyes, skin and mucous membranes. Inhalation may result in chemical pneumonitis, pulmonary edema and respiratory collapse. Nonflammable. Oxidizer. May react violently with reducing agents. Can accelerate combustion. Store below 50 C. OSHA-PEL 1 ppm; TLV-ACGIH 0.5 ppm TWA; 1 ppm STEL Sodium hypochlorite: Contact with acids releases poisonous gas (chlorine). Light sensitive. Incompatible with strong acids, amines, ammonia, ammonium salts, reducing agents, metals, aziridine, methanol, formic acid, phenylacetonitrile. Corrosive, causes burns to skin and eyes. Harmful by ingestion, inhalation and through skin contact. Skin irritant. Toxicity data ORL-MUS LD50 5800 mg kg–1 ; ORL-WMN TDLO 1000 mg kg–1; IVN-MAN TDLO 45 mg kg–1. $50 Not needed. $3000 Responds to change; PI indicates increase in pressure. $25 Methane: NFPA: 1, 4, 0; dangerous fire and explosion hazard. Does not contain oxygen and may cause asphyxia if released in a confined area. Can cause irritation and central nervous system depression at high concentrations. Nitrogen: NFPA 0, 0, 0; Does not contain oxygen and may cause asphyxia if released in a confined area. Hydrogen: NFPA: 0, 4, 0. Hopper clean, blower works well, line appears open. Not needed

439

440

Appendix D Coded Answers for the Questions Posed to Solve the Cases

117 Carefully designed to provide the minimum surface area/ unit volume so as to minimize the potential for ketene to react reversibly with condensed film of water. Area should handle the condensation duty. Gas-liquid KO pots (with barometric legs) provided between different sections of the condenser to facilitate prompt removal of the condensed water. 118 Recalibration not needed. $3500 119 Non-explosive ranging from 17 to 22% of the lower explosive limit for all tests and at all times. $700 120 Allowance for fouling on the shell side = 0.00001 m2 K/W; on the tube side = 0.00001. Use U tube instead of fixed tube sheet, because of thermal expansion. Vapor-liquid ports joining the shell sides were designed for maximum upward butane flow and full condensate downflow based on downcomer velocity considerations to give annular flow. The design is to evaporate the ethylene at 3.9 MPa g (–3.8 C) and superheat the vapor above 1.6 C. $50 121 Not needed. $5000 122 Not needed. $650 123 Yes; everything seems to be OK on our unit. $200 124 No change, no improvement. $400 125 No improvement. $40 000 126 Design checks out OK. No errors made. As expected, a trim cooler was included on the IPA liquid line. $1000 127 –99 –3 C 128 No leaks observed. 129 Steam would leak into the process fluid. $50 130 Awkward. Control is difficult but temperatures are within –2 C of targets. Steam usage on compressor increases. $2000 131 85 kPa. Usually 54 kPa. $600 132 Yes, steam pressure is 3.5 MPa g and brought to the BL of your plant. On site you reduce it to 1.38 MPa.g and further reduced to 350 kPa g for the steam tracing. I keep saying you should be put in local steam turbine drives; you are just wasting thermal energy by dropping the pressure across reducing valves! 133 Should give the correct% vaporization per pass. $700 134 400 kPa g plus pump compared with 20 kPa g on ethylene side: process fluid leaks into ethylene. 135 Pressure is slightly less than usual. On the head-capacity curve suggesting a higher than usual flowrate. However, the pressure is gradually increasing. $300 136 7 s. $150 137 Nothing. We were shut down for maintenance. 138 Checks suggests looks OK. Maybe check the equipment selected when on site to ensure it matches design specs. $200 139 No gauge present. $120 140 Estimate of power needed for pumps A and B, 10 kW each, agree with installed motors and usual efficiency of 70%. $50 141 Responds to change. $300

Appendix D Coded Answers for the Questions Posed to Solve the Cases

142 Pump should supply the design flowrate with ease. A pressure drop of 70kPa across the control valve was allowed for the design. $120 143 Instrument fault/ maldistribution/ hot-gas recirculation/ fouled tubes/ blade wrong pitch/ fan not working/ insufficient tube area/ buildup of non-condensibles in the bottom row of tubes/ tubes not sealed/ no vent break/ poor control system/ control system not well-tuned. 144 Equipment should do the job. Indeed, should overheat because some “small” fouling resistance was used in the design and since this is startup should be clean. $800 145 Well designed; trim selected carefully for caustic; should operate mid-position for design flowrate. $100 148 Not needed. 149 We used new batches of resin. We are still using from the same batch of coloring and foaming agents. $250 150 Based on the performance curve from the original vendor on file and the pressure calculations for the configuration used for this fired heater, the blower should be able to supply the air needed for the new conditions. There are no apparent blockages in the line. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 152 Trace, trace, trace, 10% v/v, trace, trace, trace, 8% v/v, trace, trace. $10 000 153 Four stage, reciprocating with kickback protection at the exit of the third stage. 10% overdesign. $800 155 Decreasing. $50 156 No observable fouling on either inside or outside although some specs of rust are on the inside suggesting eventual buildup. When put back in operation, not change in control and temperature still cycles. $30 000 157 Six months ago, inspected internals of column. Corrosion negligible; everything looked OK. Pumps overhauled with routine checking; instruments calibrated and control system checked and retuned, if needed. 159 Town-gas pressure 1.1 MPa is < steam pressure, 2.8 MPa. $400 160 Heat loss through the metal hub insert is 20% lower than through other parts of the mold simply because of the greater width. $600 162 Moisture coming in because the dryers are not working/ steam leak in the reboiler on the bottoms of the de-ethanizer/ town gas coming in too “wet”/ steam leak from E310 into the town gas/ process gas entering the site is too wet/ column was not designed to handle this high a flow/ collapsed trays. $50 164 Should be OK; allowed for in the Dp across control valve FV-4. Suction side OK, complete with vortex breaker. $300 166 Allowance for fouling on tube side = 0.0005 m2 K/W; shell-side steam = 0.0 0001. $15 168 No change. $5000 169 No change in operation. Pressures and temperatures still increasing. $250 170 P210 = 0.5 m; P220 = 34 m. $50

441

442

Appendix D Coded Answers for the Questions Posed to Solve the Cases

171 Diagram is missing most of the details. All control valves, for example, have block valves, drain, and bypass with valve. The process pump has isolation block valves, pressure gauge; all pressure taps have a tee nipple and valve for flushing; pressure gauges on hot process stream have a pigtail before the gauge. The fuel line has a pressure gauge before the burner. Automatic emergency steam-out connections for set of tubes. Draft gauges upstream and downstream of the damper. Observation ports in radiant and convection sections. If didn’t put on safe-park as first step, then dangerous potential fire/ explosion conditions are created while you experiment. $500 000 173 Steady and usual at 109 C. $50 176 No variation that we can detect. $120 177 121 C saturated corresponding to the pressure. $200 178 Extensive manuals including maintenance schedule and table of troubleshooting diagnostics. $6000 180 Not needed. $2000 181 Just completed; this is first startup. $50 183 On spec. No complaints from other customers for that batch. $350 184 Responds to change. $1000 186 Ammonia: NFPA health 3; flammability 1; reactivity, 0. Toxic, corrosive gas. Overexposure can be fatal. Low dosage: irritation to nose and throat. > 5000 ppm may result in rapid death due to suffocation or fluid in the lungs. Flammable in air for concentrations 15% to 28% v/v. Gas can ignite explosively if released near an active fire. The explosive range broadens 1) if hydrogen is mixed with the ammonia and 2) at higher temperatures and pressures. Presence of oil and combustibles increases fire hazard. Ignition energy > 0.68 J. Autoignition temperature 651 C which is lowered from 842 to 651 C by the presence of iron. At atmospheric pressure, ammonia decomposes to hydrogen at temperatures > 450–500 C. Gas has explosive sensitivity to static charge. Ammonia is highly reactive with most metals, especially mercury, gold or silver compounds. Reacts violently with tellurium tetrabromide and tetrachloride, chlorine, bromine, fluorine and with acid halides, ethylene oxide and hypochlorites. Hydrogen: 0, 4, 0; Dow 21; methane: 1, 4, 0; nitrogen: 0, 0, 0. $350 188 Temperatures and absolute pressures have been gradually increasing. $30 190 UA= q LMTD. Actual= (F cp)E100 90/ 45 C or 2 (F cp)E100. Design= (F cp)E100 96/ 38 C or 2.5 (F cp)E100. Actual UA is 80% of design. $300 191 IS: the operator of the machine several months ago when the product was produced. $60 192 Consistent with reading on temperature sensor. $500 193 Pressure gauge reads 13 kPa abs. $1000 195 IS: Overflow to DAF less than expected. $20 196 Cycle 1, 3 min: ’1: 5.1%; 5.6%; ’3: 6.2%; ’4: 6.7%; ’5: 7.1%; ’6: 7.6%; ’7: 7.2%; ’8: 7.7% ; ’9: 7.9%. Cycle 2&4: 8.0%; 8.1; 8.8; 9.0; 9.2; 9.1; 9.7; 10.3; 10.5. Cycle 3: 15.5; 15.5; 16.0; 17.3; 18.7; 19.9; 20.0; 20.1; 20.2. Cycle 5: 18.3;

Appendix D Coded Answers for the Questions Posed to Solve the Cases

199 200 201 202 203 204 205

206 208 210 211 212 213 214 215 217

218 219 220 221 222 223 224 225 227 229

19.4; 18.7; 19.8; 20.5; 22.3; 23.7; 24.2; 25.1. No data available for previous operation. $4000 Flows and temperatures are constant and the usual values. $120 Responds to change: shows condensation temperature of IPA for the pressure. $2000 Leveling out of the bottoms composition at 0.013% –15%. $500 Characteristic noise of open relief valve. Valve open. Flashy flare suggests that the pressure relief has opened somewhere on the site. $300 Not needed About 200 ppm for each of the 24 h composite for each of the seven days. $70 Could the baffles have been put in so that the windows were top and bottom instead of vertical? That would really foul up the hydraulics, but would this be consistent with the evidence? The fouling coefficients allowed are minimal and extra area was not allowed. When first put into service with clean tubes, the product would have been overcooled! $100 Behaves as expected; no surprises. $2000 Fully closed. $300 Responds to change. $200 Bottoms temperature increases by 3 C; bottoms composition is 8% organic; flowrate from pump 114-J of overhead is still negligible. Reverse jet bag filter; air:cloth ratio 2:1. No bypass installed because this is integral part of the conveying system. Bags replaced annually. No observable reasons. Inside looks a little worn but otherwise OK. Impeller looks OK and the key is in the keyway. $3000 No change All the hydrocarbon vapors are very flammable. $150 The reactor can overheat if the flow of reactants is too small. This not only does not remove the heat of reaction but it also extends the residence time. Therefore, if there is a hot spot, close valve A to increase the gas flow to the reactor. $600 IS: On furnace process exit stream. IS NOT: Elsewhere on this plant or others. Our lab analyses show variability around the specifications. $100 No apparent recirculation. $120 Vortex breaker is welded across the exit nozzle such that most of the cross section is blocked. $10 000 Trace. $10 000 See Chapter 3: thickener, Section 3.5.9; pump Section 3.2.3; belt filter Section Section 3.5.12. $20 See Chapter 3: heat exchangers, Section 3.3.3. $50 Feed forward should not cause unstable behavior. This is feedback control that can become unstable with poor tuning. $50 Not needed. $3000 We make the acid on-site; there has been no change over the past ten years; specs are the same as was used in the design.

443

444

Appendix D Coded Answers for the Questions Posed to Solve the Cases

231 Pressure and overhead temperature reduces slightly but not to the design value. Pressure control still sluggish. $8000 232 Why1? to get column operating on spec during the hot weather. Why2? to produce quality product reliably and safely. Why3? to meet customer’s demands. Conclude: focus on “to get column operating on spec during the hot weather”. $20 233 No controls, other than the controls within the screen and centrifuge system. The bottoms from the crystallizer are pumped to the nozzle feed to the DSM screen. The crystals exit from the screen and flows by gravity into the centrifuge except when the screen is being washed (when there is no flow to the centrifuge). The discharge from the centrifuge to the dryer is continuous except that there is no crystal flow during the wash cycle in the centrifuge. There is no spare equipment. The dryer runs continuously. $30 234 None, this is startup of new plant. $50 236 The block valves on both inlet and outlet of the reflux pump are one size smaller than the pipe size. Check with design plans show these should be the same size as the pipe size. The equipment for the control system agrees with specs. A vent break is included on the exit liquid manifold of the condenser. For the reflux drum, the gas-liquid inlet nozzle is on the top and at the end of the reflux drum opposite to the top nozzle for the vapor outlet to flare and the bottom nozzle for the liquid. The diagrams show a demister pad and a vortex breaker supposedly in the drum but we are unable to tell from the outside. $180 237 IS: Neutralizer temperature < 135 C; product concentration 79%. IS NOT: Neutralizer temperature = 135 C; product concentration 83%. $50 239 Flashing erratically. $500 240 No change in conveying ability. Gauge reads 13 kPa abs. $55 000 241 Responds to change. $300 242 Most details are present on the diagram. Missing, however, are for the reciprocating pump, F1400: the isolation block valves and kickback pressure relief from discharge to suction; for the collection drum V1401, vent break and drain valve; for connecting line between the condenser E1400 and cooler E1404, a block valve; for the collection vessel between the booster ejector and the contact condenser, V1402, a drain valve and vent break; for the barometric leg from the condenser, E1402, a seal pot; for the wet vacuum pump, isolation block valves. $40 243 Not needed. $2000 245 P&ID supplied; no additional data available. P&ID is less than a year old. $30 247 Responds. $120 249 No improvement. $300 250 Analysis 41% ammonia; 29% water vapor, 31% CO2 all –3%. Total 101–3%; OK within error. $50 251 Fan runs at correct speed; balanced operation; no wobble; pitch of blades changes from negative to extreme positive. $10 000 252 Butane in storage tank is “on-spec”. $5000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

253 No change. Still upset and fluctuating 254 Why1? to minimize the heat load in the boiler Why2? to supply steam for energy efficiency. Why3? so that our corporation continues profitably. Conclude: focus on Goal: “to heat the boiler water to 70 C”. $50 255 No indication of a collapsed tray; negligible liquid in column. ($5000 for scan + $1,600 for time for scan + $20 000 for lost time to arrange for scan) $26 600 256 Usual allowance for fouling on tube and shell side. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure is > steam pressure. Fuel-gas pressure, 3.1 MPa > 2.8 MPa steam. $200 257 Usual allowance for fouling on tube and shell side. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Town-gas pressure is < steam pressure. Fuel-gas pressure, 1.1 MPa < 2.8 MPa steam. $1000 258 Fully open. $300 260 Requires three stages of compression to bring the feed gas up to pressure of 34.5 MPa but only one stage of the compressor to pump the recycle gas around the loop. $500 262 Not needed 263 No upsets here. Flow, temperature and concentration should be the usual values with temp. = 108 C; concentration 41% ammonia; 29% water vapor, 30% CO2. $50 264 Because the overheads from the demethanizer are essentially “town gas” the overhead is used as an on-site source of regeneration gas for the dryers. More general information is in Chapter 1 of Process Design and Engineering Practice (1995) by Woods. $15 265 No breaker. No plans for one in the design. $1000 266 Not needed. $10 000 267 No change; when on automatic we have an unstable system whose oscillations are ever-increasing. $1600 268 Not needed. $24 000 269 None recently. $50 270 When: not since the turnaround. 271 Not needed. $650 272 Usual hydrocarbons. Flammability 2. $50 273 Fatty acids are pumped in the bottom of the column C1400, rise through the tubes of the reboiler. The heating medium on the shell side is Dowtherm. The vapors are condensed in E1400, with the overhead product dropping into V1400 where it is subcooled by the coil E1404. The product is removed from the system by reciprocating pump F1400. The distillation occurs under a high vacuum supplied by a booster ejector and a single-stage steam ejector F1402 and F1403 with an interstage barometric, direct contact condenser E1402. Condenser E1400 is a closed circuit of boiling-condensing water. The temperature of the boiling is controlled by setting the pressure P2. The backup condenser, E1401, merely condenses the more volatile contaminants that have

445

446

Appendix D Coded Answers for the Questions Posed to Solve the Cases

274 276 277 278

279 280

281

282 283 284 286 287

288

289 290 291

292 293 294 295

been removed from the fatty acids. These are drained off periodically from vessel V1401. Cooling water comes from the header and some flows into E1401 to condenses the volatiles, and some flows into the tank in which the steam coil is submerged, E1403. The bottoms of column C1400 recycles to second column. Both the recycle lines and the column are not shown. The side-draw above the reboiler on Column C1400 recycles to the feed. This also is not shown on the diagram. $250 IS: thickener-pump-belt filter system. $50 The valves look OK. Went ahead and changed them as ordered. No change in performance. $6000 Nothing unexpected. $200 000 This has been a bear to operate. The overhead temperature has been increasing gradually and I have been increasing the reflux accordingly but the temperature keeps increasing. Yet the analyzer is showing the usual overhead composition. $50 Standard ones we have always used. $150 Alumina. cycle time 12 h; two in series on-line; one off-line being regenerated with fuel gas. Fuel gas should have < 6 ppm moisture. Alumina loading 0.14– 0.22 kg water/kg dry adsorbent. $400 Hollow sound that changes to a deeper sound consistent with liquid level rising in the effect. This corresponds to the cyclical drop in temperatures and the gradual increase in pressure. $600 Both fluctuating but consistent. Fuel oil: NFPA 1, 2, 0. Explosive limits in air 0.5 to 7% v/v. Flash point > 53 C; can react vigorously with oxidizing materials. Process fluid: NFPA 1, 3, 0. Single-speed motor with variable pitch on the blades. Sensor measures the temperature of the exit IPA liquid. $1000 Allowed 130 kPa Dp for the control valve. NPSH required = 2 m water. $200 Was OK. Recalibration not needed. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Checks suggests looks OK provided the gas-liquid separation is sufficient and that there is no carryover to flare. Maybe check location of nozzles when onsite. $200 Oscillating around 65% and increasing; usually 62% and steady. $600 15 C cool and windy; this week started off mild, about 18 C but cooled down as a cold front moved in. $30. Red, indicating stopped. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. Clear. No obstruction. $6000 No files for this unit. $50 They appear to be open. $50 No improvement. $15 000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

296 Operators downstream report:. Temperature = 127 C; should be 130 to 140 C; concentration= 78% and should be 83%. Our plant: temperature = 129 C in the neutralizer; concentration= 79%. Reasonably consistent. $200 298 Not needed. $6000 300 Polycarbonate, ABS Styrenics. NFPA: 0, 0, 0. $60 301 No upsets. Feed rate and composition are within specs. $60 303 Decreasing from 1.73 MPa. $500 304 Responds to change. $300 306 Responds to change. $300. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 307 Based on the measured flame temperature in the radiation section, the temperature is not sufficient to heat the process liquid. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 308 Not needed. $5000 311 Controlling flow via blade pitch is “slow and clumsy”. $1000 312 The pressure on the head tank must be > 22 kPa g because if it is less than this pressure, then the suction of the downstream compressor is subatmospheric. This would allow air to leak into the ethylene. The vessel includes a demister, a vortex breaker. The design provides enough residence time for good liquid level control and for good gas-liquid separation. 313 Confirms that the temperature read on TI 202 is high. $1200 315 Not needed. $3000 316 IS: operators on the debut; IS NOT: operators on other plants. $50 317 See Chapter 3: thickeners, Section 3.5.9; sensors and control Section 3.1.3; pumps, Section 3.2.3. $50 318 Largest particle is 1.3 aperture and is log normally distributed with a geometric mass weighted average of 0.6 aperture size and a geometric standard deviation of 2.1. No data available for previous operation. $1800 319 Not needed. $10 000 320 NFPA ratings are: methane, 1, 4, 0; propane, 1, 4, 0; propylene, 1, 4, 1; hydrogen, 0, 4, 0; ethylene, 1, 4, 2; Explosive limit for hydrogen 4.1–74.3% in air. $200 321 I did that but either the controller isn’t responding, or the meter reading is wrong. The reflux is not increasing! $50 322 Yes. $20 323 105 kPa gauge. Our initial assumption of 200 kPa g was wrong. This changes the estimates made in the office. $200 324 Single-speed motor with variable pitch on the blades. Sensor measures the temperature of the exit from the condenser. $70 325 Noisy from the fans and air flow. Nothing unusual. $150 326 Loud knocking noise. $60 327 Accurate; no recalibration needed. $3000 328 Based on the values available this should flow well. $300

447

448

Appendix D Coded Answers for the Questions Posed to Solve the Cases

329 49 C. $400 330 Not needed. $24 000 331 Diagram reasonably consistent with what’s on the plant. No major surprises. $200 332 IS: not known when this started. $50 333 Continuous. No break or short circuit in heater. Heater should work. $23 000 334 I’m a new operator. I thought I was doing the right thing by increasing the pressure. I certainly didn’t expect everything to go wild. $50 335 Closed cycle, standalone unit. Should do the job. $700 336 Robust unit. No packaging problems are common. 337 IS: flowrate meter reads about 80% of design rate when pump A is pumping. 100% design rate when pump B pumping. IS NOT: pump A at 100% 338 Noise from the fans and air flow. Nothing unusual. $2000 339 Health Rating: 3 – Severe (Poison) Flammability Rating: 2 – Moderate; Reactivity Rating: 2 – Moderate; Contact Rating: 4 – Extreme (Corrosive) Inhalation of concentrated vapors may cause serious damage to the lining of the nose, throat, and lungs. Breathing difficulties may occur. Neither odor nor degree of irritation are adequate to indicate vapor concentration. Ingestion: Swallowing can cause severe injury leading to death. Symptoms include sore throat, vomiting, and diarrhea. Ingestion of as little as 1.0 ml has resulted in perforation of the esophagus. Skin Contact: Contact with concentrated solution may cause serious damage to the skin. Effects may include redness, pain, skin burns. High vapor concentrations may cause skin sensitization. Eye Contact: Eye contact with concentrated solutions may cause severe eye damage followed by loss of sight. Exposure to vapor may cause intense watering and irritation to eyes. Chronic Exposure: Repeated or prolonged exposures may cause darkening of the skin, erosion of exposed front teeth, and chronic inflammation of the nose, throat, and bronchial tubes. Aggravation of Pre-existing Conditions: Persons with pre-existing skin disorders or eye problems, or impaired respiratory function may be more susceptible to the effects of the substance. 340 Yes, put on safe-park. Soot buildup in the furnace could combust at a later time when excess air is present. 341 The steam is more superheated than usual but we have no sensors on the line so we cannot tell actual temperature. 343 about 400 kPa; ammonia < 20 kPa g to have ammonia temperature < 30 C. Therefore process leaks into ammonia. 344 Contamination near the hub/ moist resin/ uneven temperature over the mold/ injection too slow/ injection too fast/ faulty design of the product/ faulty design of the mold/ too much cooling/ too little cooling/ faulty resin/ incorrect foaming agent/ correct foaming agent but wrong concentration/ inadequate mixing/ cooling cycle too long/ cooling cycle too short/ feed temperature too low/ mold too cold. $200 345 I can manage by operating on manual but I have other things I must do. $120 346 Recalibration not needed. $3500

Appendix D Coded Answers for the Questions Posed to Solve the Cases

347 Minimum number of bends; relatively straight and any bends have long radii of curvature. $2000 348 Extensive leaks in 10 out of 340 tubes. $20 400 349 Water: NFPA: 0, 0, 0. $50 350 Both valves leak when fully closed.% leakage varies from valve to valve and ranges from 2 to 3.2%. $40 000 351 Sensor OK. Calibration checked out OK. $8000 352 Ammonia: NFPA health 3; flammability 1; reactivity, 0. Toxic, corrosive gas. Overexposure can be fatal. Low dosage: irritation to nose and throat. > 5000 ppm may result in rapid death due to suffocation or fluid in the lungs. Flammable in air for concentrations 15% to 28% v/v. Gas can ignite explosively if released near an active fire. The explosive range broadens 1) if hydrogen is mixed with the ammonia and 2) at higher temperatures and pressures. Presence of oil and combustibles increases fire hazard. Ignition energy > 0.68 J. Autoignition temperature 651 C which is lowered from 842 to 651 C by the presence of iron. At atmospheric pressure, ammonia decomposes to hydrogen at temperatures > 450–500 C. Gas has explosive sensitivity to static charge. Ammonia is highly reactive with most metals, especially mercury, gold or silver compounds. Reacts violently with tellurium tetrabromide and tetrachloride, chlorine, bromine, fluorine and with acid halides, ethylene oxide and hypochlorites. Hydrogen: NFPA: 0, 4, 0; Dow 21. Extremely flammable. Explosive limit for hydrogen 4.1–74.3% in air. $3000 353 Responds to change. $200 354 Responds to change 355 No apparent fouling. $20 000 356 Reads as expected $150 359 Signal is 0%. This is a fail-close valve so the valve should be wide open. $300 360 Exit moisture content in crystals from dryer increases to 6%. $3000 362 Level above level of tubes; design level. $50 363 Dry reciprocating vacuum pump. Specs suggest that a series of pumps in parallel should provide the vacuum plus reserves for on-line maintenance anticipated because of the high probability that ketene will dimerize into a gunky goo that will have to be mechanically removed from the pumps periodically. Vacuum must be kept constant. 364 68 C. $400 365 Steady and usual design value. $50 366 Other customers were shipped material produced with that number. No, we have not received any complaints or comments about “clumpiness”. Our records show that the material more than satisfied our specifications. $3000 367 Overhead product within spec most of the time. $1800 368 Calibrates OK. $1200 369 45 kPa. Usually 54 kPa. $600 370 Tab marking unclear; cannot tell. 371 Check out OK. Accurate. $20 000 373 Product breaks. $800

449

450

Appendix D Coded Answers for the Questions Posed to Solve the Cases

375 376 377 378 379 380 381

382 384 385

386 387 388 389 390 391 392

393 394 395 396 397 398

399 400 401 402 403 404

405

Responds to change No major hazards with fluids being pumped. Downstream complaints of insufficient product. $1200 No improvement. $35 000 Typical hydrocarbon NFPA: 1, 3, 0. $50 Allowance for fouling on tube side = 0.0 0001 m2 K/W; shell-side steam = 0.0006. $250 At the usual mid-range value before and after safe-park. If didn’t put on safepark as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Looks OK. 69 s. $200 Hydrogen has relatively high heat capacity; about 10 kJ/kg K compared with about 2 for ammonia vapor, about 1 for nitrogen; hydrogen thermal conductivity is about 0.2 W/m K compared with about 0.025 for ammonia vapor and nitrogen. The Pr numbers are similar. Hence, the thermal properties of a mixture of hydrogen, nitrogen and ammonia depend on the gas composition. $6000 Higher than usual. $20 Steady and usual. $50 Agrees within –10%. $1200 IS: safety relief valve popped? IS NOT: any other sensor signal. $50 160 C. $30 Keep the gas pressure drop < 1.5 kPa. Replace bags annually. Exit-gas composition from the reformer will have less hydrogen than you had previously. Otherwise, should work very well; negligible coking; good conversion and selectivity. $800 Reads 67 kPa abs; at other times it might read between 16 and 33 kPa abs. $300 Responds. $250 Usual type. $15 No mercury used. $300 Agrees. IS: mass balance does not balance on the IPA. IS NOT: all other conditions except change in type of condenser seem to be the same. No change in supplier of feed. $650 Not needed. $15 000 Needed. $200 Negligible change. $1000 No change; still cycles. $1200 533 –10% L/s. $90 Tray collapsed in stripping section; vapor space above tray 5 seems “more dense” than expected vapor space. Indeed, more dense vapor space was found between other trays in the stripping section suggesting weeping. $5000 No. $30

Appendix D Coded Answers for the Questions Posed to Solve the Cases

406 Yes, We can ensure that the inerts don’t exceed 15% by opening the purge line in the recycle loop in each separate loop. $500 407 Clear. No obstruction. $1000 408 Vent hole is clear; orifice and seat look OK; mechanism looks OK. Upon restart, cycling and poor control continue. $46 000 409 No change in operation. Pressures and temperatures still increasing. $250 410 5 C. $2000 411 Rain: load = latent heat all condensed + all condensed 2 kJ/kg K 3 C. Hot: load = latent heat partial condensed. Cannot do any more calculations based on the information in the problem statement. 412 Steady at 1.73 MPa. $500 413 Calibrates OK. $300 414 No files for this unit. $50 416 Not needed. $500 417 Flashing; not usual. $50 418 Not needed. 419 Temperature is 13 C; flow is steady 420 On this older part of the site, the water flows in drains that join the main drain at “drop boxes”. The main drain goes to disposal drain. The drop boxes allow mixing and sampling of the water. Rain water also flows from various drain across the unit into the drain system. This is an older part of the site; p and s-traps have not been installed. The safety inspector checks on “explosivity” of the vapor in all the drop boxes on site at least once per shift. $15 421 This is the commissioning startup of a new unit. $50 422 About 50%. this is the design amount. We really want to find a good use for the off-gas and minimize the use of “pure” ammonia. $600 423 Ethylene: NFPA: 1, 4, 2; butylene: 1, 4, 0; methyl chloride: 2, 4, 0; ammonia: 3, 1, 0; water 0, 0, 0. 424 Some improvement. Concentration of propane in flare gas decreases to 80% above design. $5000 425 I tried to keep the cycle time, the pressures and temperatures the same as we used for the prototype. $150 426 Stiction is the sticking and friction related to valve movement and measured as the difference between the driving values needed to overcome static friction upscale and downscale. Likely cause of small amplitude, continuous cycling. $50 427 Operates as flooded underflow. Underflow returns the particles to crystallizer; overflow returns to pregnant liquor storage. Dp and body shape seem well designed. $250 428 Heat loss based on given data of process gas gives calculation of the steam generated under actual conditions as = about 70 kg/s; consistent with gas cooling being only 90 C instead of 150 C 429 Warm and cloudy 20 C. Previously this week it was hot and clear. $50 431 Varies. Sometimes trap hot with condensate intermittent discharge; other times, cold trap and long time between discharges. Sporadic. $500

451

452

Appendix D Coded Answers for the Questions Posed to Solve the Cases

432 Robust unit. If the mixing time speed of rotation= 300 min. rpm then should be OK. 433 Yes, I have noticed that AC/1 shows variation. $100 435 Top temp –18 C usual; bottom temp same as usual. $200 436 Indicating slightly higher concentration than usual. $50 437 P&ID for the columns given in Case ’24. $1000 438 Not needed. $50 439 Calibrates OK; recalibration is not necessary. $3000 440 Not needed. $4000 441 Equipment should do the job. 442 Difficult because the flowrates are continually changing during the upsets. $50 443 Lower than expected. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 444 Usual allowance for fouling on tube side and shell sides. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure is > utility pressure. E107: Process feed, 3.3 MPa > propylene refrigerant pressure. E108: Process feed, 3.29 MPa > ethylene refrigerant pressure. $200 445 Probably should only be considered when overhead process-gas temperature is < 50 C; response is usually sluggish because condenser lag is in the control loop. Watch for fouling on the cooling-water side. $200 446 Why1? to start the production of ammonia in the reactor. Why2? to supply the demand for ammonia. Why3? so that our corporation continues profitably. Conclude: focus on Goal “to heat up the reactors to 500 C”. $50 447 Hot 35 C and humid: 80% relative humidity. Four days ago, it rained and was moderate, 28 C. Since then, it has cleared and gotten hotter and hotter. $650 448 The stuff we are receiving is not acceptable. Temperature = 127 C; should be 130 to 135 C; concentration= 78% and should be 83%. $120 449 No level. $50 450 Not needed. $50 451 Usual allowance for fouling on tube side and shell sides. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure is > utility pressure. E107: Process feed, 3.3 MPa > propylene refrigerant pressure. E108: Process feed, 3.29 MPa > ethylene refrigerant pressure. $1000 452 Missing. Perhaps misfiled. $6000 453 Valves OK. No evidence of corrosion, not plugging of valves. No leakage, valve stem moves easily. $2000 454 No change. $4000 455 No noise sounding like cavitation. $200 456 Flashing. 457 Within specifications. $35000 458 0.532 MPa and increasing. $500 459 OK. Recalibration not necessary.

Appendix D Coded Answers for the Questions Posed to Solve the Cases

460 Increasing. $50 461 10.1; 10.5; 9.9% w/w on three samples. $10 000 462 (E114) Usual allowance for fouling on tube and shell side. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure > ethylene refrigeration pressure. $250 463 Should be OK. allowed for in the Dp across control valve FV-1. $100 464 OK, recalibration not necessary. $700 465 Estimate suggests no major blockage. $1200 466 Utility pressure > atmospheric. $200 467 IS: cyclical and –10 C. IS NOT: steady and –1 C. $150 468 Centrifugal pump. Designed for range of flows and to handle the vertical height. $200 469 Was OK. Recalibration not needed. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 470 levels lower than expected. $20 471 Control should work well or should be tested. $500 472 Whistling noise; no detectable “vibration” noise. $3000 473 Naphtha: NFPA ratings for Health, 1, flammability, 3, spontaneous 0; Dow rating 16; cyclohexane: 1, 3, 0; Dow 16; benzene: 2, 3, 0; Dow 16; hydrogen: 0, 4, 0; Dow 21; hexane: 1, 3, 0; 2 methyl pentane: 1, 3, 0. $150 474 Sharp-edged orifice sized so that the design flow corresponds to conditions for a constant drag coefficient through the orifice. 475 As expected. $3000 476 Not yet but if trouble shooter takes a long time then this should be done. 477 Some condensate but mainly wet steam. $400 478 27 actual trays; anticipate lower tray efficiency in the rectification because of inerts; also allowed two extra trays. Simulation shows that unit was operating very close to design values for usual range of feedstocks and conditions. $350 479 Not needed. $6000 480 Slightly worse operation. $800 481 Thermodynamic traps on all three locations: reboiler, turbine and preheater. No bypass supplied; upstream strainer. $300 482 Suggested that a common fault is a faulty steam trap causing condensate to build up in dryer and reduce effective area for heat transfer. $120 483 Pages 4-90 to 93 in Woods “Process Design and Engineering Practice,” Prentice Hall, 1995. Gives economics, simplified P&ID and description (although the reaction temperatures in the current problem are higher than those described in this references). $500 484 Extensive manuals including maintenance schedule and table of troubleshooting diagnostics. $6000 485 Yes, all lines insulated. $1000 486 The pump F1400 seems to pump some of the product out and then the flow seems to stop. $30

453

454

Appendix D Coded Answers for the Questions Posed to Solve the Cases

487 Fan changes but a very slow response time. $600 488 Why1? to operate all columns at 150 Mg/d safely and on spec products. Why2? to supply the demand for ethylene elsewhere on-site. Why3? so that our corporation continues profitably. Conclude: focus on Why1? “to operate all columns at 150 Mg/d safely and on spec products”. $50 489 If the flowrate and reactant concentration to the reactor are too low, then the reaction quickly reaches equilibrium before the reactants leave the bed. $600 490 Not needed. $2000 491 Cold –10 C. Indeed this whole week has been cold, with snow flurries two days ago. $50 492 When: before the current operation that has problems with the moisture content in the dryer outlet. What was done? To improve the washing, the diameter of the wash pipe and the associated washing pump system were altered so that the flowrates of wash water were increased by 35%. The washing time was to be kept the same, 3 min. This was done. No changes to the dryer or to the screen. $30 493 Greater than usual. $50 494 Warm, 23 C, dry, windy. Similar weather all week. $20 495 Fan very noisy. No improvement in yield of IPA. Calculated yield about 69.8%. $20 000 496 IS: Operators on other units. IS NOT: operators on reformer unit. $150 497 Overhead pressure on the column is controlled by the amount of non-condensibles sent overhead from the reflux drum. The temperature of the condensed overhead is controlled by adjusting the pitch on the blades of the fan. The reflux rate is adjusted by the level in the reflux drum. There is no overhead liquid product. $20 498 Problem still occurs. $2000 499 Since cooling water or air cannot be used as the coolant, various “refrigerants” are used: ethylene, propane, propylene. Each refrigerant is part of a refrigeration unit consisting of the four parts: 1) an “evaporator-condenser”, 2) a refrigerant compressor, 3) a refrigerant condenser and 4) a let-down valve. In the evaporator-condenser, the liquid refrigerant inside the tubes evaporates and causing the process vapor on the shell side to condense. To allow the refrigerant to be reused in a closed cycle, the refrigerant vapor is then compressed, condensed and the resulting liquid reduced in pressure to return around the refrigeration cycle. The pressure on the refrigerant side of the evaporator is adjusted to provide the correct temperature driving force for condensation of the process vapors. More general information is in Chapter 1 of Process Design and Engineering Practice (1995) by Woods. $100 500 Demister should remove mist droplets > 500 mm. $500 501 Guarantee flow needed. Typical moisture content < 6 ppm. Heating value 37 MJ/m3 $1000 502 18 C. This agrees with design assumptions. $400 503 We followed the startup procedure carefully since we have a new catalyst with lower recycle rates of hydrogen and higher feed flows of naphtha. We think we

Appendix D Coded Answers for the Questions Posed to Solve the Cases

504 505 506 507 508 509 510 511 513 515 516 517 518 519 520 521 522 523 524 527 529 530 531

532

533

534

535

are doing a good job of maintaining the same space velocity in the reformer. $150 Operation went smoothly. No problems. No different from a month ago. Steady at 116 C. $300 Remote. $15 Accuracy of float sensor –0.5 cm; accuracy of point gauge –0.045 cm. $120 Nothing unexpected. About two to ten minutes. $200 Should do the job. $100 See Chapter 3: Section 3.9.6. $150 Steady and usual. $50 27 actual trays; anticipate lower tray efficiency in the rectification because of inerts; also allowed two extra trays. $350 3 days ago completed the annual work on the baghouse: new bags installed. Reverse jets had been serviced. No change; still cycles. $1000 All fully open (according to the handle and the valve stem). $20 27 actual trays; anticipate lower tray efficiency in the rectification because of inerts; also allowed two extra trays. $350 Steam inside tubes at 12 MPa with gas on outside at 4.5 MPa; steam leaks into process gas. Same as expected; 3/4 full and reasonably steady. $1200 Responds to change. $20 Not needed. $8000 Some product drains off but quickly stops; prime is lost and pump fails to pump. 10 ppm. $4000 Estimate seems OK. No major blockage. $300 1.48 Mg/m3. $100 1.04 Mg/m3; normal boiling temp. 134 C. $100. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 For the reduced flowrate specified in the startup manual and for the gas composition of 3:1 hydrogen to nitrogen, the cal rod heaters should bring the reactor temperature from 300 C up to 500 C in four hours. 8 am to 12 noon= 4 hours. $6000 See Chapter 3: crystallizer, Section 3.4.3; dryer, Section 3.5.5; hydrocyclone, Section 3.5.8; filtering centrifuge Section Section 3.5.11; screen, Section 3.5.6. $30 No idea. Good questions. That’s something we do not monitor, nor do we have records. I just know that the control of the pressure always tended to be sluggish right from the beginning! $150 See Chapter 3: vacuum systems, Section 3.2.2, solids conveying, Section 3.2-4; hopper design, Section 3.10. $2000

455

456

Appendix D Coded Answers for the Questions Posed to Solve the Cases

537 IS: Everything seems to be the usual behavior. IS NOT: different than usual. $50 538 Operation went smoothly. No problems. No different from a month ago. 540 Full documentation is available. $15 541 Allow 170 kPa across the steam control valve, gives pressure in coil= 1530 kPa. Inverted bucket sized on Dp= 1530 – backpressure in condensate header of 200 kPa = 1330 kPa. Condensate = 0.12 kg/s and select based on 8 times or 1 kg/s. Strainer included. 542 Well designed; trim selected carefully for gaseous chlorine; should operate mid-position for design flowrate. $100 543 Yes, excessive fouling. $4000. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 545 No feed to the dryer during the centrifuge wash cycle either before or after the change. $700 546 Hot summer day. Same weather all week. $50 549 Should be OK; allowed for in the Dp across control valve FV-4. Suction side OK, complete with vortex breaker. $300 550 P210 = 28 m; P 220 = 22 m. $50 551 Not available. $15 552 Nothing has changed but the problem seems to be getting worse as the shift progresses. $50 553 Systems respond as expected. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 554 Explosive ranging from 170 to 195% of the lower explosive limit for all tests and at all times. $700 555 See Chapter 3: reactor Section 3.6.2; exchangers/boilers Section 3.3.3 556 About 78% design rate and decreasing. usually at current production rate this would be 87% design rate and steady. $1000 558 Cycle 1, 10 min: ’1: 3%; ’2: 2.9%; ’3: 3.3%; ’4: 3.8%; ’5: 4.2%; ’6: 4.7%; ’7: 4.9%; ’8: 4.9%; ’9: 5.3%; ’10: 5.2. Cycles 2 and 3 similar: ’11: 5.4%; ’12: 5.7%; ’13: 6.0%; ’14: 6.1%; ’15: 6.3%; ’16: 6.4%; ’17: 6.6%; ’18: 6.9%; ’19: 7.1%; and ’20: 7.2%. $3500 559 No upsets. We can supply on-spec feed for up to 160 Mg/d production with no problem. $300 560 Downstream production steady; no fluctuations. 561 No feed to the centrifuge during the screen-wash cycle either before or after the change. $1200 563 Zero. $50 564 IS: other units. IS NOT: on reformer unit. $150 566 IS: past few months. Gradual not sudden. IS NOT: previous year although the slow response has been present even then. $50 567 Area sufficient to do the job with the following fouling factor allowances: shell side 0.0002; tube side 0.0005 m2 K/W. Considered propylene boiling to range

Appendix D Coded Answers for the Questions Posed to Solve the Cases

569 570 571 572 574 576

577 579 580 582 584 585 586 588 589

591 592 593 594 595 596

597

598

over nucleate and film boiling shell side 0.0002; tube side 0.0005 m2 K/W. Indeed, since this is a clean bundle, we would expect overcooling when we first start up because the allowed fouling does not interfere with the heat transfer. The observed overcooling is expected. $200 Equipment should do the job. No. I have operated this plant for six years and have never seen this before. $50 Fluctuates. $200 Responds to change. IS: bottoms and pump 114-J. IS NOT: other parts of the unit. Commissioning of the compressor last summer went well. All systems performed well regardless of the variation in atmospheric pressure and the usual range of flare-gas compositions. The seal pot used water since it was summer operation. Recommended change over for October to May to kerosene obtained from the crude unit. $500. OK. No calibration needed. $30 000 515 –10% L/s. $85 Respond to change. $500 Estimate suggests no blockage. $400 Both start to decrease. As expected. Responds to change. $100 Small amounts of rust flecks but not enough to block strainer. Removed what was there. Upon restart, cycling and poor control continues. $6000 Sulfinol removes the carbon dioxide from the reformer gas to produce a relatively pure feed of nitrogen-hydrogen to the ammonia synthesis reactor. The sulfinol needs to be regenerated occasionally. This is the cleanup column. Allowance for fouling on tube side, water = 0.0002 m2 K/W; shell-side stream= 0.0005. Care to prevent temperature cross-over. $400 Should work fine. Head-capacity curves on file. rpm 1800. $500 Temperature about 19 C. $1000 4.2 MPa. $150 Clean, trim looks fine; snug bonnet. Size is two sizes smaller than line size. $1000 Why1? to prevent possible shutdown. Why2? to keep plant operating safely and prevent upsets in operation. Why3? so that plant can handle the waste water it receives. Conclude: focus on “to keep plant operating safely and prevent upsets in operation”. $50 Why1? to safely recycle vapors for reprocessing and prevent them from going to the flare. Why2? to convert vapors into valuable products and keep the neighbors happy by preventing flares. Why3? to maintain good community relationships and keep the company economically viable. Conclude: focus on “to safely recycle vapors for reprocessing and prevent them from going to the flare”. $50 Upstream and downstream distances consistent with good design.

457

458

Appendix D Coded Answers for the Questions Posed to Solve the Cases

599 600 601 602 603 604 605 606 607 608

609

610 611 612 613 614 615 616 617 619 621 622

623 624 625 627 630 631 632

Mid-point. $300 Was OK. $2000 Steady and usual design value. $50 Both motors start. Flow, from FRC/ –100, increases when each is turned on. $30 Within –10%. $700 Nothing changed from usual operation except the liquid level in the flocculation tank is not at the normal level. $20 Agrees with head capacity. $200 No change. $600 OK, recalibration not necessary. $700 Specialized design. Should do the job. Some designs use a standby furnace to bring the system up to temperature. This design uses cal rod heaters. Thermocouples in the catalyst bed and in the feed gas entering the bed. Temperature sensors at four levels in the catalyst bed. $750 Allowance for fouling on the shell side = 0.0 0001 m2 K/W; on the steam side = 0.0006. Use U tube instead of fixed tube sheet, because of thermal expansion. Vapor-liquid ports joining the shell sides were designed for maximum upward butane flow and full condensate downflow based on downcomer velocity considerations to give annular flow. $50 OK, recalibration not necessary. $800 Fluctuates. $200 Wet and rusty flecks come out with steam. $1000 For some flowrates the pressure differential is such that the pump can be bypassed. The pump is adequately sized to handle higher flowrates. Almost closed. $200 Valve stem responds as expected. $230 Responds to change. $200 No noticeable change in sound. $800 Steady at 116 C. $300 IS: Overhead of column. Designed to provide good residence time; sufficient disengaging space; sufficient area for vaporization, liquid level so that the coil is submerged. Coil area sufficient and designed for film boiling. No change, still see weld lines. Product breaks. $900 Purge specified in the manual should maintain the desired level of methane in the gas. $6000 See Chapter 3: instruments Section 3.1.3, compressor, Section 3.2.1, knockout pots, Section 3.5.1, $150 No leaks observed. Consistent with overhead temperature and the overhead specs; increased over usual. $100 Lumps persist. $6,500 Equipment should do the job. $2000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

633 Sample 1: 13%; sample 2: 11%; sample 3: 12%; sample 4: 14%; sample 5: 13.5%. $20 000 635 Calibrates OK. $1500 636 Usual design. $400 637 Why1? to continue to operate the plant without being shut down by the inspector. Why2? to meet production deadlines. Why3? to keep the company economically viable. Conclude: focus on “to continue to operate the plant without being shut down by the inspector.” $30 640 After one hour, the temperature has risen 50 C to 375 C. $30 000 641 Pump should do the job. Based on current PI pressure the pump should be pumping 11.5 L/s. $50 642 Steam-gauge pressure upstream of control valve = 230 kPa for saturation temperature of 136.8 C. This would give a DT of 136.8–109 = 27.8 C. However, with an estimated Dp across the control valve of 30 kPa, then the DT = 25 C if the steam was saturated at this pressure. Expect wiredrawing so the steam will be slightly superheated until it reaches saturation. The DT seems to be in the nucleate boiling range. $200 643 Expected value. $50 645 Steam leak into the acid. $90 647 Should do the job. 648 Working OK sometimes. Sometimes the flowrate = 14 L/s. Other times the flowrate < 14 L/s. What’s going on? We expect reliable flow from you. $20 650 There are no standard operating procedures. However, I have been operating this unit for many years so I think I understand its peculiarities although it never has worked correctly. However, today it’s much, much worse than I have encountered before. Started it up as usual; carefully set values at usual values; ensured the steam tracing was on because it’s so cold out there. But it just won’t draw off the product. 651 Unable to control temperature with cooling-water feed valve full open. $300. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 652 Yes. From past records, the timing varies and the concentration values differ but surges in high lab values for the bottoms “correspond with” surges in high values of C3 on AC/1. $650 653 Improvement, extension of the period before break but still not satisfactory. $1000 655 Behaves as expected; no surprises. $2000 656 No change. $500 657 No contaminants. $3500 659 The setting on PIC/10 should be such that the column pressure is 1.7 MPa but does not exceed 1.8 MPa, the relief pressure for PSV 1. $50 660 After two hours, no increase in temperature above 325 C. $45 000 661 Process fluid leaks into furnace. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000

459

460

Appendix D Coded Answers for the Questions Posed to Solve the Cases

662 Vacuum leak/ variation in steam pressure to booster or other ejectors/ liquid not subcooled enough and it is flashing in pump F1400/ wet vacuum pump not pumping at capacity/ leak in exchanger E1400 causing water to flow into vacuum system/ instrument error/ more volatile material in feed/ vapor lock in suction line because of no vent break/ leg from barometric condenser not sealed/ booster not working right and air is sucked down from the booster into the pump F1400/ dry vacuum pump F14 put into service without 30-minute warmup. 663 No improvement. $6000 664 See Chapter 3: distillation Section 3.4.2, perhaps pertinent condensers and reboilers, Section 3.3.3; pumps, Section 3.2.3 and controllers 3.1.1. $50 666 Already insulated. $500 667 Polypropylene: melt point 162 C; melt flow rate 35 g/10 min; Izod impact strength, 70 J/m (1.2); HDT 105 C; barrel temperature 180 to 220 C; nozzle temperature, 200 to 220 C; melt temperature 200 to 240 C; injection pressure 60 to 80 MPa; mold temperature 15 to 50; target 25 C. $100 668 Was OK. $2000 670 Equipment should do the job. $2000 672 Ethylene leaks into the butane. $50 673 Steady and usual. $50 675 We are receiving more because of the higher through put you are using. We haven’t checked our knockout pots on the fuel gas lines to see if there is any carryover. $300 676 No difference. Solution meets specifications. $1000 677 Not needed. $1000 678 The check valve on the exit of pump B is faulty. The other looks OK. Went ahead and changed both of them as ordered. Performance back to normal. $6000 679 IS: on top of column. IS NOT: elsewhere on the unit. $50 681 OK recalibration not necessary. $800 683 About 93% design rate and increasing; usually at current production rate this would be 87% design rate and steady. $1000 684 Checks out OK. Recalibration not needed. $800 686 As usual. Same as before the shutdown. $300 688 No change, no improvement. $400 689 Everything OK. no apparent steam leaks. Buildup of cake on the tube near the inlet. Flights in good shape. $4500 690 Pressure, when converted to head based on the assumed density of the process fluid, agrees with the no-flow condition on the vendor’s head-capacity curve. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 691 Equal to the static head about 15 m whereas head-capacity curve is about 20 m. $300 694 Negligible amount drains from KO pot. $200 695 Yes.

Appendix D Coded Answers for the Questions Posed to Solve the Cases

697 Sharp-edged orifice plate sized to give CD = constant for range of flows; adequate straight length of pipe upstream and downstream. $100 698 Details of all utilities, instrument and lubrication lines are not on the diagram. There are a drain at the bottom of the neutralizer and the usual vents, sight glass and pressure-relief valves on the neutralizer. There is no measurement of flow on the cooling water. Nitric acid is steam heated in the storage tank via a trombone heater with the condensate going to drain; demineralized water goes to the lantern ring-packing seal in the transfer pump. All sample lines have water flush; the pigtails on the pressure gauges have tee-valve for water flush. $600 699 Responds to change. $200 700 Loud crackling noise, like cavitation, when < 6 g/L NaOH equivalent acid addition is required. $1200 701 Working well. $400 702 OK, recalibration not necessary. $700 703 18 –10% L/s. $80 705 OK; recalibration not necessary. $10 000 706 Followed standard procedure; we were all especially aware of the impact of trace amount of oil left in the system and did our best to ensure none was present. $1200 707 Deep thickener with central rake, designed for influent solids 1 to 4% solids; underflow 3 to 6% solids with a mass loading of 0.4 to 0.55 g/m2 s. $30 709 TDI from the new supplier doesn’t contain benzoyl chloride. The TDI from the previous supplier had benzoyl chloride as a reaction modifier. $3000. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 711 Filter cycle: pressure nozzle feed, Model 1600 DSM 120, 10-min screen, wash cycle 3 min. $100 713 500 kPa g –10. $50 714 Appears to be fully open. The direction of flow through valve is consistent with expected flow. 715 Level indicates the coil is covered 716 No leaks apparent. $100 717 No noise. Valve not open. No evidence from flare that the pressure relief has opened. $300 719 NFPA: 1, 1, 0; irritant to the eyes. 720 Respond to change. $500 721 227 C. $200 722 Four months ago. None on this unit. $20 723 Operator had turned this wide open. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 725 Allowance for fouling on tube side, water = 0.0002 m2 K/W; shell-side stream= 0.0005. Care to prevent temperature cross-over. $400 727 205 kPa g and steady. $900

461

462

Appendix D Coded Answers for the Questions Posed to Solve the Cases

728 729 730 731 732 735 736 738

739 740 742 743 744 746 747 749 750

751

752 753 756 758 761 763 765 766 767 768 769 770

Nothing. We were shut down for maintenance. No ice. $10 000 4.2 MPa. $150 Yes. $50 Not needed. $4000 50 C. $300 Usually 90 C and relatively steady. $300 Why1? so that other units can function well. Why2? to keep the overall plant operating safely and producing specification product. Why3? so that company is economically profitable. Conclude: focus on the goal statement: “to provide the usual amount of heat to the process streams of other units”. $50 Design checks out OK. No errors made. No water-cooled trim cooler was included. $100 UA= q LMTD. Actual= (F cp)E101 66/ 32 C or 2.1 (F cp)E101. Design= (F cp)E101 58/ 36.5 C or 1.6 (F cp)E101. Actual UA is 30% higher than design. $300 Should work OK. If you are concerned about surge, check the amps. $500 Water would leak into the neutralizer. $60 Not needed. $3000 At half-load: UA= Feed flow. cp. (33.5/ 27); At full load: UA= Feed flow. cp. (31/56); UA full load is 0.44 of UA for half-load. Responds to change No improvement. $8000 The temperature in the neutralizer started to decrease. We took extra samples of the product and found this was more dilute than usual. The exit coolingwater temperature is slightly less than usual. Everything else seems normal. $120 See Chapter 3: reactor, Section 3.6.2; compressor Section 3.2.1; sensors, valves and control Section 3.1.3; knockout pot Section 3.5.1; exchangers, Section 3.3.3; refrigeration cycle Section 3.3.4. $6000 Increasing concentration of heavies. Overhead off-spec. $50 Bitter cold, –15 C, blustery with snow squalls. The weather has been slightly warmer in the week but has gotten colder each day. Recalibration not necessary IS: pressure above setpoint and controller output = 0%. IS NOT: steady and controller output mid-range. $50 No observable change in level on tray 4; no observed flow in the bulls eye. No change in swinging-loop phenomena. $80 000 Design specs. seem OK. Suction nozzle 10 cm nominal diameter; discharge nozzle 7.5 cm diameter Sludge-handling pump. $20 Steady and usual value. $50 Motor running. Shaft rotating in the “correct” direction of rotation. $400 Both are steady We’re not getting much and what we get is off-spec. $50 Might be better with a float trap with a separate trap from each of the three stages. The “design” condensate load should be more than double your design

Appendix D Coded Answers for the Questions Posed to Solve the Cases

771 772 774 775 776 777 778 779 780 781 782 783 785 786 789 790 791 792

793 795 797 798 799 800 801

802 803 804 806

807

load of 1.6 kg/s. However, for an inverted bucket trap you must ensure that you have the correct Dp across the trap. Specs on size distribution of particles. Work OK. $120 Some improvement. Concentration of propane in flare gas decreases to 50% over the design value. $6000 No change. Still cannot control. $10 000 No change. Same as expected. $1200 No gauge, can’t tell. $30 Yes. What’s going on? $30 Not needed. $50 3.15 L/s with diameter 5.5 cm= 1.3 m/s. $50 Sink marks; product breaks. $2000 When: 8 months ago. During this turnaround, routine checks were done on equipment. $400 Liquid appears in the sight glass and rises. When the flow is stopped, the level gradually falls, in the sight glass, slowly over the 10 minutes. $1000 Negligible change. $1000 Steady and usual 45 C. $50 Nothing unexpected. No water leaks observed. $20 000 The pressure at the PRCV must be > 22 kPa g to prevent subatmospheric conditions at the compressor suction. If subatmospheric pressure occurs, there is a chance that air will leak into the ethylene stream. There is still no flow of sludge to the filter. $40 No improvement, $6000 Usual value. No improvement. $300 Same size. No evidence of surge. $2000 From the pressure gauge, an estimate of the pressure drops and from the pump head-capacity curve from the original vendor on file the pump should do the job and no apparent blockages are in the line. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 50 C. $300 Confirm the temps and pressures at the bottom and top of column are consistent with the compositions expected. $100 Responds to change. Suppliers were aware of our specs; we were all especially aware of the impact of trace amount of oil left in the system and did our best to ensure none was present $2000 Improvement, extension of the period before break but still not satisfactory for the customer. $15000

463

464

Appendix D Coded Answers for the Questions Posed to Solve the Cases

809 811 812 813 814 816

818 820

821 823 824 825

827 831 832 834 838 842 843

844 846 847 849 850 851 852

IS: On the ethylene exit line. IS NOT: upstream or in butane unit. $50 78 –1%. $2300 Robust unit. Should work well for cement. Alarm faulty. $300 Fluctuates. $200 Continuous crystal discharge by peeler. No crystal discharge during the wash cycle (that is used to clear the fines from blinding the media.) cycle times, 10 min filter; wash media cycle 3 min $120 Fortunately a spare was available in stores. No change. When capacity was increased to 150 Mg/d we encounter the same problems as before. $80 000 Lower than expected. Occasion puffs of “afterburn” with local temperatures much higher than normal in the convection section. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Difficult because the flowrates are continually changing during the upsets. $50 Should do the job. Inverted bucket with upstream strainer. No bypass. Sized on double the condensate flow. $200 Check on design calculations show liberal allowance for the pressure drop across the control valve so that it operates mid-range. Pump should have no trouble supplying the design reflux. $100 IS: Gas samples are explosive and pose immediate hazard. IS NOT: gas samples have non-explosive gaseous mixtures. $30 The hopper should be well designed. It has rained recently and if moisture got into the powder you might run the risk of bridging. $3000 Hot 33 C and humid. This hot spell has continued all week. $150 No improvement. $7500 No observable decrease in pressure. $6000 No change, no improvement. $400 Emergency/ action to prevent explosion. Yes, there are a variety of hypotheses to explore later to discover why the temperature runaway: cooling-water failure/ coolant-inlet temperature too high/ cooling-surface fouled/ addition too fast/ operating procedures not followed/ sensors wrong/ different supplier of TDI/ contamination of TDI in the storage tanks/ different polyol/ contamination of polyol in the storage tanks or lines. $300. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 Equipment should do the job. $2000 Not needed. $15 000 Usual value but fluctuating; past records show similar values. Bypass valve shut; block valves open. $80 Calibrates OK. $8000 Working well. $400 Designed to shut down if the suction pressure exceeds 112 kPa. $500

Appendix D Coded Answers for the Questions Posed to Solve the Cases

853 854 855 857 858 860 861

863

864 865 866 867 868 871 872 873 874 875 876 878 880 881 882 883 885 886 887 889 890 892 894 896 897

Fully open. $300 OK. Recalibration not necessary. OK. Recalibration not necessary. Should be OK. allowed for in the Dp across control valve FV-1. $100 Sound changes about half-way up the condenser. $600 More cycling. $600 Rake speed is kept consistent with feed concentration of 400 ppm but because the torque has increased I increased the sludge pump rate, even up to wideopen, as given in Standard Procedures. The supernatant was still around 200 ppm and the belt filter was flooded. $120 Half-load: DT goes from 75 to 5 C. Starts at film boiling, flux decreases and then increases as shifts to nucleate near the exit. Full load: DT goes from 75 to 41 C. Starts at film boiling, and flux decreases near the exit. Fluctuates. $200 The reverse jet works automatically. When trouble shooter is out on the plant the cycle seems to be working. Fluctuates. $200 No change in swinging-loop phenomena. Loss of reactant gas to bleed. $100 000 OK; head-capacity curve OK relative to estimated pressure requirement. $240 Can’t tell. No level indicator nor sight glass was installed. $30 Propane flowrate is way below normal; the quality is OK. $120 Cycle time increases; product breaks more easily. $1000 Startup seemed to be methodical and carefully following the guidelines supplied by the vendor of the new catalyst. $150 Appears to be fully closed. The direction of flow through valve is consistent with expected flow. 45 C. $3400 None available. $30 Three months ago; usual maintenance on pump. $200 Trace amounts of water; no ethylene. $5000 Usual. $50 U tube, horizontal. $100 Low level that is rising. $200 93 C; about 3 C higher than usual. $100 No evidence of surge. $2000 No change. Still upset and fluctuating Power agrees with expected draw for the flowrate. $200 8, 83, 105, 43 ppm $6000 No change. $2000 Amps as expected excepted when low flowrates of acid are required. $1000 P3 is < atmospheric, as it should be. Therefore air leaks into the furnace; excess fuel would not leak out of the furnace. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000

465

466

Appendix D Coded Answers for the Questions Posed to Solve the Cases

898 899 900 901 902 903

904 906

907

909

910 911

912 913 914 915 916 917

919 920 922 925 926

Just start the pump and let it run. $10 No improvement. $20 000 Heavies concentration exceed specifications. $800 Not successful. $1000 Boiling water on outside tubes at 12. + hydraulic head MPa with gas on inside of tubes at 4.55 MPa; water leaks into process gas and becomes steam. Part of the design of the reactor. Incoming gas flows across tubes carrying hot reactor effluent; then on the tube side with the hot gas from the catalyst bed on the shell side. The heated feed gas then goes up and then down through the catalyst bed. Thermocouples monitor the temperature of the exit gas as it cools down. $800 54 MPa. $200 Fixed roof, usual design with breather, heater, level gauge, sensors. Tried to design to minimize water condensation inside the vessel. Ensured roof has good seal to prevent rain from entering vessel. Should do the job. $300 Design load = Feed flow. cp. 67 C; Actually get (at half-load) = Feed flow. cp. 33.5 C; Actually get (at full load) = Feed flow. cp. 31 C; Slightly better for half-load than full load. Before the change the feedrate to the dryer was relatively constant for 10 minutes; then feed dropped to zero for the 3 minutes of wash cycle; then cycle repeats. After the change, the feedrate to the dryer only lasts for 3 minutes, then the wash cycle for the centrifuge, then feed to the dryer and so one. During the 3 min feed to the dryer, the rate seems to be all over the place. First, it seems about the same as before; another time it is about 30% higher and another time it is about double. $1200 19 C. $300 No formal records kept. The operator had worked with polypropylene and UV stablizers often, previously. No lumps appear in prototype when viewed by transmitted light. $100 Valve A fully open. No change in temperature. $6000 Everything OK. $3000 Propylene: NFPA: 1, 4, 1. Extremely flammable. $50 Fluctuates. $200 No upsets. Steam pressure 1.2 MPa g. 3 C superheat. Should be OK even with increase in column pressure by 100 kPa; allowed for in the Dp across control valve FV-4. Suction side OK, complete with vortex breaker. $300 Not needed. 495 C gradually increasing to 510 C. Bed is cold. Usually 510 increasing to 550 C at exit of bed. $1000 Nothing, except we are not getting the amount of heating we expect. $150 Not easy without shutting down plant. $3000 Estimates agree with Dp. $2000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

929 Before “safe-park” = 0%; under “safe-park” conditions = 5%. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 930 Full documentation is available. Model 2GYUO;, 7 L/s; 10 m head, 1 kW; 1750 rpm, efficiency= 70%; NPSH required at 7 L/s = 1.1 m water. At zero flowrate, head = 12 m. $15 931 Yes 932 No difference. Solution meets specifications. $1000 933 Responds to change. $20 935 Bypass valve was closed; both block valves were open. $800 936 No pressure decrease. $15 000 938 Very fine particles, < 150 mm, free flowing with angle of repose 15 to 30, mildly abrasive with Moh’s hardness 2–3. $1000 940 Sample 1: 0.95%; 2: 0.98%; 3: 1.10%; 4: 1.00%; 5: 0.95%; 6: 1.05%; 7: 1.06%; 8: 0.96%; 9: 0.88%; 10: 0.99%; 11: 1.10%; Insert swing: samples during 10-minute swinging loop: S1: 0.98%; S2: 0.99%; S3: 0.99%; S4: 0.98%; S5: 0.99% 12: 1.06%; 13: 1.02%; 14: 1.01%; 15: 1.02%; 16: 1.10%; 17: 0.94%; 18: 0.98%; 19: 0.95%; 20: 0.89%; 21: 1.01%; 22: 1.02%; 23: 1.01%. The swinging occurred in the middle of hour 11. $80 000 941 Operation went smoothly. No problems. No different from a month ago. 943 We’ve tried everything. We have the reflux set at the highest value but, as you can see, the reflux rate, FIC/4, is below normal. $50 944 Level mid-drum; design level. $50 945 Negligible variation detected over five minute. $1200 946 Usual value and steady; past records show similar value. 947 Not needed. $4000 948 Agrees. $200 949 Not needed. $6000 950 Moderate; 21 C; cloudy; humid. Typical weather for all week. $50 951 This is the startup after a shutdown. New catalyst was packed in the reformer. Routine checks made of fouling on tubes in exchangers; checking of the sensors. Trapped air was vented from all exchangers before startup. Larger-size impeller put in naphtha pump to handle the increased flowrate. $400. 952 No change. $3000 953 Clean, clear, works well. $500 954 Standard design, 50 cm width; stilling well and float-operated recorder for height of liquid level. $400 955 FC-1 = 6.8 L/s. Before we put the system on “safe-park” it read 8.2 L/s. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 956 50.2; 49.8; 50.4% w/w. $10 000 957 0.8 Mg/m3. Color, crystal clear. $150 958 Trace amounts of butane. $5000 959 Estimate seems OK. No major blockage. $300 960 21 C. $300

467

468

Appendix D Coded Answers for the Questions Posed to Solve the Cases

961 962 963 964 965 966 967 969

972 973 974 976 977

979 980 981 982 983

984 985 986 987

988 989

990

Within normal specs. $800 Steady at 70 C. $200 Responds to change. $100 IS: after two months operation; summer. IS NOT: immediately at startup although “control was not very good”. $50 Product breaks. $1500 Allowance for fouling on tube side = 0.0001 m2 K/W; shell-side steam = 0.0 0001. Uses tower water. $160 See Chapter 3: boilers and condensers, Section 3.3.3; sensors and control, Section 3.1.3. $50 Allowance for fouling on tube side = 0.00001 m2 K/W; shell-side steam = 0.0006. Simulation shows that unit was operating very close to design values for usual range of feedstocks and conditions. $250 Not needed. $50 No idea. There is no pressure gauge on the tank. $15 None that we know about. The feed has been in storage tanks during the shutdown. $50. See Chapter 3: instruments and controllers, Section 3.1.3; pumps, Section 3.2.3. $50 Calibrates OK. $1200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. IS: feed flowrate is gradually being reduced. IS NOT: the normal feedrate. $50 Steady and usual. $50 Estimate 1.5 m/ (1.3 m/s) = 1.1 s. $50 No additional data available. $200 This system has worked extremely well in all our installations. Where we have had trouble has been when the on-site utilities have not been reliable: steam, cooling water, condensate handling. Oh, we did have two cases where the product foamed severely. That caused cycling and reduced capacity! Not needed. $2000 Packed column providing good contact between ketene vapor and glacial liquid acetic acid to produce acetic anhydride. IS: always since startup. (E113) Usual allowance for fouling on tube and shell side. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure > propylene refrigeration pressure. $250 Valves stems respond to change. $600 For ethylene, 1.6 C corresponds to 4.4 MPa abs; 8.5 C, to 5.1 MPa abs; –3.8 C, to 3.9 MPa g. For n- butane: 0.58 MPa abs corresponds with 57.8 C; 0.97 MPa abs, to 78.5 C. For steam at 0.8 MPa g corresponds with 170 C. $50 Bumps appear on the product opposite to the knock out pins. Product still breaks. $800

Appendix D Coded Answers for the Questions Posed to Solve the Cases

991 This process always worked OK. $400. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 992 Head-capacity curve available. $40 993 Value of 3.5 m compared with 2.5 m required. $100 994 Less than design. $50 995 Sized on condensate discharge with a safety factor of 2 times the nominal flowrate of condensate. $150 996 IS: OK at 8 am; IS NOT: OK at noon or later. $3000 997 Rectangular box 0.5 m 1 m with coarse screen to remove metal and large material that might damage impeller of downstream pump. $20 998 Control system tuned by process-control specialists. $400 999 Not needed. $500 1000 503, 547, 602, 698, 587 ppm. $6000 1001 Equipment should do the job. 1002 IS: operators on this plant and operators on downstream IPA processing plant. IS NOT: at this time no complaints from upstream or downstream vent scrubber personnel. $650 1003 Estimate of head required agrees with that supplied by pump. $1000 1004 Resin polypropylene plus UV stablizer and pearlized pigment. No blowing agent. $50 1005 Blower-performance curve available, specs available. No pressure gauge on the exit line so we cannot tell. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $50 000 1006 Use either of the pumps A or B. We have the extra pump available in case one pump needs maintenance. Then we can swing the other pump on-line and keep the process going. Occasionally, if capacity demands increase beyond what one pump is capable of producing, start both pumps and operate in parallel. $30 1007 No improvement. $15 000 1008 Why1? to contact Fuller’s earth with oil in the deodorizer; Why2? to contact Fuller’s earth with oil; Why3? to remove the color bodies from the oil; Why4? to provide oil suitable for downstream processing. Conclude: focus on goal “to suck Fuller’s earth into the deodorizer.” 1009 Before: screen 10 min; wash 3 min. Now: screen 3 min; wash 3 min; $30 1010 Given in Chapter 2 and on page 2-68 of “Process Design and Engineering Practice”, Woods, Prentice Hall (1995). 1011 Bypass valves: butterfly valves. These do not seal tight enough and cause gas leakage even when they are “shut”. Replaced with globe valves. $6000 1012 Reads 24 C. $1000 1013 Responds to change. $100 1014 Steam pressure the nominal 1.5 MPa g; we have a slight superheat here so that the steam should be saturated at your unit. $500

469

470

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1015 Yes, for this production rate. Water should be able to condense all the vapor product. $50 1016 No noise sounding like cavitation. $200 1017 See Chapter 3: distillation, Section 3.4.2; condensers, Section 3.3.3; control and sensors, Section 3.1.3. $700 1018 IS: exit gas is 730 C. superheat temperature is 443 C. IS NOT: exit gas design value of 600 C; superheat is not 353 C. 1019 Estimate 1.25 kg/L 1.3 m/s 0.055 m/ 30 10–3 = 2900. Probably turbulent. I would be more comfortable if the Re was three times higher. $50 1020 Yes. $30 1021 None done. This is just startup. Plan to repack the circulation pumps on the absorber-reaction system every time the furnace is decoked about every 10 days. 1022 Cooling time lengthened to 51–60 s and product looks OK although the production decreases because of the increased cycle time per piece. Further tests show the product still breaks. $800 1023 All air and water tests done prior to startup showed no leaks. Care was taken to remove the air and water from all process equipment prior to startup. 1024 Set at 20% open; usual value and steady. $600 1025 Fluctuating wildly. $300 1026 For the air: Air flowrate is unknown; we can measure the air temperature and humidity in and out, estimate the heat capacity but we really can’t estimate this without plant data. The files show that the new exchanger should condense all the IPA with ease. $3200 1027 Sharp-edged orifice plate sized to give CD = constant for range of flows; adequate straight length of pipe upstream and downstream 1028 Stage 1: 0.58 kg/s; from stages 2 and 3 combined 1.0 kg/s (neglecting about 10% loss as flash steam) or total condensate load of 1.58 kg/s. $700 1029 Not needed. 1030 Head = 46.5 m at zero flow. $50 1031 Area sufficient to do the job with the following fouling factor allowances: shell side 0.0002; tube side 0.0003 m2 K/W. Considered propylene boiling to range over nucleate and film boiling. $200 1032 Cycle 1, 10 min. Cycles 2 & 3 similar: ’1: 5.1%; ’2: 5.6%; ’3: 6.2%; ’4: 6.7%; ’5: 7.1%; ’6: 7.6%; ’7: 7.2%; ’8: 7.7%; ’9: 7.9%; $3700 1034 Reads OK. $10 000 1036 Allowance for fouling on tube side = 0.0001 m2 K/W; shell-side steam = 0.00001. Slightly overdesigned with about 10% more area than needed. $700 1037 This is not measured but it comes from storage and the temperature should be pretty steady. $800 1038 Dry reciprocating; should supply needed vacuum. 1039 Flowrate of product increases to 0.08 L/s and is steady but still is not up to the expected 0.15 L/s 1041 Before: screen 10 min; wash 3 min. Now: screen 10 min; wash 3 min. unchanged. $30

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1042 No improvement. 1043 P&ID supplied; P&ID is less than a year old. Simulation report shows that unit was operating very close to design values for usual range of feedstocks and conditions. $30 1045 Should produce target amount of dry sludge and target filtrate of 100 ppm. $400 1046 Flow is constant and at design value. temperature going to the chiller unit has Temp= 100 C. We are not pleased that the temperature of the stream that we are getting from your chiller unit is 10 to 11 C. Do something on your chiller unit or we are going to have to shut down! $500 1048 325 C. $3000 1049 Responds to change. $200 1050 Slightly less. $200 1051 Thermowell location and “cleanliness” looks OK; vortex breaker is clean and in place. Level sensor looks OK. $25 000 1052 Higher than usual 0.522 MPa and increasing. $50 1053 Hydrogen has relatively high heat capacity; about 10 kJ/kg K compared with about 2 for ammonia, methane or nitrogen; hydrogen thermal conductivity is about 0.2 W/m K compared with about 0.02 to 0.03 for ammonia, methane and nitrogen. The Pr numbers are similar. Hence, the thermal properties of a mixture of hydrogen and others vapors depends on the gas composition. $400 1055 No change. $3000 1056 Yes. $70 1058 NPSH supplied = 3.5 m; end suction centrifugal; impeller one size smaller than housing in anticipation of future expansion. Pressure allowance for control valve. Check valve but no pressure gauge on exit line. $50 1059 Respond to change. $500 1060 Moderate, 20 C; cloudy. This past week has been moderate with temperatures 18 to 22 C; rain two days ago. $20 1061 Should do the job. 1062 Estimate agrees with Dp. $2000 1063 None. $50 1065 Hot 25 C; sunny. Similar all week; we have had a hot and dry spell. All in the turnaround complained. $20 1067 Estimate seems OK. No major blockage. $1200 1068 Open. $500 1069 As expected. No improvement in the mass balance. $8000 1070 0.75 Mg/m3. $1000 1071 Responds to change. 1072 6 kW. $100 1073 Valve A fully open. No change in temperature. $6000 1074 Set at 20% open; usual value and steady. $600 1075 Density 1.478 –10% Mg/m3. $10 000 1076 221 C. $200 1077 Steady and usual. $50

471

472

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1078 Strengths of cast samples within specs. Particle-size distribution within specs. Concentration of components within specs. 1079 Preheater 1: 175 kW; preheater 2: 120 kW; Stage 1: 1100 kW; stage 2: 1080 kW; stage 3: 930 kW. $1000 1080 Two days ago. Before the turnaround, the South loop was constructed first; was commissioned and performed well at design rates. The North loop has just been completed, the reactor catalyst has been reduced and is now ready for operation. The tie-ins have been made to connect the two loops to allow for great operating flexibility. The combined “two-loop” plant has operated for two days during which time every time the flows were increased beyond 80% design capacity, the swinging-loop behavior occurs. $800 1081 Yes. $50 1082 Today; warm and humid; past week has been torrential rain. $50 1083 Equipment should do the job without rat holing or bridging. 1084 0.5 MPa. $150 1085 After 1 hour the temperature is 335 C; up 10 C. $25 000 1086 For the new reactor conditions there will be less energy available for export via exchangers E202 and E203. $1000 1087 Responds. $400 1088 Agrees; seems consistent. $200 1089 No change in operation. Pressures and temperatures still increasing. $250 1091 Good flowrate; inlet temperature 15 C. $200 1092 TC-3 = 205.9 C. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 1093 Some improvement. Now can control –4 C. $50 000 1094 No change. Still cannot control. $10 000 1095 Currently at full rate. 1096 Flows required for the reaction are consistent with design values. $50 1097 When I opened the valve, I would have expected to hear a noise of air coming in and, with the light on, when I looked through the viewport I could see the surface of the oil clearly and I could see the discharge end of the Fuller’s earth conveyor. No solids or powder were coming out of the pipe. I didn’t hear any noise. $ 4000 1099 No change. $50 1100 Goes into the top. $120 1101 Not available. They were put in so long ago no one remembers. $600 1102 dPI slightly higher than usual. $50 1103 Not needed. $5000 1104 OK. recalibration not necessary. $3000 1105 No improvement. 1106 When: 4 months ago: Checked over the control system and fans. Little adjustments or changes made; Pitch on the blades seems to change when set point changed. $50 1107 Slight variation: 0.035, 0.041; 0.039%. $3000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1108 IS: over the past few months PIC set point needs to increase. Low-temperature trip when full load of ethylene fed to the unit. Outlet ethylene temperature has been slow to recover from change in flowrates. IS NOT: PIC set at 0.8 MPa; ethylene temperature quickly recovers from change in flow rate. $50 1109 Fully open. $110 1110 Thermodynamic traps on all three locations: reboiler, turbine and preheater. No bypass supplied; upstream strainer. $300 1111 Sensors wrong, T3/ sensor fault F2/ heat-transfer area too small/ heat-transfer area fouled/ poor tuning of controller/ air flow too small to support combustion/ flameout/ gas velocity on outside too small/ excess air cooling the radiant and convection sections/ damper failed closed/ liquid-fluid velocity on inside too small/ decrease in thermal properties of process fluids/ process fluid flow increased. 1112 Allowed 130 kPa Dp for the control valve. NPSH required = 2 m water. Supplied 4.5 m organic minimum. $200 1113 Set at 10% open; usual value and steady. $600 1114 No feed, crystals or water was observed to flow from the screen to the centrifuge during the wash cycle for the screen. $700 1116 See Chapter 3: pumps, Section 3.2.3; reactor Section 3.6.3; exchangers. Section 3.3.3. $300 1117 74.3%. $3000 1119 Basics are OK but it is missing key details. There are block valves, a drain and bypass with control manual valve for all control valves: CW, steam, Level. There are local temperature and pressure sensors on the top of the column. The temperature is shown in the control room. Vortex breakers are specified for the reflux drum. The condensed liquid flows into a nozzle at the top of the reflux drum. The vent-break line includes a small orifice plate (instead of a valve to adjust the pressure balance). The valves are CW: Fail Open, FO; steam: FC; Level, FC. The steam trap on the reboiler is a float trap. $120 1120 Product sump has holes; liquid weeps from the sump. The trays are corroded with the hole size increased in all of the trays. One tray collapsed. $3000 1122 Should do the job. Enough flexibility in operating conditions that the bottoms concentration of < 2% organic should be met and should provide negligible organic overhead into the steam-ejector system. Adjustable parameters include the boilup rate, the vacuum, the live steam, the reflux water rate. Furthermore, we can recirculate the overhead product back as feed. This plant is very flexible. 1123 Heavy fouling. $10 000 1124 Well designed based on pressure, residence time plus vortex breaker and demister pad. $150 1125 None seem to have been prepared for this new addition. Someone assumed it would just be like the “old” unit. But it isn’t! $200 1126 Product breaks. $1000 1127 Checked that the height in the seal pot that was used by maintenance = that calculated for kerosene using handbook value for density. $400

473

474

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1128 Stage 1: 123 C steam; 109 to 115 C glycerine inside tube which is < 25 C; it is likely nucleate boiling; similarly 114–(98 to 108) C in the second stage; 103– (80–60) C in the third stage. $140 1129 No upsets. Everything is working normally. $300 1130 Last maintenance 3 months ago; no planned maintenance for two months. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 1131 Two months ago in May. Startup of new plant. $75 1132 IS: Control engineers. IS NOT: others. $150 1133 Wide open. $100 1134 The diagram seems to be a good representation of what is on-site. No surprises. Around all the control valves are isolation block valves, drain, bypass with valve. The steam drum has a level sensor with low and high level alarms. 1135 Vertical depth of the tube bank= 1.2 m. Top of syphon to top of the tube bank= 0.1 m. Bottom of the tube bank to the top of the overhead receiver = 1.2 m. $3200 1136 The ethylene vapor from the ethylene head tank goes to a downstream compressor as part of the ethylene chiller system. 1137 1800 nominal rpm, end suction; shrouded impeller; direct drive. NPSH required for design condition= 2.5 m. $60 1138 Gurgling noise consistent with condensate-steam flow. $500 1139 FC-1 looks OK; F-2, tab markings obscure; F-3, looks OK. If didn’t put on safepark as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 1140 Temp of –5 C and pressure 500 kPa g; consistent with sat. propylene vapor; suggests instruments OK and no contamination. $75 1142 Cavitation, vapor lock on suction, excessive pressure loss on suction side. $200 1143 Rpm minutes = 300. 1145 Seem consistent with expected gradients and responses. $400 1147 Butane outside the tubes should be film boiling because the temperature difference for steam–butane is > 50 C. $50 1148 This is first time startup. $50 1149 43 C and usually 45 C. $120 1150 Loud crackling noise, like cavitation, when low acid additions are needed. $500 1152 Gradually increasing to 52 C from the usual 47 C. $50 1153 No change and poorer operation in the long run. $30 000 1154 Yes, we’ve had occasional upsets that would have released stock into the water. “But, you folks measure that and can handle that!” These have been going on for about 20 days. $300 1155 55 C. $500 1156 Steady over 10-minute period at 110 C. $50 1157 IS: Exit reactant feed temperature. IS NOT: elsewhere. $150 1158 Detailed report of the startup of the South loop only. The plant worked well; met all specifications. Everything worked at the design rate. At full design

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1159 1160 1161 1162 1164 1166 1167 1169 1171 1172 1173 1175 1177 1178 1179 1180 1181

1183 1185 1186 1187 1189

1190 1191 1193

rate, the conditions for the South loop were: lowest pressure in loop: 2.75 MPa g; temperature at the exit of the cooling water exchanger = 30 C; at the exit of the refrigeration exchanger = 10 C; loop-gas analysis at the inlet to the reactor: H2 = 62%; N2 = 21%; methane = 13%; ammonia = 4%; discharge temperature of the fourth and recycle stages of the compressor = 40 C; valve A= 20% open; valve B = 10% open; inlet temperature to catalyst bed in reactor = 505 C; exit temperature from catalyst bed in reactor = 550 C; Dp across catalyst bed = 76 kPa g. A second internal report recommends linking the two loops to provide flexibility in operation and provides the economics benefits of such operation. $2000 Product breaks. $1000 Correct. $30 Partially plugged with construction garbage: gloves, bolts and gaskets. Confirm the temps and pressures at the bottom and top of column are consistent with the compositions expected. $100 Sporadic fluctuation; no observable cycle time. Boiling temperature of IPA= 88.4 C. $800 Observed high frequency (and large fluctuations) suggest a controller problem. $50 No improvement. $100 Estimate suggests no blockage. $400 Feed concentration the same 50% w/w IPA. $8000 0.525 MPa and increasing. $500 Responds to change. No noise. Valve not open. No evidence from flare that the pressure relief has opened. $300 For the given areas, the fouling coefficient allowances and the conditions shown in the given information, the three stages should work fine. $600 Confirm unsatisfactory yields of IPA. $1200 Valve stem responds as expected. $240 Case 1: 8:28 am: South loop temp 563 C; closed valve A for South loop by 8:29 am; 8:33 am; hot-spot temperature 555 C and swinging seems to stop. Case 2: 4:32 pm: North loop; temp 562 C; closed valve A for North loop by 4:33 pm; 4:35 pm hot-spot temperature 572 C and swinging just gets worse. $80 000 221 C. $150 The setting on PIC/10 should be such that the column pressure is 1.7 MPa but does not exceed 1.8 MPa, the relief pressure for PSV 1 $50 DHf aq, 500 = –338 kJ/mol. $300 Should be about half-load to be consistent with the evidence. Clean, impeller as expected, bearings OK. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 No leaks. $80 000 Reynolds number = 8 104; turbulent. $50 6 kW. $100

475

476

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1194 With the centrifuge cycle at 10 min but with the increased volume of wash water for the centrifuge, the crystals leaving the dryer have moisture content averages at 3.5%. $ 2000 1195 170 and 171; a few minutes later 201 and 202, respectively. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 1196 Not needed. $500 1199 A centrifugal pump should handle this condition but perhaps a reciprocating pump would better serve this application. Check that the pump is located close to the exit of the thickener and ensure you do not have NPSH problems. $75 1200 No change. $2000 1201 No evidence other than the log. Seemed to follow standard procedure. The level in the poly was correct when I came on shift. $150. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 1202 Slightly more C3 in this sample than usual. $2000 1203 Design should handle usual flowrate and 90% reduction of target influent suspended solids. No coagulant used. $500 1204 None done. This is just startup. 1205 Should do the job. Feed temperature varies from room temperature to 100 C. 1206 The steam-ejector system consists of a booster ejector, and two ejectors with barometric condensers. The steam for all ejectors comes off the top of the steam main. The cooling water for the condensers and the deodorizer comes from the cooling tower. There are eight, 25-ton deodorizers inside this part of the plant and on the second floor. The Fuller’s earth is in a hopper in an adjacent building. The conveying line has very few bends; all bends have a large radius of curvature. 1207 OK; recalibration not necessary. $10 000 1208 Set at 10% open; usual value and steady. $600 1209 Improved control but still the same type of general oscillations. $35 000 1210 No evidence of fouling. $3000 1211 If there was no tube leak, water would dribble out. If there is a tube leak, fuel gas will gush out. A vapor smelling like fuel gas gushed out. $2000 1212 New installation. Carefully pressure tested, cleaned out, inspected before startup. The rest of the system was checked over, adjusted as needed before the new system was started up. $50 1213 Responds to change. 1214 Top temp –100 C usual; bottom temp same as usual. $200 1216 Diagram is schematically correct. Around all control valves are block valves, drain and bypass with valve. The reactor design uses cal rod heaters. Thermocouples in the catalyst bed and in the feed gas entering the bed. Temperature sensors at four levels in the catalyst bed. The catalyst bed, the gas preheater and the water cooling exchanger are integrated into one, high pressure vessel.

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1217 1218 1219 1220 1222 1223

1224 1225

1227 1228

1230

1235 1237 1238 1239 1240 1244 1245 1246 1247 1249 1250

1252 1253

1254

Level indicators on the gas–liquid separators are used to control the flow of liquid. $10 000 TI-3 decreases to 48 C promptly; the flowrate FIC/4 increases rapidly to usual value. $600 No improvement. $25 000 Clean. $2000 1.5 MPa g steady. $700 IS: at the hub at the weld line. $60 Design capacity with allowance for pressure drop through the chiller on the tube side and the elevation and frictional losses in the fittings and piping. Should do the job. Able to bypass the pump if the pressure in the feed drum is sufficient for the desired flowrate. Should do the job. $45 FC-5 reads the usual value and it reads the same value just before the system was put on “safe-park”. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Clean; no improvement. $200 I brought this on-line following the procedures. However, shortly thereafter it started cycling like mad. I was keeping the feed steady. The feed composition is on spec. But the product isn’t! That’s when I phoned you to sort this out. $120 Small amount of scale removed. When operation resumed, negligible difference in system behavior. Still getting off-spec material on some hot days. $4000 Should handle new conditions easily. $900 Total area = 100 tubes 2.83 cm2. Assume tube velocity to minimize fouling = 1 m/s. Estimate of total water flow= 28 L/s. $100 OK. recalibration not necessary. $3000 OK. Recalibration not necessary. $3000 CH4: 16%; C4 s: 4%; C2H6: 12%; C5 +: 3%; C3H7: 6%; C3H8: 7%; N2: 48%; H2S: 3%. $12000 Thermodynamic trap; upstream strainer. No bypass supplied. $300 P&ID supplied; the data are being generated now during the commissioning. $30 Output is 100%; fully opened for a fail-safe valve, FC. $100 Usual. $100 Midrange and steady. $120 None on the gas feed lines from the ammonia and urea plants; steam lines hooked up to the heater in the storage tank (with condensate going to drain); demineralized water to lantern ring-packing seal in transfer pump. $300 49 C. $400 Pressure is slightly less than usual. On the head-capacity curve suggesting a higher than usual flowrate. However, the pressure is gradually increasing. $300 Some fouling. Water and air pressure tests show no leaks. $3000

477

478

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1255 Nothing unexpected because we had not increased process liquid flowrate. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000. 1256 OK, recalibration not necessary. 1257 Slight decrease in response, more sluggish and slight decrease in ability to evaporate ethylene. $500 1258 No interruptions. Pressure should be 550 kPa g. $50 1259 Just before we started up, changes only made to the condenser and associated piping. $650 1260 No diagram supplied. Blends are stored inside in vertical hoppers, star feeders at the bottom leading to batchwise, dense-phase conveyors that bring the different blends to blender. The lines are gently curved. The mixer agrees with the specs. The blended product is conveyed by dilute phase, pressure conveying to a reverse jet bag filter located above the hopper for the bagging machine. The conveyed material enters the bottom side of the filter, rises up the inside of the bags and the “clean” air goes out through the bags and exhausts to atmosphere. The equipment agrees with specs. The bagging machine agrees with specs. No general surprises. The area is surprisingly clean and dust free. 1261 This is a fail-open valve. The signal is 0. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 1263 Accurate; no recalibration needed. $3000 1264 23 C; cloudy and overcast; rain forecast. It sprinkled over the last couple of days. $300 1266 Vent-break line already in place. $50 1267 Boiling temperature 117.9 C; pKa = 4.76; heat of vaporization 405 kJ/kg. Acetic acid vapor partially dimerizes (so that the molar mass varies between 60 and 120). 1268 Frequency: once every 24 hours; sporadic; major changes within 5 minutes of when hot spot first appears; takes about 1 to 3 hour to settled things out assuming “all the right adjustments have been made by the operators”. $600 1270 Pumped to the exchanger. $150 1271 Slightly over sized. The liquid is subcooled so that there should not be any NPSH problems. The role is to remove liquid product so that liquid does not back up and overflow into the vacuum system. $150 1273 No noise sounding like it is passing. $200 1274 This is first time startup. $50 1275 High inlet temperature and humidity; Exit temperature and humidity as expected. $4000 1276 Was OK. $2000 1278 Steady pressure at normal values; amount of superheat about 3 C. $50 1279 Damper appears to be 1/3 closed. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1281 1282 1284 1285 1286 1287 1289 1290 1291 1293 1295 1297 1298 1299 1300 1301 1302 1303 1304

1305

1306 1307 1308 1309 1310

1311

Nothing unexpected. $200 000 Not needed. Valve stem moves to open valve. There is 100% valve travel. $3000 Lumps persist. $2500 No changes. Steam on the drum is steady, and we are supplying steam with a slight superheat, as usual. $300 Balances –10%. $50 Filled with sludge. $300 Flowrate relatively constant; the pH is jumping all over the place! $100 Pulses of excessive gas loading on the vent scrubber. $1200 Same as we supplied you with over the past ten years. $250 Both have equal activity. $80 000 Sized to do the job. 10 L/s at head of 10 m ( = 100 kPa). $30 Not needed. $6000 Slightly more than usual. $75 No improvement. $3000 Signal is 0%. This is a fail-close valve so the valve should be wide open. $300 –25 –2 C Same as TC/5 within –1 C. $500 The diagram seems to be a good representation. In addition, there is a control system. The steam trap is an inverted bucket with an upstream strainer; isolation valves, no bypass. The barometric leg is submerged in a “hot well”. Vacuum system consists of a booster ejector with two ejectors with two, direct-contact interstage barometric condensers. These condensers feed the hot well via barometric legs. All the steam lines come off the top of the steam main. $200 Reciprocating pump. NPSH needed << calculated NPSH supplied; steam traced suction line (with 350 kPa g steam) to prevent freezing in suction line; upstream vortex breaker in sump on tray 4; design pumping rate 0.07 to 0.15 L/s. The diagram does not correctly show that there is a strainer in the suction line to 114-J. An analysis of the pressure profile from the tray 4 to storage shows that the product should flow by gravity to the storage tank without the need for pump 114-J. 128 C. The usual value is 130 to 140 C. $120 Responds to change. $100 Product breaks. $1100 605 C. $1200 The feed concentration seemed to be about 200 ppm but because the torque has increased I increased the sludge pump rate as given in Standard Procedures. The supernatant was still around 120 ppm and the belt filter was flooded. $120 Not enough steam/ wash water carryover from the centrifuge/ cycle from screen not coordinated with cycle in centrifuge/ feed crystals too wet from the screen/ rotational speed of the dryer has increased/ more fines in the centrifuge causing the filter cycle to be too long; fines carryover to the centrifuge causing blinding in the centrifuge/ centrifuge rpm faster than usual/ crystal

479

480

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1312 1313 1314 1315 1316 1317

1318

1319 1320 1321

1322 1323 1325

1326 1327 1328 1329 1330 1331 1332 1333 1334

1335

size change; crystals change shape so that filtering is different/ centrifuge operates but has run out of feed/ dryer feed too cold/ vendor supplied faulty equipment Raw material: particle-size distribution within specs. Slightly more than usual. $150 Fcondensed. 200 kJ/kg = 200 Fcondensed. Not needed. $4000 Pressure negligible. $1000 The pressures and temperatures are rising and we can’t seem to control these. The pump F1400 is knocking its head off but doesn’t seem to be pumping anything out. I wish I wasn’t on this shift! Sampling error/ analysis error/ hazard coming from upstream plants/ leaks in vent valves from any location having “fuel gas” type materials upstream of the valve, e.g., the shell side of most exchangers/ leak in the valve on the knockout pot/ leak in the tubes from fuel gas to water in E131. IS: much less than expected and not in production. IS NOT: 500 C at noon and steady production. $3000 IS: after 3 hours since startup. IS NOT: immediately at startup. $50 Why1? to ensure safe operation; Why2? to provide steady, reliable operation; Why3? to meet production requirements; Why4? to sell products and keep company financially viable; Conclude: focus on Why1? “To ensure safe operation.” $50 Good flowrate. $250 No change. Flowrate and temperature should be the usual values. The temperature should be about 18 C. $200. See Chapter 3: reactors, Sections 3.36, 3.3.7; mixers, 3.7.1; control and valves, Section 3.1.3 $200. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 See Chapter 3: sensors and controls, Section 3.1.3; pumps, Section 3.2.3. $200 Not needed. Steady flows and temperatures expected for this time of year. Since this is spring, the temperatures are lower than usual. $60 Everything started fine. We have not changed anything. $50 I think early on today’s shift, but now there is no flow. $20 Not needed. $5000 IS: in neutralizer. IS NOT: reported elsewhere as yet. $50 Recalibration not necessary. Yes, I have noticed that AC/1 shows variation suggesting variable C3 in the debut overheads or. maybe the analyzer is just jumping around. But that system controls itself – so I don’t worry. $50 Why1? to provide steady, reliable operation; Why2? to meet production requirements; Why3? to sell products and keep company financially viable; Conclude: focus on goal “To keep bottom level steady and stop tray-temperature cycling.”

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1336 Diagram is reasonably complete. Additions include a strainer just upstream on the suction side of the pump 114-J. The suction line also a bypass of pump 114-J. Block valves are also on the exit of pump 115-J to direct the flow to drum disposal or to product storage. A drain valve is present beside the feed control valve FIC /101 (which is incorrectly shown as PI/101). The control valve on the reflux water has block valves, drain valve and bypass with a valve. The cooling water on the reflux water comes from a common header with the CW to the barometric condensers and the water goes to the hot well. Steam is generated on-site at 3.5 MPa g, brought to the battery limits and pressure reduced to 1.38 MPa g for use within the battery limits. A temperature sensor is present on the steam to the reboiler. A temperature sensor is on the vapor from the reboiler into the column. A pressure sensor is at the bottom of the column with a separate tap for the differential pressure. There is a level indicator at the bottom of the column. All process steam comes off the top of the steam headers. The pipe rack is 3.6 m above grade. Just before the product storage tank is a “bulls eye” in the line that allows one to see the flow to storage. There is a sight glass at tray 4 that includes the vertical distance including the product sump and the seal pan for the downcomer from tray 4. The liquid from tray 4 overflows the weir on the downcomer seal into the product sump. The weir on the product sump overflows onto tray 5. 1337 Responds to change. $200 1338 Lost. Cannot be located. $500 1340 Sample 1: 13%; sample 2: 10%; sample 3: 7%; sample 4: 3%; sample 5: 8%. $20 000 1342 Diagram reasonable. It is missing the isolation block valve on the pump suction. Pressure gauge has a pigtail with a shutoff valve with a tee nipple and valve installed between the gauge and the pigtail to allow backflushing of pigtail. $40 1343 Moist and humid today: a continuation of the weather over the last week. $60 1345 Not needed. $10 000 1347 Alarm sounds indicating high pressure. $3000 1349 Oscillating corresponding to the recorded output. $120 1350 The obvious ones: the cooling water in and out; steam out the top. There are water lines available to clear sample lines but these are not directly attached to the neutralizer. $300 1351 During the turnaround. Previously about four months before the turnaround, some routine maintenance was done on the naphtha feed pump and gas feed line to the furnace. $300 1352 Samples showed the propane concentration in the fuel gas was about 1.5 times higher than usual but this could be simply sampling and analytical error. $100 1353 Extensive manuals including maintenance schedule and table of troubleshooting diagnostics. $800 1354 Should do the job. Steam pressure to ejectors is 1.38 MPa g. 1355 Fcondensed. 2. (120–60) = 120 Fcondensed.

481

482

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1356 Confirm unsatisfactory yields of IPA. $1200 1357 Usual concentration of butane. $50 1359 Five months ago; routine cleaning of exchangers, checking and calibrating instruments; usual maintenance and adjustments of the ethylene feed pump. $50 1360 Why1? to be able to sell propane not fuel gas; Why2? to keep the company profitable; Why3? to pay my salary; Conclude: focus on goal “To reduce the propane loss to fuel gas.” $50 1362 Temperature quickly drops from 325 C. $10 000 1363 See Chapter 3: sensors and control, Section 3.1.3; evaporators, Sections 3.3.3 and Section 3.4.1; reactor, homogeneous (no trouble shooting suggestions available); condensers, Section 3.3.3; vacuum pumps, Section 3.2.2. 1364 Correct residence time for gas-liquid disengagement and residence time for control. Demister pad and vortex breaker included. $50 1365 Tricky. The leak goes either way depending on the PIC setting. For a high PIC setting the leak would be butane into the condensate; for a low PIC setting the steam would leak into the butane. There is no butane detector on the condensate. $50 1366 Why1? to provide steady, reliable operation; Why2? to meet production requirements; Why3? to sell products and keep company financially viable; Conclude: focus on goal “To meet specs on overhead and bottoms.” $50 1367 Lumps persist. Time for injection increases; overall cycle increases. $3000 1368 Yes. $2000 1369 I can’t seem to control anything! I’ve tried putting it on manual but it is so sensitive and I seem to get what I get no matter what! That’s why I called you. $120 1371 Pressure and overhead temperature reduces to design. Pressure control still sluggish but it is in the design range. $12 000 1372 Equipment should do the job. Simultaneous fill and weigh. 1374 Steady and higher than usual, 50 C. $50 1375 Clear. No obstruction. $7000 1376 No. 1377 Top temp –40 C usual; bottom temp higher than as usual. $200 1378 Steady, 1.8 –0.1 kg/s. $60 1379 The oscillations began small and increased in amplitude until it met sensor range. Problems related to valve stiction and hysteresis more likely to lead to continuous cycling at smaller amplitude. $50 1381 Head-capacity curve available. NPSH requirements available. Operating at 1800 rpm nominal. No pressure gauge on exit line so cannot check. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 1382 Allowance for fouling on tube side = 0.00001 m2 K/W; shell-side steam = 0.0006. $250 1384 2:01 am South reactor: hot spot noted 580 C; 2:02 Production in South loop increasing; 2:02 North reactor “drying” with temp falling below 500 C; 2:03

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1385 1387 1388 1389

1390 1391 1393 1395 1397 1399 1400

1401

1402 1403 1405 1407 1408

1409 1410

am. Startup heater put ON for North loop reactor; valve A on North loop OPEN to decrease gas flow and help reheat North loop; 2:04 am hydrogen analyzer on South loop oscillating and showing a slow increase in hydrogen; 2:05 am production in North loop decreasing; hydrogen in North loop oscillating and showing a gradual decrease in hydrogen; 2:06 am valve A on the South loop CLOSED. 2:07 am South reactor temperature 590 C; hydrogen and production increasing on South loop; 2:20 am North reactor temperature 500 C; 2:22 am startup heater on North loop OFF; valve A on North loop closed to 20% open setting; 2:30 am South loop reactor temperature 560 C; hydrogen and production in South loop decreasing toward design values; 2:45 am South temp 555 and North temp 545 C; 3:05 am both and North and South reactors 550 C exit temperature. No upsets or unexpected behavior over next two hours. $40 000 Same before and after change. $1200 No evidence of solidified fatty acid. $3000 Improvement, extension of the period before break but still not satisfactory. $12 000 Last turnaround three months ago. Process has been operating “fine” since then. If didn’t put on safe-park as first step, then dangerous potential fire/ explosion conditions are created while you experiment. $500 000 Insufficient liquid in column. $5000 No improvement. $25 000 Usual value. Last summer; a compressor installed to recycle flare gas as fuel gas. New demister pad installed in the knockout pot. $300 –59 –2 C No change; gets worse. $80 000 None done recently. four years ago the rpm of the pumps was changed from 1800 rpm to 3600 rpm (and the motors changed accordingly from 1 kW to 10 kW) to account for pressure for increased production. $15 Files missing. Engineers say that the reactor was overdesigned and should handle the increased feedrate “No problem!” If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 No sounds of cavitation. $200 1.6 m/s, reasonable. $50 Should give the correct % vaporization per pass. $700 IS: On the debut. IS NOT: upstream on the deprop. No problems stated on reactor $50 This has been a bear to operate. The overhead temperature has been increasing gradually and I have been increasing the reflux accordingly but the temperature keeps increasing. Yet the analyzer is showing the usual overhead composition. $50 Steady and usual design value. $700 None

483

484

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1411 We follow the new operating procedures that affect only the centrifuge. Wash time remains the same; volume of wash water increased by 35%. Filtering time reduced by 1/3 to= wash time. $30 1412 75.3%. $3000 1413 Reads 438 –3 C 1414 510 increasing to 550 C at exit of bed; usually consistent and “reliable”. $600 1415 Warm 20 C; similar temperature and slight wind all week. $15 1416 IS: nothing happened; IS NOT: Fuller’s earth seen flowing into deodorizer. 1417 Operation went smoothly. No problems. No different from a month ago. 1419 227 C. $200 1420 High liquid level on tray 4 and in the product sump. $5000 1421 Not for 8 months. $50 1422 The temperatures recorded in the control room indicate that the feed temperatures are “normal”. The exit “cooling water” temperatures are extremely high. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 1423 3 to 7.5 m/s as expected from the vendor data. $3000 1424 IS: at startup and during the first week. IS NOT: corrected after one week with focus on control. $150 1425 Based on the fluid velocity of 1.6 m/s and a distance between injection and the sensor of 2.4 m= 1.5 s. $50 1426 Agrees; seems consistent. $200 1427 Fluctuates. $200 1428 Not needed. Depends on the safety officer and magnitude of the values. $2000 1429 Half-load: inside h= (0.5)0.8 hfull = 0.57 hfull ; outside: heat flux (half-load) >> Heat flux (full load) because of shift to nucleate boiling for half-load. 1430 See Chapter 3: distillation, Section 3.4.2, perhaps pertinent condensers and reboilers, Section 3.3.3; pumps, Section 3.2.3 and controllers 3.1.1. $50 1431 Coke on the outside of the tubes and on the inside furnace walls. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 1432 Estimate seems OK. No major blockage. $100 1433 Well designed; should do the job. Use hydrogen fuel. Thermal efficiency= 32%. 1434 Consistent with overhead temperature and the overhead specs; relatively steady since startup, about 230 kPa g. $200 1436 Decreasing. $50 1437 Respond to change. $500 1440 Seven months ago. Routine calibration of all instruments; changed the seal on the transfer pump and adjusted clearances; cleaned the outside and inside of the cooling coils. Inspected neutralizer for corrosion. $150 1442 Line clear; no plugs. $35 000 1443 No improvement in process. 1444 Why1? to save lives and equipment, keep insurance down, good operations. Why2? to stay in business. Why3? to pay my salary and give me continued

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1445 1446 1447 1448 1449 1450 1451

1452

1453 1454 1455 1456 1458

1459 1460 1461 1462

1464 1465 1466 1467

peace of mind. Conclude: focus on Why1?: “to save lives and equipment, keep insurance down, good operations.” If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Aluminum with two feed gates. Two cooling lines each side. $150 Midrange and steady. $120 Dark black with yellowish tinge. $1000 Should work well. Baffles were in vertical. Vessel has air bleeds. Before you started up you would have opened the bleed to ensure all the air was out. $200 This plant has been very steady. No changes. Variable speed; air cooled; forced draft. No trim cooler. Good design of hydraulics with vent break (not shown on diagram). 20% excess area. $50 Calculated height of kerosene in the seal pot was checked to ensure that this was consistent with a pressure differential of 6.8 kPa. No visual or level transmitter was installed that allows us to verify the height. $300 Everything OK until temperature TC/3 started to droop. Looked at the flare. It’s black. I went on the plant but couldn’t see anything obvious. My first reaction is to reduce the process flow but called you before I reduced flowrate. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Our lab: 0.028; Excell Lab: 0.030; Univlab: 0.026. $20 000 Slightly more than usual and consistent with the increased pressure. $75 Well designed based on pressure, residence time plus vortex breaker and demister pad. Complete with syphon break, vents and drains. $150 Should handle range of flows for design conditions. $500 Feedrate increases slightly with time averaging about 65 during one cycle and then averaging about 78 during the next cycle and then fluctuating in the range 60 to 85 depending on when the analysis is done. Before the change it was about 68 during any cycle. $2000 7 MPa; usual value for startup. $3000 No change. $9000 All seems to be working properly. $30 Diagram shows general layout. Both control valves have isolation block valves, drain and bypass with valve. There is no pressure gauge on the reflux pump before the check valve on the discharge line. The condenser has four rows of finned tubes. The reflux drum has a demister and vortex breaker. There is a vent break on the exit liquid header of the condenser to “seal” the tubes. There is no pressure gauge on the overhead receiver. $100 OK; head-capacity curve OK relative to estimated pressure requirement. $240 Head = 45.0 m at zero flow. $50 Not pertinent, just completed a turnaround. $50 See Chapter 3: reactor, Section 3.6.2; compressor, Section 3.2.1; sensors, valves and control Section 3.1.3; knockout pot, Section 3.5.1; exchangers, Section 3.3.3; refrigeration cycle, Section 3.3.4. $500

485

486

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1470 Everything works fine as long as we can depend on pump B operating well. We really are in big trouble if anything happens to B and we are expected to use A. As it now stands, we don’t have a backup! $15 1472 Clean; negligible corrosion. No apparent blockages in sparger holes. $30 000 1473 Internals look normal; clearance = normal; wear rings = good shape; negligible erosion; key in place. No pluggage. $4000 1474 Head-capacity curve suggests the increased naphtha flowrate can be handled easily. $700 1475 Should work well. Baffles were in vertical. Vessel has air bleeds. $200 1476 Fluctuates. $200 1477 IS: reported variation in concentration of propane in deprop bottoms > specs. IS NOT: everything else is running smoothly 1478 Well designed based on pressure and residence time. $150 1479 Product continues to swell after it is ejected from the mold; resulting product is mis-shapen. $900 1480 Very fine particles, < 150 mm, free flowing with angle of repose 15 to 30, mildly abrasive with Moh’s hardness 2–3. Harmful by inhalation. Contains silica, crystalline cristobalite. Irritating to eyes and respiratory system. $500 1481 IS: 7 hours after startup and after an upset on the upstream deprop that sent too much C3 to the debut; IS NOT no information known because this is startup. $50 1482 Flow is the control value of 14 L/s. $50 1483 Two months ago. $60 1484 When: 6 months ago; routine checks done during the turnaround. $400 1485 41% ammonia; 29% water vapor, 30% CO2 all –3%. $2100 1486 Reciprocating pump. NPSH required is specified. Details for maintenance, trouble shooting and precautions about operating with a closed valve on discharge and the importance of checking that the pressure relief from discharge to suction is working. Nominal flow 0.15 L/s but this can be decreased by setting the recycle valve. Precaution: drain the pump cylinders of all liquid when shut down to prevent liquid from freezing in the pump in cold weather. 1487 IS: T3 is lower than expected. IS NOT: changes elsewhere that have been noted 1488 Unable to sample directly. The vent line is at the highest point, namely on the inlet cooling water. The exit discharges below ground. $200 1489 Was OK. $2000 1490 Dryer should do the job $100 1491 Steady and slightly higher than usual. $50 1492 This thing cycles no matter what I try. $500 1493 Hydrates are ice-like solid clathrates with the water forming the cage around the hydrocarbon molecule. Species like methane, ethane, isobutane, propane (and butane in the presence of other hydrocarbons), plus water, can form hydrates. Hydrates form at high pressures and relatively low temperatures. For example, at 3.1 MPa hydrates form with isobutane at temperatures < 3 C; with methane, temperatures < 3 C; ethane < 14 C; propane < 6 C. Hydrate for-

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1494 1495

1496 1497

1499 1500 1501 1502 1503

1504

1505 1506 1507 1508 1509 1510 1511 1512

mation can be inhibited by alcohols (such as methanol), glycols and ionic salts (including NaCl). Hydrates do not form with very soluble gases such as ammonia and hydrogen chloride. $500 Some improvement. Now can control –9 C. $120 000 NFPA ratings: Health, Fire, Stability: Propane: 1, 4, 0; Methane: 1, 4, 0; Hydrogen: 0, 4, 0; Butane: 1, 4, 0; Pentane: 1, 4, 0; Extreme flammability hazard; avoid sparks, open flames. Hydrogen: Evacuate all personnel from affected area; do not enter areas containing flammable mixtures of hydrogen and air. Vent area to prevent form of flammable or oxygen deficient atmosphere. Eliminate all potential sources of ignition. Butane skin-contact with gas may cause frostbite (liquified gas). May cause slight eye irritation. Inhalation causes drowsiness, excitation or unconsciousness due to anesthetic and asphyxiation properties of this gas. $50 510 increasing to 550 C at exit of bed; usually consistent and “reliable”. $600 Checked that it is not on surge or sonic conditions based on the average flaregas composition given in the problem statement. Confirmed that the decrease in gas temperature, because of cold temperature, is not a factor. Realize that high molar mass in feed decreases range of operation. Surge is related to power used. $800 Installed pump and motor should handle it OK. $900 Not needed. $2000 Same as expected. $1200 Going crazy. $50 Not needed. $20 000. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. Water has relatively high heat capacity; at the low-pressure steam has a high latent heat. Ethyl acetate vapor heat capacity about 1.24 kJ/kg K compared with steam at 2 kJ/kg K. Pr = 0.8 for ethyl acetate and 1.1 for steam. If the steam was at “atmospheric pressure,” then there is not enough pressure to push the steam through the pipe and exchanger. Usually 200 kPa g is the minimum practical pressure; sat. temperature for steam= 134 C. Heat capacity for water = 4.2 kJ/kg K. $50 Cold, –28 C, dry, windy. January. The weather has been like this all week. $150 Steady and mid-vessel. $200 Closed cycle, standalone unit. Should do the job. $6000 Now: IPA in= 2250 kg; IPA out bottoms = 225 kg; IPA out overhead to further processing = 1350 kg or total of 1575 kg out. $4200 Sounds “hollow” all the way up the condenser. $600 Identical to pump A. Head capacity and NPSH data as expected. Allowance of 140 kPa for the control valve. $15 Usual amp draw under existing circumstances. $300 No change. $600

487

488

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1513 Valve position is fully open; valve is put in correct direction; valve is two sizes smaller than the line. $100 1514 Feedrate constant now; feedrate constant before the change, Both feedrates the same = about 100. $2000 1515 Steady and lower than usual at 104 C. $50 1516 Feed flow rate 531 –1 L/ s; effluent flow rate 514 –1 L/s. $2000 1517 Area and fan should be fine for your operation. Allowances were made for the hot day-time temperatures like today. Should be no problem. Should ensure, however, that the bottom series of tubes in the condenser are flooded to “seal” and prevent the vapor from passing through the tubes uncondensed. To ensure this, the liquid header should either be near the top of the tube bank or should have an exit pipe that is elevated. Naturally you will provide a vent break at the top of this seal so that the condensate can syphon out continuously. If no vent break is provided, the liquid seal is periodically broken, all the liquid syphons out and uncondensed vapor escapes. $800 1518 Head-capacity curves available. Strainer needed on the suction to remove metals and materials that might damage the impeller. Net positive suction head information given. $30 1519 Should work well. Baffles were in vertical. Vessel has air bleeds. Before you started up you would have opened the bleed to ensure all the air was out. $200 1520 Cycle 1 3 min: ’1: 2.1%; ’2: 2.6%; ’3: 3.2%; ’4: 3.7%; ’5: 4.1%; ’6: 4.6%; ’7: 5.2%; ’8: 5.7%; ’9: 5.9%. Cycles 2 and 3: 6.0%; 6.1%; 6.8%; 7.0%; 7.2%; 7.1%; 7.7%; 8.3%; 8.5%. Cycle 4: 8.5%; 8.5%; 8.0%; 9.3%; 8.7%; 8.9%; 9.0%; 9.1%; 9.2%. Cycle 5: 4.3%; 5.4%; 6.7%; 7.8%; 7.5%; 8.3%; 8.7%; 9.2%; 9.1%. No data available for previous operation. $3500 1521 Same as expected; 3/4 full and reasonably steady. $30 1522 The major mold concern was the thick wall at the hub. Two gates were placed in the mold at the hub to account for the thick walls. Several different molding machines were available on site to produce this product. A cold runner system was used. Eventually a set of production conditions were developed to produce a reliable product. $200 1524 Allowed 130 kPa Dp for the control valve. NPSH required = 2 m water. End suction. $200 1525 Liquid has filled the glass; suggesting either a plugged sight glass, faulty reading or a very high level of liquid that would “overflow both the seal pan on the downcomer from tray 4 and the product sump”. 1526 Dp across the new catalyst consistent with old catalyst. Max temperature for new catalyst comparable with max temperature of old. Procedure for loading catalyst well documented. Same catalyst support used as previously. $650 1528 Should work well provided there is no vapor lock in the feed line. Caution: never operate against a closed discharge valve. Include a bypass. $200 1529 TI/2 = –59 –2 C; downstream operator reports at design flows the temperatures are –60 to –55 C; 1530 Hot and humid all week including today. $50 1531 Not needed. $4000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1532 Furnace: sized to take the increased flowrate. Heat-exchange tubes: sized to take the increased flowrate. With oil firing the emissivity goes through a maximum in the early parts of the flame. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $50 000 1533 Float trap provides continuous discharge, is suitable for low pressures; not suitable for high pressure and not suitable where there is a chance for water hammer. Range 0.06 to 5 kg/s. At 2.7 kg/s, even our largest size trap may not be best because we recommend double the size and double your load is slightly more than the capacity for this trap. $200 1534 OK. Recalibration not necessary. $3000 1535 Thermodynamic traps on all three locations: reboiler, turbine and preheater. No bypass supplied; upstream strainer. $300 1538 UA design= heat load/LMTD = 118/ 450 = 260; UA actual = heat load/LMTD = 81/550 = 147 or 56% of design. Either the area is dramatically reduced; heattransfer coefficient on gas side is less (especially in second section) or high fouling coefficient. 1540 See Chapter 3: distillation, Section 3.4.2; pumps, Section 3.2.3; reboiler and exchanger, Section 3.3.3; vacuum system, Section 3.2.2; sensors and control, Section 3.1.3. 1541 No apparent leaks. $20 000 1542 Warm, 22 C; sunny. This nice weather has continued all week. 1543 Yes. Valve stem closes slightly. $450 1545 Moderate. $150 1546 324 C. $3000 1547 Unlikely that flashing would occur. $300 1548 Responds to change. $200 1549 No change. $300 1550 Much lower total amount exchanged than expected. Heat lost in reformer exit gas is same as that gained by stream to other units –10%. $800 1551 Steady and mid-vessel. $200 1552 Recalibration not needed, no adjustments made. $2000 1553 OK. Recalibration not necessary. $3000 1554 IS: this operator. $50 1555 Negligible fouling; no condensed liquid; baffle spacing correct and baffles secure. $6000 1556 Orifice plate in correctly. No improvement. 1557 Product breaks. $10 000 1558 Balances –10% although too high a percentage of propane is going to fuel as gas instead of forward as liquid. $50 1559 Clean; no improvement in process. $5000 1560 Unmixed without mechanical mixer. $30 000 1561 Controller output suggests negligible stiction of < 0.1%. $1600 1563 Steady and usual value. $50 1564 150 amp. Usual value. $6000

489

490

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1566 Should do the job. $2000 1569 Shell and tube. Should work well. Baffles were in vertical. Usual allowance for fouling on tube side and shell side. Correction factor for not-true countercurrent flow is > 0.85. Vent/blowdown lines on both the tube and shell side. $700 1571 IS: at drop boxes B and C. IS NOT reported elswhere on the site by the safety inspector. $30 1572 Polypropylene NFPA: 0, 0, 0. Nuisance dust. Autoignition about 357 C. Electrostatic charge can potentially build up when handling. $50 1573 Much higher in C3 than design. $50. 1576 Almost completely plugged. $150 1577 Everything OK. No evidence of leaks. $6200 1580 Pulp fibers had formed a slight plug at the end of the sampler tube. Cleared it out and carefully put the sampler line in the feed. Suspended solids 482 ppm. $14 400 1582 Yes, T8 > T2. $50 1583 IS: sporadically; IS NOT: never. $300 1584 Not needed. $8000 1585 No change. $600 1587 Both correct. $30 1589 During the turnaround, routine inspection of the column internals, checking pumps. calibrating sensors. $650 1590 1.02 Mg/m3. $1000 1592 Both seem to function OK. Are left CLOSED. 1594 Yes. The downcomer appears to be “sealed”. The water temperature seems “normal”. $60 1595 Identical to pump B. Head capacity and NPSH data as expected. Allowance of 140 kPa for the control valve. $15 1596 Why1? to prevent possible explosion. Why2? to save lives and keep equipment intact. Why3? employee’s lives and health are paramount, to prevent lawsuits, to protect equipment, to maintain reputation for safety. Conclude: focus on Why1? “to prevent possible explosion”. $50. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 1597 Pyrometer reads 228 –3 C; temperature sensor reads 227 C. $300 1598 No improvement. $18 000 1599 Mid-range. $90 1600 Design should handle separation well. Level sensors. Sufficient residence time for both gas and liquid with well-designed demister mesh pads and vortex breakers for gas and liquid, respectively. $6000 1601 The pressure is stable and sufficient to ensure good combustion in the burner. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 1602 510 increasing to 580 C at exit of bed. Usually 510 increasing to 550 C at exit of bed. Hot spot in bed! $1000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1604 Column settles out; concentrations of overhead and of bottoms are consistent with design but we cannot meet the downstream production required. $4000 1605 Column gradually settles down. All values return to normal. $600 1606 With the overhead temperature increasing I have increased the reflux to the maximum but nothing seems to happen. This is out of control! $50 1607 Pressure about 1.72 kPa g; 5 C superheat; no upsets and steam supply should be reliable and constant. 1608 Evidence of splay marks on the surface. Product still breaks. $800 1610 Decreasing. $50 1611 Not needed. $500 1612 18 C and steady; past records show this ranges all over the place from 15 to 100 C. 1613 Excessive off-specification gaseous overhead stream especially on hot dry days. $20 1614 OK. $2500 1615 Overflow effluent = 210 ppm; belt filter still flooded and cannot handle feed. $6000 1616 Sufficient area supplied for design flowrate; natural circulation; designed for mix of nucleate and film boiling on the shell side – the mechanism and flux shifts as the liquid goes through the tubes. Changes from high-flux film to lower-flux film, then transition and then to high-flux nucleate. Consideration given for the vapor space to keep vapor to exit port reasonable. 1617 No. $60 1618 As best I can tell. $150. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 1619 0%. $80 1620 We’ve had four days of operation since startup after the turnaround. It has been 3 times the South loop and once the North loop. $500 1621 Being collected now during the commissioning. No other internal reports. $50 1622 0.5% w/w added to the resin hopper; this is added especially because of the thickness of the hub 1.25 mm. $200 1623 T4 = 207.2 C. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 1624 No lumps. $3000 1625 –101 C 1626 See Chapter 3: solids conveying, Section 3.2.4; fans and blowers, Section 3.2.1; bag filters and cyclones, Section 3.5.2; mixing solids, Section 3.7.3. 1628 Steady but higher than design. $50 1630 Pressure tap line is clean and clear. $350 1631 Equipment should do the job. $100 1632 Usual range 0.06 to 5 kg/s. Should do the job. Water hammer is unlikely; usually sized on double the condensate rate. Here condensate rate is about 2.7 kg/s so it should be OK. $800 1633 Steady but much below design. $50

491

492

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1634 Responds to change. $100 1635 The pressure should provide a head of about 47m of sludge; the static height it blows against is about 15 m. Therefore there should be sufficient air pressure to overcome the head and the friction to blow out the line. $50 1636 Should work well. Baffles were in vertical. Vessel has air bleeds. Sealing strips used on baffles. $500 1638 Clear, no plugs. $6000 1642 Mild, 18 C; sunny; predicted rain; Spring time. Mixture of sunny and overcast weather over the past week. $50 1643 Exit overhead is 0.65 kg/s which produces, according to flash drum, is 0.43 overhead and 0.24 liquid underflow or 0.67 kg/s. This is within the accuracy of the measurements. $600 1644 Not out of alignment. Alignment OK. No change. Level in flocculation tank still below normal. $2000 1647 Column settles out; concentrations of overhead and of bottoms are consistent with design but we cannot meet the downstream production required. $4000 1648 Calculated power consistent with power draw for “no flow”. $120 1649 No fouling. $5000 1650 350 kPa g steam; sat. temp. = 148 C; 1.48 MPa g; sat. temp= 198 C. 30 C = 4.2 kPa abs 1651 Thermowell location and “cleanliness” looks OK; downcomer is sealed; tray is level. $20 000. 1652 Clean; new looking. Water and air pressure tests show no leaks. $20 000 1653 See Chapter 3: distillation, Section 3.4.2, perhaps pertinent condensers and reboilers, Section 3.3.3; pumps, Section 3.2.3 and controllers 3.1.1. $50 1654 IS: the acetic acid-acetic anhydride system. 1655 Partly closed. $300 1656 Hot, sunny. 28 C. This week has been one deluge of water after another. I didn’t think it could rain that much! Today, however, it has cleared up to reward us with a nice sunny day. $50 1657 Close valve A to increase the gas flow and sweep the hot spot out of the catalyst bed. $300 1658 Same before and after. Bottoms concentration IPA= 10% w/w. $8000 1660 Allowance for fouling on tube side = 0.0001 m2 K/W; shell-side steam = 0.00001. This is the main condenser. The overhead product is drawn off by the reciprocating pump. $130 1662 Lumps persist. $3000 1664 Higher than usual value for startup. $3000 1666 Agrees with head capacity. $200 1667 Why1? so that the design flowrate of feed can go to the reactor. Why2? so that the reactor can produce the design rate of product. Why3? so that we have product to sell to customers. Conclude: focus on Why1? “to supply the design flowrate of feed to the reactor.” $50 1668 This is a closed system. Unable to get a reasonable check on the amount of condensate. No drain valve to use to measure condensate. $300

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1669 1670 1671 1673 1674 1675 1676 1677 1678 1680 1681

1682 1683 1685 1686 1688 1690 1691

1693 1695 1697 1699

1700 1701 1702

OK. Recalibration not necessary. $2000 Equipment should do the job. $100 Estimate seems OK. No major blockage. $100 Allowance for fouling on tube side = 0.00001 m2 K/W; shell-side steam = 0.0006. $250 Usual value but fluctuating; past records show similar values. No change. $110 Should give the correct% vaporization per pass. $700 Responds to change. OK. Recalibration not necessary. $3000 IS: on deprop. IS NOT: other units. $50 The Dp for clean, carbon-free reaction coils = 5 kPa; but rapidly increases to 10 to 20 kPa because of carbon formation. This seems consistent with design values and with values used to size/select the vacuum pumps. Erratic. IS: startup of new double-load system. $50 Same as expected. Consistent with the overhead composition at this pressure. $1200 OK. Recalibration not necessary. $3000 TI/3 = –102 –2 C; TI/4 = –101 –2 C. Accurate; no recalibration needed. $3000 Polycarbonate’s impact strength is particularly sensitive to moisture in the feed resin. Blend of polycarbonte and ABS: melt volume flowrate: 12 cm3/ 10 min; molding shrinkage parallel, 0.5 to 0.7%; molding shrinkage normal, 0.5 to 0.7%; tensile stress 60 MPa; Izod notched impact strength (at 23 C) 640 J/m; Izod notched impact strength (at –30 C) 267 J/m; HDT (0.45 MPa) 127 C; (1.84 MPa) 110 C; injection-molding melt temp, 260 C; mold temp. 80 C; injection velocity 240 mm/s. Drying temperature: 105 to 110 C; drying time 3 to 4 h; max. moisture content 0.04%; melt and nozzle temp. 275 to 300 C; rear barrel, 250 to 290 C; middle barrel, 255 to 295 C; front barrel, 260 to 300 C; mold temp. 60 to 90 C; back pressure 0.3 to 0.7 MPa; screw speed, 40 to 70 rpm; shot to cylinder size, 30 to 80%; vent depth 0.038 to 0.076 mm; hold pressure 50 to 75% of injection pressure; injection pressure 70 to 140 MPa; clamp 45 to 75 MPa. $150 Temperature increases consistent with vapor–liquid data for ethylene. 3/4 open whereas usual is mid-range. $300 Responds to change. $200 Explosive conditions 180% above lower limit were found at drop boxes B and C. The composition is unknown; the test instrument showed it was explosive. No explosive conditions were found upstream at box A when it was tested as 20% of explosive lower limit. It must be your plant. $300 Everything OK. Clean as a whistle. $20 000 Steady and usual value. $50 Temperature starts to drop: 185 and 130 C; Temperature near lowest temperature: 175 and 128 C; temperature starts to rise; for a brief period the tempera-

493

494

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1703 1704 1705

1706 1707 1708 1710 1711

1712 1714 1715

1717 1718 1719

1720 1722 1723 1724 1725

tures are 190 and 183 C, then quickly drop to about 185 and 130; then cycles repeats. However, this is not a reproducible cycle. Sometimes the temperatures on either side are about equal, at 128 –4 C. It seems to depend on the time of day. $8000 No improvement. $7000 IS: sporadically any shift. $300 Cycle 1, 10 min Cycle 2 & 3 similar within experimental error in sampling and analysis: ’1: 0.5%: ’2: 0.6%; ’3: 1.2%; ’4: 1.7%; ’5: 2.1%; ’6: 2.6%; ’7: 3.2%; ’8: 3.7%; ’9: 4.9%; ’10: 5.0%. $4200 Hot upstream, cool downstream. Yes, T6 > T2. $50 IS: this plant and downstream IPA processing. $650 Height is the same in both reactors. $13 000 Alumina dryer includes several beds of alumina; with two on-line and one regenerated via hot gas heated by steam. All control valves have block valves, bypass plus valve, drain valve. FRC/1 is feedback control. The vapor and liquid ethylene are attached to a compressor, condenser and drop-down valve in a conventional refrigeration cycle. Much lower than design. $50 Respond to change. $500 The piping is complex and consists of a range of 3-way and 4-way valves to allow any of the dryers to be on-line or regenerated. On-line is straightforward. The feed gas enters the top of the first dryer, exits the bottom, enters the top of the second dryer in series and then out the bottom to a knockout pot. Regeneration is more complex because two different sequential activities occur during regeneration. First, hot “fuel gas” goes through the bed and sends off the adsorbed water. When most of the water has been desorbed from the bed of alumina, the bed must now be cooled before it can be put back on-line. Fuel gas is used for both these functions. First, the desorption is done with “hot” fuel gas; then the cooling is done with “cooled” fuel gas. The fuel gas from both functions is returned to the fuel gas system. Case ’24 gives the details of a similar – but slightly different – system. More general information is in Chapter 1 of Process Design and Engineering Practice (1995) by Woods. $100 Ethylene inside the tubes should be film boiling because the temperature difference for ethylene–butane is > 50 C. $50 Lumps persist. $2000 Usual composition caproic, capryllic and capric fatty acids with small amounts or lauric and myristic fatty acid. This feed does not contain more volatiles than expected. $1500 I am following the usual startup procedure. Nothing is different about the cooling-water condensers. $3000 Same as expected. $1200 This is startup of a new plant. No maintenance has been done recently. dPI slightly higher than usual. $50 Seem consistent with expected gradients and responses. $300

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1726 Responds to change. $300 1727 Not needed; old impeller is 15 cm diameter and shows little erosion or corrosion. This is the correct impeller based on vendor files. No change. Level in flocculation tank still below normal. $8000 1729 See Chapter 3: fired heater/furnace, Section 3.3.2; blower Section 3.2.1; sensors and controllers, Section 3.1.3. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000. 1730 IS: The crude unit was having troubles a couple of months ago but not now. It was producing off-specification products. $150 1731 Usual value. $80 1732 Everything is normal; you should be receiving on-spec feed at the design rate. $200 1733 Shell and tube. Allowance for fouling on tube side = 0.0001 m2 K/W; shell-side steam= 0.00001. $6000 1734 q process fluid = q latent heat. For E100: Fp,E100 (latent heat) or FpE100 438 kJ/kg. Hence, Fp,E100 = q process / 438; Actual Fp,E100 = (F cp)E100 90/ 438 or 0.205 (F cp)E100. Design Fp,E100 = (F cp)E100 96/ 438 or 0.219 (F cp)E100. For E101: Actual Fp,E101 = (F cp)E101 66/ 438 or 0.1506 (F cp)E101. Design Fp,E101 = (F cp)E101 58/ 438 or 0.132 (F cp)E101. Total propylene flow: Actual= 0.205 (F cp)E100 + 0.151 (F cp)E101. If the flow and heat capacities of the process streams are about equal then= 0.356. Design= 0.219 (F cp)E100 + 0.132 (F cp)E101. If the flow and heat capacities of the process streams are about equal then= 0.351. $350 1735 No. Same supplier we’ve had for years. $120 1737 Not successful. $1000 1739 Yes. $3000 1740 Not needed. $3000 1741 Faint weld line, but product still broke. $800 1742 First section: Design (F cp) (1000–750) versus (F cp) (1000–820) or 250 versus 180 or 72% actual versus design. Second section: Design (F cp) (750–600) versus (F cp) (820–730) or 150 versus 90 or 60%. First section Design versus second section; 250 versus 150 = 62.5% of heat load lost in first section. First section actual versus second section actual: 180 versus 90 or 66% of heat load in first section. This incorrectly assumes that cp is independent of temperature but this gives us an idea of what is going on. 1743 Yes, we received a discount price of 20% off for the TDI from another supplier. Indeed, today is the first day you are using the new shipment of TDI. The polyol we are purchasing from the same supplier that we have always used. $300. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 1744 Not needed. $3000 1745 No change. $300 1747 Fully open. $200 1749 Temperature = 41 C. $1800

495

496

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1750 1.1 kg/L 1751 Hazardous because of pressure, temperature and composition. Seems to be expected process gases. $2000 1752 Very much so; it just can’t handle this underflow. Water is everywhere. $700 1753 No change in performance. $300 1754 Negligible improvement. Product breaks. $3000 1755 End suction. Head-capacity curve available. Details of NPSH required are supplied. 1756 Bypass valves: butterfly valves. $300 1757 It was idle for 6 to 12 hours. I suspected a lot of settling had occurred, so I blew back the lines into the thickener with air. $50 1758 Since the new installation. $650 1759 Shell side 228; tube 164; leak shell to tube side or steam into glycerine. Same pattern occurs in other effects. $70 1761 Hot 32 C; sunny. It has been this way for the whole week. $20 1762 Insufficient vacuum/Dp too low/ pressure sensor wrong/ plugging conveying line/ valve not open/ no air to convey the powder/ no powder in the hopper/ powder is bridging in the hopper. $500 1763 Some improvement. Now can control –6 C. $75 000 1765 Fluctuating. $50 1766 Steady and mid-range. $300 1767 Diagram is a reasonable general picture. What’s missing are the vent/blowdown lines on both the shell and tube side of all exchangers; the drain from the knock-out pot that goes to the sewer. Each line has a single gate valve that is normally shut. All control valves have block valves, drain and bypass with valve. All equipment has isolation block valves on the inlet and outlet nozzles. All steam comes off the top of the header; all condensate going into the condensate main enters the top of the main. Steam, town gas, condensate and process lines are on elevated pipe rack. $50 1768 Usual allowance for fouling on tube side and shell side. Correction factor for not-true countercurrent flow is > 0.85. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure is > utility pressure. Fuel-gas pressure, 1.1 MPa > 0.65 MPa water. $200 1769 Excessive fouled with blobs of stuff. $4000. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 1770 IS: hot and sunny. IS NOT: rainy day 1771 P&ID supplied; Simulation report suggests that all equipment on this unit is working close to design capacity. P&ID is less than a year old. No details available in this office for upstream unit where catalyst was changed. $30 1772 IS: On the debut. IS NOT: upstream on the deprop. those problems have now been corrected. $50 1773 Usual allowance for fouling on tube side and shell side. Correction factor for not-true countercurrent flow is > 0.85. Vent/blowdown lines on both the tube

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1774 1775 1776 1777 1778 1780 1781 1782

1783 1785 1787 1788

1789 1791 1792 1795 1796

1799 1800 1801

1802 1803 1804 1805

and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure is > utility pressure. Fuel-gas pressure, 1.1 MPa > 0.65 MPa water. $1000 Negligible. $3000 Reflux rate increases but not to required rate; level in drum holding constant. $20 000 Insulation seems to be sealed; intact. $120 Well designed. Demister included. Globe valve on periodic drain to sewer. $600 DT > 25 C (1000 or 750 or 600 less 325). Therefore all film boiling. –102 –2 C Designed to handle the usual 50 kg/h of non-condensibles. Should do the job well. $110 Sized to take the increased flowrate. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $50 000 < 0.3%. $800 OK, recalibration not necessary. $700 See Chapter 3: pump, Section 3.2.3; sensors and control system, Section 3.1.3. $50 Design: F 4.2 kJ/kg K (70–18) C or 218 F; Actual: F 4.2 (42–18) C or 100 F. This is a loss of F cp (70–42) or F 4.2 28 C = 118 F. The heat ratio of actual/design= 14/52 = 0.27. $200 Valve stems suggest that all valves are closed. No apparent leaks. $12 000 About one month ago, changed the oil in the hydraulics. Continually add a suitable lubricant to the tie bars. $60 Estimate seems OK. No major blockage. $300 Valve is fully open; valve responds to change. $400 End suction, centrifugal. 1800 rpm with impeller one size smaller than casing. NPSH data and head-capacity curve available. An allowance was made of 70 kPa Dp across the control valve. Should do the job. Special attention to seal to prevent acid leakage out. $300 Not needed. $3000 IS: excessive Dp in T103, the C2 splitter. IS NOT: other major variables The blower can easily supply the amount of air to provide at least 10% excess air for the range of fuel rates expected. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 OK, recalibration not necessary. $700 IS: temperatures in reactor bed, “hot spots”; production of ammonia. IS NOT elsewhere on plant. $300 No. We cannot afford to have duplicate. $30 Inadequate water flush/ flush does not go into the correct lines/ pump stopped/ strainer plugged/ line plugged and cannot be cleared with the flush/

497

498

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1807 1808 1809

1810 1811

1812 1813 1815 1817 1818 1819 1820 1821 1822 1823 1824 1827 1828 1829 1831 1832 1834 1835 1837

flush not working/ block valves leaking/ valves around pump closed/ pump clogged. $50 Much lower total amount exchanged than expected. Heat lost in reformer exit gas is same as that gained by stream to other units –10%. $800 Sensor OK. Calibration checked out OK. $8000 “We have not detected any contamination. The pressure is 1230 kPa g.” You therefore conclude that the leak may be from the process stream into the propylene on the very unit that is underperforming. $250 Yes, insulated. The insulation seems to be intact. $500 The raw material for the cat-cracking unit, from storage tank T200, is pumped and preheated via pump A followed by a preheater. A spare pump B is hooked up in parallel so that the process can continue when pump A is being serviced. $50 No apparent fouling inside or outside of tubes. $20 000 Work OK. $120 323 C. $3000 Both appear to be closed. Valve direction consistent with direction of flow. See Chapter 3: evaporators, Section 3.4.1; heat exchangers, Section 3.3.3; vacuum, Section 3.2.2; steam traps, Section 3.5.1; pumps, Section 3.2.3. $50 Overflow effluent = 330 ppm; torque increased. $8000 Steady and usual. $50 Yes, consistent. None done recently; but pump B is scheduled for maintenance next week. $15 None available. $50 Oscillating around 58% and decreasing; usually 62% and steady. $600 IS: both chiller exchangers. $50 See Chapter 3: distillation, Section 3.4.2; exchangers, Section 3.3.3; sensors and control, Section 3.1.3. $50 Not upsets here. Flow, temperature and concentration should be the usual values with temp. = 98 C; concentration 100% ammonia. $70 About one month ago, changed the oil in the hydraulics. Continually add a suitable lubricant to the tie bars. $50 Four stage, reciprocating with kickback protection at the exit of the third stage. 10% overdesign. $6000 Monitor the level in feed drum V-29 so that the amount fed to the column does not exceed the flow to the drum. $50 Yes, although it is not shown on the diagram. $500 The polymerizer is a jacketed stirred tank. The top entry stirrer has double mechanical seals; the amps are recorded. Besides a view port and light and access hole, at the top are two nozzles to vent. The one nozzle has a pressurerelief valve; the other backup emergency line has a manual block valve. Another top nozzle is hooked to a manifold with three lines with block valves: vacuum; vapor recovery and equalizer line to the vapor space in the feed vessels. Another top nozzle is manifolded to three lines, each with manual block valves: emergency air, compressed air and nitrogen. A easy-clean angle block

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1838

1839 1841 1842 1845 1849 1850 1851 1852 1853

1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865

valve is attached to the nozzle for each of the last three nozzles on the polymerizer. A pressure-recorder sensor is also connected to the top of the reactor. The pressure gauge has a pigtail with a shutoff valve with a tee nipple and valve installed between the gauge and the pigtail to allow backflushing of pigtail. Inside the vessel are four vertical wall baffles extending 0.1 D into the vessel, with space between the baffle and the wall for easy cleaning. At the bottom of the polymerizer is an easy-clean angle valve. Four/three lines attach to this valve: a premix feed line, a blowdown line/ feed line to downstream processing and a process water line. These each have manual block valves and check valves. Hot or cold water is fed to the jacket. The temperature sensors indicate the temperature in the liquid in the reactor and a second sensor is used to control the flow of heating or cooling water via a TRC. The temperatures of the fluid entering and leaving the jacket are measured. The TRC valve has block valves, drain and bypass with a valve. There is a pressure-relief valves on the jacket. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. They both are decreasing. The cooling-water exit temperature is decreasing and the steam production rate is decreasing. All else seems to be the same. $50 Usual. $50 Responds to change. $300 None reported. $150 Set at 10% open; usual value 10% and steady. $800 Temperature = 127 C. $1800 Usual value. Seems steady at 1.2 MPa g. $50 Set at 40% open; usual value 20% open. $800 Preheater 1: tube is > 200 kPa; shell is about 120 kPa; leaks from tube into shell; steam into glycerine; Preheater 2: tube is > 190 kPa; shell is about 170 kPa; steam leaks into glycerine. $60 Sounds like cavitation. Amps consistent with “no flow”. $80 Fuller’s earth dumps into the bleacher at the design rate. $100 000 165 C. $150 Mild, 18 C and raining. Previously this week we have had thunderstorms and wind. $150 IS: at exit of pipe reactor. IS NOT: elsewhere. $50 Should do the job. $2000 Reads 995 –7 C Resin premixed as polycarbonate and ABS; coloring agent and exothermic blowing/foaming agent. $80 P&ID supplied; P&ID is less than a year old. $30 Rising. $50 No evidence of fouling. $3000

499

500

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879

1880 1881

1882 1884 1886 1888 1889

1892 1894

1896 1900 1901 1902

Well within design specs even though this is a hot and humid day. $300 No change in swinging-loop phenomena. $100 000 OK; head-capacity curve OK relative to estimated pressure requirement. $240 Appears to be fully open. The direction of flow through valve is consistent with expected flow. Steady and higher than usual, 105 C instead of 101 C. $50 Glycerine: NFPA: 1, 1, 0. $50 OK, recalibration not necessary Watch out for oil contamination. The lubricating oil from the compressors will poison the catalyst if it ever reaches the catalyst. $2000 540 ppm. $4000 No change. $110 Equipment should do the job. $100 In= 18 C; Out = 42 C. $1300 Shut. $400 No apparent leaks on the four accessible flanges. Six flanges on the overhead line to the condenser are relatively unaccessible. No apparent leaks on three valve stems. $3200 Valve was OK. No improvement. $1000 Not needed at the beginning until after we have collected data. Indeed if put on safe-hold we won’t be able to collect data to figure out what is wrong. $3000 Not needed. $3000 South loop: OFF; North loop: ON. $600 No observable reasons. Inside looks a little worn but otherwise OK. Impeller looks OK and the key is in the keyway. $3000 No valve. Would have to stop the process and insert a smaller orifice in the resistance disk in the line. $300 Temperature-sensor error/ power failure/ cal rod heater failure/ gas flow too high/ refrigerant condenser too cold/ cooling-water condenser too cold/ hydrogen concentration < 64%/ startup instructions not followed/ startup pressure too high. IS: DAF tank. Careful and methodical. Should not be a problem. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 IS: first time startup; IS NOT: ever worked Solvent to remove carbon dioxide; mixture of tetramethylene sulfone, diethanolamine and water; boiling temperature of pure diethanolamine = 270 C. Responds to change. $200 Process pump will be able to handle the anticipated flowrates and maintain liquid flowrates > 1.5 m/s. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1903 From the reactant flowrate, measured upstream; the calculated steam usage (and condensate flow) are reasonably consistent with the design values. $3000 1904 Fouled on the outside of the tube. Scrape the scaling and analyze. Negligible fouling on the inside of the tubes. 1906 Motor running. Shaft rotating. Pump noisy. $400 1907 Extensive manuals including maintenance schedule and table of troubleshooting diagnostics. $800 1908 Temperature leaving the cooling tower is about 23 C; flowrate from the towers is about usual for daytime; the temperature is lower at night. We don’t have a direct line to your units so we cannot say exactly how much is going to your plant. $200 1909 Plate is installed correctly and orifice diameter correct. $2000 1910 Confirm the temps and pressures at the bottom and top of column are consistent with the compositions expected. $100 1911 Largest particle is 1.2 aperture and is log normally distributed. No data available for previous operation. $2100 1912 Nothing unusual. $50 1913 100% open. $300. 1914 Operation returns to design operation. PIC = 0.8 MPa for “usual” design flowrate. Response to change remains sluggish. However, as time passes, increased settings on PIC are needed to maintain even the usual capacity. $10 000 1915 Density 1.53 kg/L; viscosity 85 mPa s. $50 1916 Slight improvement. 73% yield. $32 000 1917 < 150 mm diameter, angle of repose 15 to 30 (free flowing), moderately abrasive, fluidizes, hygroscopic and packs under pressure. 1918 Moderate, sunny; today and all week. $50 1920 OK. $2500 1921 Seven hours into first time startup after changes: new catalyst; new feed supply. $50 1923 Internals look normal. Clearances are normal; wear rings are in good shape; no pluggage; negligible erosion. $300 1924 Conditions get worse. Condensate seems to build up in the exchanger. $4200 1925 The operating procedures were followed. $3000 1926 dPI Steady and usual. $50 1927 Pressure steady about 200 kPa g; no upsets. 1928 Shell and tube. Allowance for fouling on tube side = 0.0001 m2 K/W; shell-side steam= 0.00001. $800 1929 Functions OK. Is left fully OPEN. 1930 0.5 mm diameter perforations; usual for steam. $400 1931 Agrees. $300 1932 “We have not detected any contamination. The pressure in the process line is 230 kPa g.” You therefore conclude that any leak would be from the propylene to the process fluid. $300 1934 8.2 MPa. $150

501

502

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1935 Temperature and flowrate are steady and about the same as usual. $200 1937 Feedback control; orifice plate sensor; PID. 1938 Initial increase and subsequent decrease in temperature. After about one to two minutes (allowing sufficient time to purge the fire box of the excess fuel) the temperatures level out and production returns to normal. 1939 40%. $200 1940 Orifice is sized so that, for the range of anticipated flowrates, the discharge coefficient is independent of Reynolds number. $300 1942 Half the design value. 1943 Calculated yield 42%, 48%, 32%. $40 000 1944 Expensive and no improvement. $8200 1945 No. We have had to order in new batches of all feed materials. $150 1946 326 C. $3000 1947 Fluctuating about 45 C. $50 1948 Pipe is clear; injection line meets design specifications. $6000 1949 Nothing that looks like it would vibrate or chatter. $4000 1950 30 to 35 C; past records show similar values. 1951 Fully open. $300 1952 No evidence of fouling. Problem persists. $15 000 1953 Intense thunderstorm. Previously this week has been rainy most of the week. $30 1954 Raining. It has been cool and mild, slightly sunny all week. $30. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 1955 Responds to change. 1956 Much, much less hydrogen for the new conditions compared with previous conditions. $4000 1957 Steady and usual value. $50 1958 Design superheat = 100 kg/s 7.5 kJ/kg K 28 C = 21 MJ/s. Actual= 70 kg/s 8.5 kJ/kg K 118 C = 68 MJ/s or 220% more heat transferred. Question! 1959 Eight months ago. Routine check of rake, skimmer, calibration of instruments and sampler. Usual turnaround checks and adjustments on pump and belt filter. $60 1960 Shell and tube. Allowance for fouling on tube side = 0.0001 m2 K/W; shell-side steam= 0.00001. $6000 1961 Standard literature giving usual steam usage for different vacuum conditions. $200 1962 IS: operators on the debut; IS NOT operators on other plants. $50 1963 OK. recalibration not necessary. $3000 1964 Not available. 1965 Product did not break; overall cycle time increased from 68 to 78 s. $2000 1966 Estimate suggests no blockage. $400 1967 No change. Still cannot control. $10 000 1968 Cycle 1 10 min: ’1: 3%; ’2: 2.9%; ’3: 3.3%; ’4: 3.8%; ’5: 4.2%; ’6: 4.7%; ’7: 4.9%; ’8: 4.9%; ’9: 5.3%; ’10: 5.2. Cycles 2 and 3 similar: ’11:

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1969 1971

1973 1974 1975

1976

1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

1987 1988 1989 1990

1991 1992 1993 1994 1995

5.4%; ’12: 5.7%; ’13: 6.0%; ’14: 6.1%; ’15: 6.3%; ’16: 6.4%; ’17: 6.6%; ’18: 6.9%; ’19: 7.1%; ’20: 7.2%. No data available for previous operating conditions. $2500 I followed the standard startup procedure. I have used the same procedure for the past six startups. $3000 Inverted bucket with bypass and upstream strainer. Blows intermittently. Glove test shows hot upstream and cold downstream. Confirm that it should be working OK. $500 Usual values. No changes here. Tried to operate as usual but pressure gauge in flare line = 112 kPa abs; compressor stopped. $200. Could have increased the capacity by installing a larger-diameter impeller or shifting to 3600 rpm instead of the original 1800 rpm. NPSH requirements for larger-diameter impeller similar to previous. $400 IS: at production rates > 80% the whole process is difficult to control. Usually starting with the South loops the production, and temperatures “swing”. IS NOT: steady production and temperatures on both loops. $300 Should do the job; unclear as to why the washing system doesn’t work. $2000 IS: gas temperature entering reactor. IS NOT elsewhere on plant. $3000 OK. Recalibration not necessary. $3000 400 kPa. $200 Water would flow into the reaction product. Set at 10% open; usual value 10% and steady. $800 Cold runner mold with reciprocating screw injection; L/D of 20:1 with compression ratio of 2 to 3:1. 100 ton, Toggle unit. Includes resin hopper. $150 Responds to change. Direction of rotation and arrow are consistent. $100 No power failure. $150. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 Yes. $110 New plant startup; no data. $50 Allowance for fouling on tube side, steam= 0.00001 m2 K/W; shell-side stream = 0.0005. Care to prevent temperature cross-over. $200 Flash steam flows to a steam header that runs from the ethyl acetate plant to this location. Steam line comes off the top of the header. Steam traps along the line to try to ensure steam is saturated and clean when it arrives at the exchanger. $150 Motor running. shaft rotating in the “correct” direction of rotation. $400 Seems to be independent of operators and the shift. $150 IS: undercooling on E100; overcooling on E101. IS NOT: meeting target exit temperatures on either exchanger. $50 Hot 30 C, humid; yesterday was the same. $30 Not needed. $500

503

504

Appendix D Coded Answers for the Questions Posed to Solve the Cases

1996 Design checks out OK. No errors made. Top temperature and pressure are consistent with the expected overhead concentration. $200 1997 OK, recalibration not necessary. $700 1998 No. $50 1999 Sized to take in increase flowrate. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $50 000 2000 Design checks out OK. No errors made. No changes to feed location, feed temperature, type of trays. Top temperature and pressure are consistent with the expected overhead concentration of IPA. $1000 2001 IS: just this morning. IS NOT previous morning. $30 2002 Yes. $400 2003 Adiabatic flame temperature calculated for 10% excess air but the measured temperature in the combustion chamber is much lower. If didn’t put on safepark as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 2004 Although this is a new plant, no information about commissioning and tests were recorded. $30 2005 For current actual conditions = 0.71 kg/s –10%. $1200 2006 Allowance for fouling on tube side, water = 0.0002 m2 K/W; shell-side stream= 0.0005. Care to prevent temperature cross-over. $400 2007 I tried to keep the cycle time, the pressures and temperatures the same as was used for the prototype. $250 2008 Feels warm. $200 2009 Cooling time lengthened to 51–60 s and product looks OK although the production decreases because of the increased cycle time per piece. Lumps persist. Brown streaking at the same location. $2000 2010 Part of the design of the reactor. Incoming gas flows across tubes carrying hot reactor effluent; then on the tube side with the hot gas from the catalyst bed on the shell side. The heated feed gas then goes up and then down through the catalyst bed. Thermocouples monitor the temperature of the exit gas as it cools down. $6000 2011 Higher than design. $50 2012 As measured on the gauges Dp =110 kPa. $1000 2013 IS: reflux. IS NOT: elsewhere on the unit as yet. $50 2014 0.205 MPa g, the condensate header: 134 C; Steam main: 1.5 MPa g: 201 C. At 1.2 MPa g: 192 C. $300 2015 Warm 25 C; clear, windy. It was raining and thunderstorm about four days ago. Since then it has cleared and settled into nice warm weather. 2016 Estimate seems OK. Flowrate from the “other units” seem same as previously. $700 2017 Internal report of commissioning showed that all units function well and work on design specification. The steam generated from the waste-heat boiler was 50 to 120 kg/s with 28 to 30 C superheat. 2018 Not needed. $3000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2019 2020 2021 2022

2024 2026 2027 2029 2030 2031 2032 2034 2035 2036 2037

2039 2040 2041

2042 2043 2044 2046

2047 2048 2050

2051 2052 2053

No leaks. $5000 No change. $1200 62%. $600 Alumina. cycle time 12 h; two in series on-line; one off-line being regenerated with fuel gas. Fuel gas should have < 6 ppm moisture. Alumina loading 0.14– 0.22 kg water/kg dry adsorbent. $100 Confirm the temps and pressures at the bottom and top of column are consistent with the compositions expected. $100 Steady and usual. $50 Works OK with flow 14 L/s when either pump B or A is working. Issue repair order for next maintenance. Current level= 2.5 m; usual level= 2.5 m. $40 Steady and design value. Responds to change. $100 Acknowledge that they recently opened a new leg of the line from Western Canada. The line was hydraulically tested before the line was put into service. Yes. $3000 Not needed. $3000 Steady and higher than usual. $50 Why1? to ensure safe operation, correct production rate and quality product. Why2? to meet customer’s demand. Why3? to keep company economically viable. Conclude: focus on “to ensure safe operation, correct production rate and quality product”. $40 No improvement. $12 000 Yes. $120 Vendor has supplied standardized triple effects to many satisfied customers. Vendor cautions that the purchaser install the unit “appropriately” into the existing local environment. $300 Valve stem is about 1/4 open. No. $50 None recently. $50 Usual, no sounds attributed to “cavitation”. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Responds to change No change. $13 000 Yes, we do expect the town gas to have < 6 ppm moisture. However, some slugs of water may have been left over from the hydraulic test of the new line and the result might be periodically high moisture levels but this should level out to < 6 ppm as the residuals slugs of water disappear. P&ID supplied; Simulation report suggests that all equipment on this unit is working close to design capacity. P&ID is less than a year old. $30 The operator turns on these lines, I think I hear flow. I shut off the flush after about five minutes. There is still no flow of sludge to the filter. $40 Cool cycle decreases from 45 to 42 s. Product still breaks. $800

505

506

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2054 High-pressure alarm ceases; overhead temperature and pressures realign to design values. $5000 2055 No apparent fouling inside or outside of tubes. $18 000 2056 Silica-based catalyst. 2057 Level seems to be constant and suggests coil is covered with liquid. 2059 Hazardous because of pressure, temperature and composition. Seems to be expected process gases. $2000 2060 54 kPa. $600 2061 IS: since startup. 2062 Pressure gauges not installed. Expect gas Dp, on average, = 0.025 kPa. Should increase when gas flow increases. $50 2063 Inlet temperature consistent with data from utilities; outlet temperature below design. $1200 2065 Not needed. $50 2066 IS: startup. $30 2067 For all exchangers, reboilers and condensers (E130, 131, 107, 108, 113, 114) Process pressure is > utility pressure. For example, E131: fuel-gas pressure, 3.1 MPa > 0.65 MPa water. $100 2068 Not needed. 2069 Process liquid would leak into the cooling water. $50 2070 Design checks out OK. Demister included. $50 2071 PC-100 = 517 kPa; PI-200 = 510. Dp= 7 –1 kPa which is reasonable. $75 2072 Checks OK. Recalibration not needed. $680 2073 Responds to change. $20 2075 No change in sample properties. 2076 IS: The kickback valve on the compressor seems to be working OK now. $150 2077 Midrange both before and after safe-park. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 2078 Slightly higher flow than usual. $50 2079 Well-designed for center-line injection. $100 2080 Not needed. 2081 IS: independent of operators. $50 2083 The scaling from the inside of the tube contains silica. 2084 Mid-tank; usual value. 2085 When: 8 months ago. Previously the regeneration gas for the dryers was supplied by an independent, off-site utility who supplied gas via pipeline. However, since we have basically “town gas” as the overhead from the demethanizer we stopped purchasing the town gas and repiped the overhead from T101 through heaters to the regeneration line. In addition during this turnaround, routine checks were done on equipment. $200 2086 Valve stem is fully open. $30 2087 Lumps persist. $1200 2088 All five samples are consistent with excessive C3 in all samples –10%. $4000. 2090 Temperature = 93 C. $1900

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2091 Negligible fouling. 2092 Sufficient pressure difference. Should work. $800 2093 Two gates, aluminum mold. Product to have impact strength of 640 J/m at room temperature. $70 2094 No. $80 2095 IS: startup. $50 2096 IS: startup, no evidence that it ever worked. 2097 End suction, centrifugal, 1800 rpm. Should have no NPSH problems. $300 2099 No sounds of cavitation. $120 2100 Internals look normal. Clean as a whistle. Some evidence of slight erosion just downstream of valve. $20 000 2101 Flue gas is hazy, black flue gas. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 2102 Yes. From past records, the timing varies and the concentration values differ but surges in high lab values for the bottoms “correspond with” surges in high values of C3 on AC/1. $700 2103 Much greater than design. $600 2104 Amount fluctuates erratically. 2105 Usual value. $120 2106 Controlled and steady. $200 2107 Sample ’1: 650 ppm; ’2: 720 ppm; ’3: 690 ppm; ’4: 705 ppm; ’5: 698 ppm. $2000 2108 Test ’1: flow= 0.5 kg/s; Test ’2: flow= 0.01 kg/s; Test ’3: flow= 2.1 kg/s. $1000 2109 0.97 MPa. $100 2111 Reads higher than expected. $150 2112 52 trays = 120 kPa. $200 2113 54 kPa. $600 2114 Confirm the temps and pressures at the bottom and top of column are consistent with the compositions expected. $100 2115 Startup of new process. $300 2116 48 trays = 104 kPa. $200 2117 Looks OK. No improvement in operation. $1000 2118 See Chapter 3: injection molding, Section 3.9.6. $200 2119 No blockage. $15 000 2120 Temperatures in bed are consistent with the exit temperature of about 1000 C. 2122 As expected for increased flow of naphtha to provide the same space velocity. $300 2123 Steady. $300 2124 No reports available. $300 2125 Yes. $100 2126 Diagram is reasonably complete. Missing are isolation block valves for the strainer trap and a drain line valve within the block valves. Upstream there is a

507

508

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2127 2128 2129

2130

2132 2133 2134

2136 2138 2139 2140 2141 2143 2144

2145 2146 2147

2148 2149

flowmeter to measure the boiler feedwater. The steam comes off the top of the steam header. $50 No. $500 IS: cycling steam flowrates, temperatures and pressures. IS NOT: glycerine feed. $50 Dimensions are correct. Missing are the isolation block valves around the pump; the pressure gauge on the exit of the pump before the block valve; the block valves, drain and the bypass with valve on the control valve; and the drain, feed nozzles and vent on the acid storage tank. Also missing are the safety showers, safety equipment storage for gloves, protective clothing related to the hazards associated with this process. $120 Usual allowance for fouling on tube side and shell sides. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure is > utility pressure. 3.3 MPa > LP steam at 0.85 MPa. $1000 Difficult because the steam flowrate is not recorded. $50 Slightly higher than usual value. Self-standing column with the air-cooled condensers about 10 m above ground. The overhead line from the top of the column goes vertically down and then across to the feed header of the condenser. The exit from the liquid header containing the condensed IPA goes directly to the overhead receiver. From the overhead receiver, the condensed IPA is pumped to downstream units for further processing. Any uncondensed inerts from the overhead receiver go to the vent scrubber system. There are sensors on most lines. There are isolation block valves around all pieces of equipment. For all control valves there are isolation block valves, drain and bypass plus valve. $650 No improvement. $1200 Steady and mid-range. $300 No change. $3000 Yes. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 More than enough supplied: 2.5-cm diameter line, welded connections with four elbows. Vertical height 11.1 m. Typical feed is 3 to 6% solids. Polymer conditioner added to aid filtration. Typical product is 30% “heavy solids”; 20% light solids; 50% water. $50 Pressure on the steam drum is steady at 3.4 MPa g. We’ve had no upsets. The flow is steady. The degree of superheat is not excessive but sufficient to prevent you from having wet steam in your turbine. $300 Not needed. $10 000 OK. Recalibration not necessary. $3000 Selected based on a condensate load of 2.5 kg/s; discharge to atmosphere; common header from all stages with single inverted bucket; sized orifice based on Dp= 125 kPa. Upstream strainer. $120 No leaks. No problem found in the linkage. No change in operation.

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2150 2151 2152 2153

2154 2155 2156 2157 2158 2159 2160 2161 2162

2163 2164

2167 2168 2170 2171 2172 2173 2174 2176

2177 2179

Responds to change. $2000 Responds to change. $100 Trace, trace, trace, off-spec, off-spec, off-spec, trace, off-spec, trace, trace. $2000 It has really been trouble since startup! It seems to randomly fluctuate all over the place. I can’t sort out any trend. The feed is steady as a rock; the pressure is steady; everything else seems to be steady except for the production rate! OK. recalibration not necessary. $3000 Seal is partially failed. Clearance OK. $2000 62%. $600 q= (F cp)E101 DT; Actual= (F cp)E101 (70–4) or (F cp)E101 66; design= (F cp)E101 (68–10) or (F cp)E101 58 or Actual is 14% higher than design. $300 Same as usual concentration. $2500 Not successful. $1000 Yes, responds. $230 IS: gradual decrease in flow and today there is no flow; IS NOT: past three weeks. $50 Why1? to close the mass balance on the IPA column. Why2? to ensure that there is not a loss of chemicals from any unit. Why3? so that our corporation continues profitably. Conclude: focus on Goal: “to prevent the apparent loss of IPA”. $650 Standard literature. $200 Diagram is a reasonable general picture. What’s missing are the vent/blowdown lines on both the shell and tube side of all exchangers; the drain from the knock-out pot that goes to the sewer. Each line has a single gate valve that is normally shut. All control valves have block valves, drain and bypass with valve. All equipment has isolation block valves on the inlet and outlet nozzles. All steam comes off the top of the header; all condensate going into the condensate main enters the top of the main. Vapor samples can be taken from all drop boxes and sewer gates. $50 Not needed for safety purposes. $200 Need safety interlock and emergency shutdown. $200 IS: boiler on exit of reformer gas. IS NOT: upstream Some liquid. No change in pressure drop in column T103. $300 On spec. No complaints from other customers for that batch. $300 Tapping suggests height is mid-vessel, at the design location. Minor improvement. Concentration of propane in overhead to flare decreases to 1.25 design value. $5000 Pressure gauge reading wrong/ cavitating pump/ wear rings worn causing internal circulation/ impeller worn/ blockage in line/ change in density of liquid so that pressure gauge reading converts to a different value of “head”/ faulty check valve/ faulty block valves/ pressure-sensor line plugged. Pressure-relief valve is not faulty. No improvement. $1000 Confirm that the system is temperamental and difficult to control. Sometimes we just don’t condense enough of the high boiler and the vapor from the over-

509

510

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2181

2183 2184

2185 2186 2187 2188

2190 2191

2193 2196 2198 2199

2200 2201 2202 2203 2204 2205 2206 2207 2208

head drum has too many high boilers in it. We have to flare it. It seems to work OK in the winter and when its raining. $30 On file, a thick manual but it does not include a section on trouble shooting. The specifications suggest that the new condenser should do the job very well without a trim cooler. $1000 Steady. $300 Chin lists A. three methods that vary the vapor rate; B. seven methods that vary the area; C. three methods that vary the temperature difference; D. six methods that vary the heat flux and E two other methods. For this clean and moderate pressure system, where there is “full condensation” the recommended options should be from options B, C, or D. Chin recommends B2. Method D1a was selected. $230 No change reported upstream. $20 Responds to change. $200 Bumps appear on the product opposite to the knock out pins. Lumps persist. brown streaking at the same location. $2000 Since the startup following the new supplier of TDI, I was meticulously careful to follow the standard procedure. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 Clear. $2500 P3 = expected value, slightly below atmospheric pressure. If didn’t put on safepark as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Operators confirm at design flows the temperatures are –60 to –55 C; at half the design flow the temperatures are –90 to –96 C. All five samples are consistent with excessive C4 in all samples –10%. $4500 Within specifications. $3000 Design= 218 F = UA LMTD or UA/F = 218/104 = 2.1; Actual= 100 F = UA LMTD or UA/F = 100/ 90 = 1.1. The UA actual/ UA design= 0.5. If the flow through the tubes, F, has not changed, then either the outside heat-transfer coefficient dropped by half or the area became flooded and half the area is lost. $500 IS: at high capacities of 150 Mg/d. IS NOT: low flowrates, 70 Mg/d. Both valves seem to be fully open. $90 OK, recalibration not necessary. $700 Sounds like the inverted trap is functioning very well and consistently. OK. rechecking not needed. $800 Two gauges agree within –4 kPa. $300 Moisture < 0.03%. $3500 No leaks, no blockages, no fouling. No improvement in process operation. $4000 In E100: DT ranges from 105 to 10 C; hence goes through nucleate and film boiling. In E101: DT ranges from 73 to 15 C; hence goes through nucleate and film boiling. $50

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2209 IS: Crystal moisture is 4.5% coming out of rotary dryer. IS NOT: <1.5% 2211 Why1? to provide steady, reliable operation with bottoms meeting specification; Why2? to meet production requirements; Why3? to sell products and keep company financially viable. Conclude: focus on Why1? “to provide steady, reliable operation with bottoms meeting specification”. $50 2212 Hazardous because of pressure, temperature and composition. Seems to be expected process gases. $2000 2213 No sounds of cavitation. $90 2214 Not needed. $5000 2215 No data for this configuration. Internal report for initial installation of single chiller unit shows all performance data agreed with specifications. No surprises. Very reliable, very stable unit. $300 2216 Test OK. $40 2217 Appears to be closed. 2218 See Chapter 3: control, Section 3.1.3; condensers, Section 3.3.3. $50 2219 No upsets. Steady operation. Usual pressure delivered to site. $30 2220 OK, recalibration not necessary. 2221 Bottoms temperature increases by 3 C; bottoms composition is 8% organic; flowrate from pump 114-J of overhead is still very low. 2222 Set at 22 kPa g. 2223 See Chapter 3: For these symptoms, it is difficult to identify specific sections. Perhaps related to steam distribution, Section 3.3.5; and steam traps, Section 3.5.1; exchangers, Section 3.3.3 or valves, Section 3.1.3. $50 2225 Level is not steady, sometimes it doesn’t even show in the sight-glass. 2226 I am following the usual startup procedure. Nothing is different about the refrigeration condensers. $3000 2227 30% higher than design. $50 2228 This diagram presents just the basic connections among the equipment. Missing are sensors, controls, isolation block valves, vents, drains, safety relief and SIS systems. As usual, there are isolation block valves, drain and bypass with valve around all control valves. A line brings fresh hydrogen into the feed line to the reactor. The sensors shown on the diagram are the bare minimum needed to understand, and solve, this case. $300 2230 148 C with some fluctuations; past records show similar values. 2231 Routine check on stirrer two weeks ago. $50. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 2232 Respond to change. $500 2233 56 –1%. $2000 2234 No controls, other than the controls within the screen and centrifuge system. The bottoms from the crystallizer are pumped to nozzle feed to the DSM screen. The crystallizer is a continuous process. The screen is a batch process because it needs to be washed and cleaned periodically. The crystals exit from the screen flows by gravity into the centrifuge except when the screen is being washed (when there is no flow to the centrifuge). The centrifuge is a batch

511

512

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2237 2238 2240 2243 2244 2245 2248

2249 2250 2251 2252

2253

2254 2255 2256 2257 2258 2260 2261 2263 2264 2265 2266 2267 2268 2270

process because of the washing cycle. The discharge from the centrifuge to the dryer is continuous except that there is no crystal flow during the wash cycle in the centrifuge. The dryer is a continuous process. $30 IS: since we started ISO testing. IS NOT: before when we didn’t sample this frequently Two gates, QC-7 aluminum mold. $50 IS: at startup after major turnaround. IS NOT: before shutdown. $50 Yes. $20 OK. Recalibration not necessary. $3000 IS: oil deodorizer. IS NOT: apparently elsewhere on plant. Our customer demand was for full load. I didn’t think we could handle it. However, I set the PIC at our max. But we just couldn’t evaporate and heat the ethylene enough. The unit just shut down because we couldn’t meet the minimum pipeline temperature! $100 On spec. No complaints from other customers for that batch. $400 Hydrochloric acid NFPA 301; causes burns and irritation to eyes, skin and lungs. Decomposes to HCl vapor. $50 Slightly higher flow than usual. $50 TDI: NFPA 211; vapor is extremely irritating, may cause respiratory reaction, eye irritation. Contamination or excessive heat may result in dangerous pressure. TDI caused cancer in animals. For stability, avoid contact with acids, bases, amines and oxidizing agents. Thermal decomposition may produce carbon monoxide and/ or oxides of nitrogen. STEL 0.02 ppm; TWA 0.005 ppm $50 Special design to include internal baffle. Care needed to account for extremes in thermal expansion. Allowance for fouling on shell side, water-boiling = 0.0002 m2 K/W; tube-side process gas = 0.0005. An allowance of 10% more area than needed for design flowrate. Single pass on the tube side with unique baffle design for secondary part of the boiler. See Chapter 3: sensors and control, Section 3.1.3; heat exchange, Section 3.3.3; steam traps, Section 3.5.1. $700 P&ID shows control system for the unit. Appears normal; clearances and settings appear normal. Accurate; no recalibration needed. $3000 About 87% design rate; steady. $600 No change. $9000 Sensor is OK; negligible noise in the sensor. $800 No leaks. $5000 Temperature = 101 C. $1800 Before pump= 10.1 –5% glycerine; into the first stage is 9.9% glycerine. $3000 Typical hydrocarbon with NFPA rating 1, 3, 0. Product did not break; overall cycle time increased from 68 s to 88 s. $2000 Not needed. $2000 Accurate; no recalibration needed. $3000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2272 Today is warm and humid. When the previous parts had been made, the weather was variable spring weather: windy, rain, cloudy, sun with variable temperatures 15 to 22 C. $50 2273 2.7 kPa absolute. $30 2274 No upsets. The flowrate and temperatures are consistent with what you usually receive. $300. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 2277 Slightly higher than usual. $400 2278 Design should handle separation well. Level sensors. Sufficient residence time for both gas and liquid with well-designed demister mesh pads and vortex breakers for gas and liquid, respectively. $800 2279 IS: Out of rotary dryer. 2280 No leaks, no blockages, no evidence of foaming, no fouling. No improvement in process operation. $4000 2281 The product should flow by gravity to the storage tank without the need for pump 114-J. 2282 Usual cycle is to power the heaters, add feed to hopper, start mixer, mix 35 s, stop rpm and screw advances to push melt into the mold, screw reverses, feed drops from hopper into mixer, mix 35 s, and so on. The resin is predried for 3 hours by direct contact with hot air at 95 C. Resin is dried, and components are fed to the hopper. The feed mix enters the extruder that melts and mixes the melt through a combination of barrel heaters and screw mixing. After the melt is mixed the melt is pushed through the nozzle, sprue and runner system and enters the mold through the gates. Air, initially present in the mold, is discharged from the mold through the air vents in the mold. Coolant runs through the upper and lower mold. Once the cooling cycle is completed, the mold is open, the part is discharge; the mold is closed and the cycle repeats. $150 2284 No change. $200 000 2287 Diagram is a reasonable representation. $200 2289 339 kPa; usually much, much less. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 2290 Yes. $120 2293 Fan runs at correct speed; balanced operation; no wobble; pitch of blades changes from negative to extreme positive. $350 2294 Based on the flowrates from the unit, the space velocity seems to be the same as previously used. The mass flowrate through the exchangers should be the same as previous. $900 2298 No improvement. 2299 All OK. $120 2300 No improvement. $300

513

514

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2301 Flow FC/1 = 6.8 L/s. Flow F/7 = 6.5 L/s. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 2303 Moderate, 21 C; severe thunderstorms. The previous week has alternated sunny and cloudy days. The temperature has been about the same. $20 2304 Steady and usual design value. $50 2306 129 C. The usual value is 135 C. $120 2309 dPI Steady and lower than usual. $50 2310 231 C. $150 2312 Accurate; no recalibration needed. $3000 2314 5 s; slightly longer than usual because of the thickness of the product, especially the hub. $200 2315 Not needed. $4000 2316 Estimate suggests no major blockage. Flowrate is consistent with space velocity used previously. $800 2317 IS: “all the non-condensed gas goes to flare”. IS NOT: negligible going to flare. 2318 About once every 24 hours; sometimes on my shift; sometimes on someone else’s. $300 2319 Responds to change. $100 2320 High and rising. $80 2321 99.9 –1%. $2000 2322 90 kPa (9 m head) $15 2324 Well designed based on pressure, residence time plus vortex breaker and demister pad. $150 2325 Signal output is 100% indicating fully open. This is a fail-close valve. $30 2327 Hydrogen: 70.2%; nitrogen 23.1%; methane and other 5.9%. $20 000 2329 I’ve only been on this job for 3 months. I have never seen anything like this before. $50 2330 No. $50 2331 Not needed. $10 000 2332 This is startup. 2333 Both OFF. $300 2334 Valve-stem movement suggests valve works and is fully open. $2000 2335 Why1? to ensure safe and reliable operation; Why2? to meet production requirements; Why3? to sell products and keep company financially viable; Conclude: focus on Why1? “To ensure safe and reliable operations.” 2336 IS: on deprop. IS NOT: other units. 2337 IS: cool to –96 C at half flowrate and –60 C at full flowrate. IS NOT: cool to –96 C at full flowrate 2338 Flow is constant and at design value; temperature going to the chiller unit has Temp= about 70 C. We are not pleased that the temperature of the stream that we are getting from your chiller unit is so cold! We need it to be 10 to 11 C and not the 4 to 5 C that we are currently getting. Do something on your chiller unit or we are going to have to shut down! $500 2339 No change. $10 000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2340 Tapping suggests that the height is half-way up the tube bundle. Half the tubes are covered. 2341 Not needed. $24 000 2343 No sounds heard. $20 2344 Units are neither over- or under-designed. $200 2346 Estimates agree with Dp. $200 2347 Under design conditions, the cooling coil removed the appropriate amount of heat. $600 2349 No improvement. $6000 2350 OK. Recalibration not necessary. 2352 Responds to change 2353 Sample: 0.2% suspended solids. $300 2354 Set at 0% open; usual value 20% open. $800 2356 Confirm that the system is temperamental and difficult to control. Sometimes we just don’t condense enough of the high boiler and the vapor from the overhead drum has too many high boilers in it. We have to flare it. It seems to work OK in the winter, when its raining and often on the night shift. $40 2358 IS: pump and control system IS NOT: apparently upstream. $50 2359 That particular batch tended to blob. Several complaints from other customers for that batch. $350 2360 The increased duties in the secondary reformer cause traces of silica from the catalyst to be vaporized in the presence of water vapor and carried along in the process gas stream. As little as 1 ppb of such silica can foul the surfaces of the critical waste-heat boiler. The first effects of this fouling do not normally appear until after three months of operation and the tendency disappears after about nine months. See Chem. Eng. June 5, 1967, p 125. 2361 IS: cycling of the amount of product. IS NOT: steady production 2363 All valves appear to function OK. $800 2364 Yes. All valves appear to be closed on tube side. $800 2366 Steady and almost fully closed. $200 2367 23 actual trays; allowed two extra trays. $350 2368 IS: when only A is running; B is stopped. IS NOT: when B is running and A is stopped 2369 The level seems to be about 1 m above the exit nozzle and it is rising. $300 2370 Takes about 2 days. Nothing to be seen. $40 000 2371 Steady as usual at this end; the production might be a little lower than usual. The feed you receive should be on spec. –1%. $50 2372 Diagram reasonably complete. Isolation block valves on process line entering and leaving the shell side of the preheater. Isolation block valves, drain and bypass with valve around the control valve V100. $50 2373 Diagram is a reasonable representation. In addition, the torque of the rake is measured and displayed in the control room. The sludge flow is measured at the outlet of the pump. The pump has isolation valves with a pressure gauge on the discharge line upstream of the block valve 2374 Yes.

515

516

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2375 Yes, both were batch from BZ684. $2000 2377 61 C. $30 2379 Connect with flow controller FC-5. Good move! However, if didn’t put on safepark as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 2380 Allowance for fouling on tube side = 0.00001 m2 K/W; shell-side steam = 0.0006. Should have sufficient area to boil off propane to give on-spec bottoms. $250 2381 Seven hours into first time startup; changes have been made to correct faults on the upstream deprop. Case ’41 $50 2382 112 kPa. $400 2383 Insulation in good condition; noise and temperatures as expected. It’s a hot day out here and it’s even hotter by these exchangers. $300 2384 No distinctive noise. Valve not open. No evidence from flare that the pressure relief has opened. $300 2385 No change in either light ends or in flowrate. $178 2386 Looks OK. No improvement in operation. $1000 2387 502 ppm. $4000 2389 Some improvement; by adjusting level below complete coverage of the bundle, T/8 reads 9 –2 C; With care we can get it closer to the target 10. $2000 2390 Responds to change. $200 2391 IS: problem is at boiler feedwater heater exit. IS NOT: elsewhere on the unit. $50 2392 The temperatures and flows go wild in the 5 minutes after we first notice a hot spot but it takes us about 1 to 3 hours to get things settled out after an upset. $300 2394 Feedrate the same: feed concentration the same 50% w/w IPA. The feed rate is 2.78 kg/s –10%. $2000 2395 Tray collapsed in the stripping section/ too much bottoms fed to the debutanizer/ too much overheads in the feed/ feed valve FV-1 stuck/ pump F-26 not working/ check valve on the idle pumps allows backflow/ no feed left in feed vessel V29. 2399 IS: C2 splitter, T103. IS NOT: T101 and T102 2400 Not needed. $20 000 2401 Steady and higher than usual, 108 C instead of 101 C. $50 2402 Header 1.7 MPa g; steam drum 1.72 MPa g. 2403 Diagram seems to be a reasonable schematic representation. For all the control valves there are isolation block valves, a drain valve and a bypass with valve. The chlorine storage tank has drain and fill lines attached to nozzles on the tank. The pump has isolation block valves with a pressure gauge on the discharge line followed by a check valve. $250 2404 Reads as expected $150 2405 Steady at 1.73 MPa. $500 2406 Not needed. $10 000 2407 Level in indicator shows that the butane should cover the full bundle. $100

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2408 2409 2410 2411 2412 2413 2414

2415 2416

2417 2418 2419 2420 2421 2422

2423 2425 2427

2428

2429 2430

2431 2432

IS: now. IS NOT: values recorded for making 100 kg/s of steam. Responds to change. $200 Acetone, n-hexane, methanol. $200 No. $20 PID and carefully tuned. $400 180 to 185 C; fluctuates. Flowrate and temperatures are as expected. We increased the biocide slightly because the days are getting warmer, but nothing surprising is happening over here. I hear you are having fun! $230 Higher than usual value for startup. $3000 Dirty machine/ dirty hopper/ moist feed/ too many volatiles in feed/ lubricant or oil on mold/ incorrect mold lubricant/ feed contaminated during material handling/ faulty raw material from supplier/ poor shutdown procedures/ screw rpm too high/ sensor error/ shutoff valve dirty or clogged/ injection pressure too high/ gates too small. $50 About 87% design rate; steady. $600 Higher than expected. Higher than before the shutdown. $300 Temperature is 23 C; flow is steady. No. $20 No change; still cycles. $1000 IS: with this shift starting up. IS NOT: with previous shifts last week. $50. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 See Chapter 3: distillation, Section 3.4.2; exchangers, Section 3.3.3; sensors and control, Section 3.1.3. $100 No characteristic noise. Valve not open. No evidence from flare that the pressure relief has opened. $300 Response to change has been sluggish since startup. Over the past few months PIC at 0.8 MPa for the “usual” ethylene load. Brief period about two months ago when demand was below usual and PIC setting was at 0.7 MPa. Since then, for the usual demand the PIC settings have been increasing slightly as time progresses until recently the PIC setting was 0.89 MPa for the usual load. $150 Usual value even at the increased flowrate. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Process pressure > atmospheric. $200 Sample 1: 0.96%; 2: 0.99%; 3: 1.04%; 4: 1.02%; 5: 0.98; 6: 1.01%; 7: 1.04%; 8: 1.06%; 9: 0.89%; 10: 0.97%; 11: 1.08%; 12: 1.05%; 13: 1.03%; 14: 1.02%; 15: 1.05%; Insert swing: samples during 10-minute swinging loop: S1: 1.02%; S2: 1.03%; S3: 1.02%; S4: 1.03%; S5: 1.03% 16: 1.08%; 17: 0.99%; 18: 0.96%; 19: 0.93%; 20: 1.01%; 21: 1.02%; 22: 1.04%; 23: 1.06%. The swinging occurred during the start of hour 15. $80 000 No upsets. Steam pressure is about the same at the two locations. $500 As measured on the gauges Dp= 140 kPa. $1000

517

518

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2433 All valves appear to function OK on shell side. $800 2434 Drop box B: 55% hydrocarbon consisting of 64% methane; 32% hydrogen and 2% ethylene. This is the same composition as the overheads from tower T 101 that is used as regeneration gas for the dryer. Drop Box C: same as for B within experimental error. $1500 2435 12. MPa g, saturated steam temperature 325 C. Latent heat of steam, 1188 kJ/kg; heat capacity steam 7.5 kJ/kg K. 2437 No change; original valve looked OK. $45 000 2439 Not needed. $2000 2440 No apparent contamination. 99.9% ethylene. 2441 Signal output is about 80%. This is a fail-close valve. $30 2442 All five samples are about 2.5 times design value –10%. $4500 2443 Higher than usual 0.532 MPa and increasing. $50 2444 OK. Recalibration not necessary. $3000 2446 Allowance for fouling on tube side = 0.0001 m2 K/W; shell-side steam = 0.00001. As the name suggests, to condense any vapors that have not condensed and to ensure that these do not solidify in the knockout pot V1402 or in the barometric condenser E1402. $120 2447 25 C. $200 2448 Usually very fast. If there is a mega-change in residence time, then perhaps a consideration. $120 2449 On specification. $160 2450 Steam 228 kPa abs, 123 C; 164 kPa abs, 114 C; 115 kPa abs, 103 C. $100 2451 Without data from the instruments on the unit, office-based calculations of the exchangers before the change suggest the exchanger should do the job. If the flowrates or hydrogen or water composition of the exit gas from the reformer change, then we need to revisit the calculations. $200 2452 Specialized design. Should do the job. Some designs use a standby furnace to bring the system up to temperature. This design uses cal rod heaters. Thermocouples in the catalyst bed and in the feed gas entering the bed. $6000 2453 Allowance for fouling on shell side, process gas = 0.0005 m2 K/W; tube-side steam superheating = 0.0005. 2454 Consistent. Heat picked up in cooling water = heat to be removed from reactant gas. 2456 Sounds as though air is flowing; gauge pressure gradually increases. $15 000 2457 On specification. $160 2458 Design checks out OK with pressure drop allowance of 140 kPa for the control valve. Check valve on discharge. No gauge. $70 2460 Yes. $20 2461 Steady and slightly higher than usual. $50 2462 100 kPa. $300 2463 Plugged at sampler inlet. $120 2465 No improvement. $30 000 2466 46 –1%. $2000 2467 Controller output suggests negligible stiction of < 0.1%.

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2468 Cooling surface area is double that needed. Fully equipped with charge vessels, instrumentation, pressure relief and SIS. $200. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 2469 Responds to change. $200 2470 OK. Recalibration not necessary. 2471 This plant has never worked right. 2472 Utility pressure > atmospheric. $50 2473 Cold runner mold with reciprocating screw injection; L/D of 20:1 with compression ratio of 2 to 3:1. Toggle unit. Includes resin hopper. For this product, the shot is 40% of the machine capacity. Screw diameter, 6 cm; maximum rpm, 270 rpm; maximum pressure 137 MPa; stroke length 20 cm; injection speed 600 cm3/s. $200 2475 Speed same as specs. 2477 Yes. All valves appear to be closed on shell side. $800 2481 IS: lumps are in the product when viewed by transmitted light. IS NOT: seen when viewed by reflected light. $50 2482 Not needed. $3000 2484 This is not a double-suction pump. This is an end-suction pump! Can’t happen. $50 2485 U tube, horizontal. $100 2486 Allowed 130 kPa Dp for the control valve. NPSH required = 2 m water. Simulation shows that unit was operating very close to design values for usual range of feedstocks and conditions. $200 2487 Amps reading agrees with expected values. $400 2488 Usual design value. $600 2489 Units are neither over- or under-designed. Mechanical baffles and extra head space allowed to account for any foaming. $200 2490 Typical for the overhead material: 0.1 kPa; T = 85 C; 0.5 kPa; T = 118 C. Liquid heat capacity 2 kJ/kg K; latent heat about 200 kJ/kg. $80 2491 Much greater concentration of C4 and C5 than before. $600 2492 If less neutralization occurs, then there will be less heat to remove. Could we be removing too much heat and overcooling? $90 2493 No fouling. 2494 Warm and cloudy 20 C. We had a thunderstorm about four days ago; then the weather continued clear and mild. $10 2495 TIC- 3 increases; TI-5 increases. 2496 No change. 2497 Warm, sunny 28 C. Some rain sprinkles at the start of the week; but it cleared up to this nice weather. $2 2498 Valve at full open position. Valve is one size smaller than line size. Full bypass, block, sample lines are in place. $120 2499 The diagram is correct; there is no pressure gauge. However, the block valves on both inlet and outlet are one size smaller than the pipe size. Checks with design plans show these should be the same size as the pipe size. $120

519

520

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2500 No improvement. $21 500 2501 All columns plus ancillaries should be able to handle up to 160 Mg/d ethylene capacity. $800 2502 Same as expected. $30 2503 Lower than expected. Lower than before the shutdown. $300 2504 Steady and almost fully open. $200 2505 Alarm is not on. $50 2506 Ampere draw lower than expected for design flow. Estimated power corresponds to flowrate about 8.5 L/s –30%. $120 2507 Accurate; check out OK. $20 000 2508 New plant. $50 2509 Everything started fine. However, we really seem to have problems with acid addition when low flows are required. The pump sounds as though it is cavitating under these conditions. We tried putting the controller on manual but this didn’t seem to help. $150 2510 Not needed. $2000 2511 C1400 about 0.6 kPa; now 5 kPa; E1400 vapor side, about 0.55 kPa; now 3.4 kPa; E1401 vapor side about 0.45 kPa; now 2.7 kPa. $120 2512 When: just completed. What: cleaned all the condensers; replaced valves A by globe valves instead of the previous butterfly valves. The butterfly valves did not seal. Routine maintenance on the compressors. Checked the calibrations of all temperature sensors. $5000 2513 Pump has no pluggages; adjusted the clearances, although negligible adjustments needed. $5500 2514 Since cooling water or air cannot be used as the coolant, various “refrigerants” are used: ethylene, propane, propylene. Each refrigerant is part of a refrigeration unit consisting of the four parts: 1) an “evaporator-condenser”, 2) a refrigerant compressor, 3) a refrigerant condenser and 4) a let-down valve. In the evaporator-condenser, the liquid refrigerant inside the tubes evaporates and causing the process vapor on the shell side to condense. To allow the refrigerant to be reused in a closed cycle, the refrigerant vapor is then compressed, condensed and the resulting liquid reduced in pressure to return around the refrigeration cycle. The pressure on the refrigerant side of the evaporator is adjusted to provide the correct temperature driving force for condensation of the process vapors. $400 2515 Heat removed consistent with theory. 2516 Fully opened. $300 2517 Usual value. 2518 Shell and tube. Should work well. Baffles were in vertical. Usual allowance for fouling on tube side and shell side. Correction factor for not-true countercurrent flow is > 0.85. Vent/blowdown lines on both the tube and shell side. $700 2519 OK. Recalibration not necessary. $2000 2520 IS: 7 hours after startup; IS NOT: no information known because this is startup. $50 2521 7 s. $150

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2522 Yes. $3000. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 2523 IS: after short time of use. $60 2524 Standard design. Should do the job easily. $100 2525 1800 rpm, 15 cm diameter impeller, 0.75 kW; 10 L/s at head of 10 m. NPSH (required) = 1.2 m. Head-capacity curve available. At zero flow, h= 45 m. $45 2526 1000 C. 2527 IS: Amount of heating of streams to other units is less; IS NOT: the usual amount of heating. $150 2528 Slight hydrocarbon contamination from somewhere on site during the last couple of months. $120 2529 IPA: NFPA: 1, 3, 0. Dow value: 16. $650 2530 No change. Slightly different cycle but still cycles. $150 000 2531 Should give the correct % vaporization per pass. $700 2532 No. $120 2533 Inverted bucket. $50 2534 Overflow effluent = 350 ppm; belt filter OK. $5000 2535 IS: temperature gauge is higher than expected; stirrer should be operating; IS NOT: controlled at 127 C; stirrer is not operating (based on the lights). $50. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 2537 IS: Everything upstream and downstream seems to be the usual behavior. IS NOT: different from usual. $50 2538 No leaks. 2539 Negligible non-condensibles. Concentration same as always (within the usual sampling and analytical error). $9000 2540 Respond to change. $500 2542 F/2 reads ... and it reads 50% higher just before the system was put on “safepark”. If didn’t put on safe-park as first step, then dangerous potential fire/ explosion conditions are created while you experiment. $500 000 2543 No details available. $50 2544 See Chapter 3: refrigeration, Section 3.3.4; sensor and control, Section 3.1.3; turbine, Section 3.3.1, exchangers, Section 3.3.3. $120 2545 540 ppm. $4000 2546 From a lake that is fed by several streams. $100 2547 Recommends no bypass; (bypass is included in this design); use of a drain valve inside the block valves (no drain valve included). Inverted bucket is the recommended type although float trap is another option. The other recommendations are consistent with the design chosen. $1000 2548 Yes, carbon formation but not more than expected for three days of operation since startup. 2549 Reads 1.5 C. $100 2550 No change. $6000 2551 Hydrogen has relatively high heat capacity; about 10 kJ/kg K compared with about 2 for naphtha or organic vapors; hydrogen thermal conductivity is about

521

522

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2552 2553 2554 2555 2556 2557 2558

2559

2562 2564 2567 2568 2569 2570 2571 2572 2574

2575 2576 2577 2578 2579

2580

2581 2582

0.2 W/m K compared with about 0.02 for naphtha, benzene and other organic vapors. The Pr numbers are similar. Hence, the thermal properties of a mixture of hydrogen and other organics depends on the gas composition. $200 Clean, clear valve stem; trim in good condition. $5000 Film boiling since DT approx. (220–120) = 100 C. Relatively clean. $12 000 For pure methyl chloride: 630 kPa abs; 530 kPa g. Since the mixtures is a gas, the concentration of butylene decreases the vp to < 450 kPa g. Fewer sink marks, less air trapped in part. Pack in the mold is improved. Product breaks. $800 Valve appears to function OK. $200 Nothing much. It should be straightforward. Exit pressure from the third stage was 7 MPa during startup. Dp across the reactors was higher than I usually see during startup. $3000 Natural circulation; well designed. Should do the job. Heat flux 50 kW/m2; overall heat transfer 1 kW/m2 K. Hydrocarbon is relatively clean; moderate pressure at 230 kPa g. Fouling assumed 0.0003 m2 K/W for process fluid; 0.0001, for steam. Float trap for the condensate. $100 The design has the exit pipe inserted into the hopper. However, the sides are designed such that neither ratholing nor bridging should occur. $2000 None on this unit. IS: effluent overflow and underflow from mill clarifier. IS NOT: elsewhere. $50 IS: at suction to booster ejector. $30 No change. Level in flocculation tank still below normal. $5000 Usual concentration of butane. $50 New plant. $50 Very sluggish response. $800 Before “safe-park” the signal output was “high” ; after “safe-park” the output is “mid-way”. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Lumps persist. $3000 Can’t see any leaks $2400 Steady at 31 –1 C. $4000 Not needed. $40 000 Why1? to provide smooth control of pH. Why2? to maintain the pH within target values for the feed to the effluent treatment plant. Why3? so that the effluent plant can handle the waste water it receives. Conclude: focus on Why2? “to maintain the pH within target values for the feed to the effluent treatment plant”. $50 Steam at 240 kPa condenses/ boils about 115 kPa so that DT for condensation is about 5 C. Seems consistent for the design conditions. Now there is a large DT. $40 47 –1%. $2000 Responds to change. $200

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2583 Diagram is reasonably correct and complete. There is a block valve isolating the pump from the KO drum; vents on the top of the cooling-water condensers, the propylene drum and exchangers E100 and E101. There are block valves, drain and bypass-with-valve around both control valves on LC 3 and 4 and on the steam to turbine. The turbine-compressor has the usual controls and instrumentation. The steam comes off the top of the steam header and the low-pressure steam from the exhaust goes into a low-pressure steam line for use in heating other parts of the plant. $2000 2584 No. 2585 No change. $600 2586 Negligible fouling; no condensed liquid; baffle spacing correct and baffles secure. $6000 2587 Trace; within spec. $1000 2588 IS: ever-increasing oscillation in OPRC output. IS NOT: steady. $50 2590 Steady at 0.6 kg/s. 2591 Standard ones we have always used. $250 2592 No dirt or contamination. Valve is clean. $3000 2593 Yes, the impeller shaft is rotating in the direction consistent with the “arrow” on the casing. $300 2594 OK; head-capacity curve OK relative to estimated pressure requirement. $240 2596 Why1? to get flow to DAF to expected value Why2? to keep plant operating safely and prevent upsets in operation. Why3? so that plant can handle the waste water it receives. Conclude: focus on “to get flow to DAF to expected value”. $50 2597 Checks OK. Recalibration not needed. $680 2598 Feed drum is under pressure (because of pressure relief); probable Dp= 40 kPa across drier followed by 5 to 10 kPa across condenser = 400 kPa g on the feed drum. 2599 Baffle should be fully closed. 2600 No improvement. $45 000 2601 Zero. $50 000 2602 Pump has no pluggages; adjusted the clearances, although negligible adjustments needed. $5500 2603 Reads 0.8 MPa. $100 2604 FRC reads wild fluctuations. 2605 342 kPa g with some fluctuations; past records show similar values. 2606 Steady and mid-range. $300 2607 30 C. $300 2608 IS: sludge not flowing to filter. IS NOT: feed should be going to filter. $50 2609 Yes. $120 2610 Yes, both read 112 kPa abs –2 kPa. $550 2611 Improves. Level in the reflux drum gradually decreases to normal. Reflux flowrate still “flat out”. $2000 2612 Not needed. $50 2614 Midway. $200

523

524

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2616 2617 2619 2621 2622 2623

2624 2626

2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2638 2640 2641 2642 2644 2645 2647 2649 2650

Tapping suggests height is the same as level gauge reading. Yes. Valve appears to be closed. $200 Not needed. $4000 IS: just after startup. IS NOT: no previous information. $50 Overflow effluent = 95 ppm; belt filter operation OK for a while. Then, everything changes and does not work no matter what we try. $3000 When the production rate in the “hot loop”, say the South loop, is 20% higher than that in the other loop, say the North loop, then there is no more reaction in the North loop. $300 Nothing done recently. $10 Moderate, windy. 20 C; This weather has occurred all week. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 IS: in August; IS NOT: in January when simulation done Should do the job. $300 11 s. $200 As viewed from the view port, the liquid level appears to be 20 cm below the Fuller’s earth entry pipe. $2000 Seems to be near the bottom but very difficult to tell. $300 Designed on condensate flowrate and for pressure differential across trap. Upstream strainer. No bypass. Discharge to header. $600 Weather today and all week has been moderate and sunny with highs of 23 to 25 C OK. Recalibration not necessary. Recalibration not needed. $1200 IS: operator of this plant. IS NOT: Others. Usual value. $120 Butene about 2.2, methyl chloride about 1.5; water about 4.3 and ammonia about 4.5 kJ/kg K IS: the seal pot did work OK last summer after the compressor had been installed. $150 IS: after they modified the process for more wash water in the washing cycle of the centrifuge. Slightly higher than usual value. Usual. $300 Bypass valve closed. Excessive scaling on the inside of the tubes. IS: no specific location in the part is identified. $50 There are no standard operating procedures. However, I have been operating this unit for many years so I think I understand its peculiarities although it never has worked correctly. However, today it’s much, much worse than I have encountered before. Started it up as usual; carefully set values at usual values; ensured the steam tracing was off. But the flowrate pumped out is just not up to expectations and the bottoms concentration is still too high.

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2651 Allowance for fouling on shell side, process gas = 0.0005 m2 K/W; tube-side steam superheating = 0.0005. Simulation shows that this unit should do the job; it has done so in the past. 2652 No improvement. $4000 2653 Not needed. 2655 Serious burns. The hot water sprayed over the operator when the top up valve was opened. $5000 2656 Within specifications. $3000 2658 Pressure higher in the propylene so propylene would leak into the cooling water. $50 2660 Flush lines and clean out strainer as needed. Done three weeks ago. Extensive preventative maintenance done on the belt filter press. $20 2662 Four rows of finned tubes. Fan diameter-bundle width; gas Dp= 0.025 kPa; Net free area for air flow =50% face area of bundle. Effective MTD = 20.5 C; air velocity 3 m/s. $100 2663 Pyrometer reads 232 –3 C; temperature sensor reads 231 C. $300 2664 Valve stems turn. However, valves are one size smaller than line size. $150 2665 Not needed. $40 000 2667 NFPA ratings are: methane, 1, 4, 0; propane, 1, 4, 0; propylene, 1, 4, 1; hydrogen, 0, 4, 0; ethylene, 1, 4, 2; Explosive limit for hydrogen 4.1–74.3% in air; the lower explosive limits are: for methane 5.0% v/v; propane 2.1% v/v; propylene, 2.0% v/v and ethylene 2.7% v/v. $50 2668 Moderate: 19 C; thundershowers predicted. It’s spring! This week has been chilly temperatures and overcast. $50 2670 Correct amount of TEP added. Carbon buildup as expected. 2671 Not needed. $2000 2672 Test OK. $40 2673 Run a charge of acrylic through the extruder. Yes, I do this whenever we start up each day. $250 2674 Occasionally flashes. $50 2676 IS: effluent 200 ppm; flooded belt filter. IS NOT: 40 ppm; not flooded belt filter. $50 2677 No. $150 2678 No blockages. Clean. No improvement in process operation. $3000 2680 We are gathering such data now. No previous reports are available. 2681 IS: high-pressure alarm on debut; IS NOT: any other sensor signal. $50 2682 OK; should work. $50 2683 No improvement. $1020 2685 Same as expected. Consistent with the overhead composition at this pressure. $30 2686 Condensate buildup/ trap malfunction/ sensor error/ steam pressure too low/ inerts in the steam/ inert buildup in the exchanger/ fouling of the tubes on the outside/ fouling of tubes on the inside/ steam superheat too high/ steam flowrate < design/ water flowrate > design. $50 2687 No change; still cycles. $2000

525

526

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2688 Checked out TIC-3 and it should have the baffle completely shut. I really don’t want to reduce the flowrate through the reformer because we have high demand for ammonia these days. 2690 35 m/s, about 3 kg/kg loading and pressure drop is less than that provided. The vacuum in the bleacher should be more than adequate. $2000 2691 Negligible organic. $500 2692 Not needed. $4000 2693 Cool cycle increases from 45 to 47 s; product breaks. $800 2694 Relatively clean. $12 000 2695 None. We are gathering data as the plant is operating. We have already noted that the pressure-control system is “not very good” and appears to be sluggish. $200 2696 Inverted bucket; intermittent discharge. Sized for double the design flowrate of condensate. Orifice size selected for a Dp of 1.2 MPa with an allowance of 100 kPa pressure drop across the control valve. 2697 Mid-range consistent with full design flow. 2698 Mass balances closes within 10%. $100 2699 Design F cp DT = F (steam) latent heat or 28 L/s 1 kg/L 218 = 6175 kJ/s = F(steam) 2263. Hence F(steam) estimated as 2.7 kg/s; for actual F (steam) = 0.75 kg/s. $500 2700 IS: startup of new plant. $50 2701 Steady and mid-range. $300 2702 Sharp edge facing upstream. OK. 2703 No change. $50 000 2704 Aluminum with carbon steel for the center of the hub with two feed gates near the hub. Two cooling lines each side. $200 2705 Steady and mid-vessel. $50. 2706 Should do the job. Two extra plates in column to allow for growth. Top pressure = about 230 kPa g. $200 2708 q= (F cp)E100 DT; Actual= (F cp)E100 (100–10) or (F cp)E100 90; design= (F cp)E100 (101–5) or (F cp)E100 96 or Actual is 94% of design. $300 2711 Negligible improvement in heat transfer; increase in gas pressure drop and resulting upstream adjustments to flows. $20 000 2712 Traditional fixed-bed catalytic bed with water jacket and numerous thermocouples so that we can monitor the bed temperatures from the control room. Silica-based catalyst. 2714 Not needed although hazardous conditions might arise. Caution is needed. $50 2715 Not needed. 2716 Ammonia: NFPA health 3; flammability 1; reactivity, 0. Toxic, corrosive gas. Overexposure can be fatal. Low dosage: irritation to nose and throat. > 5000 ppm may result in rapid death due to suffocation or fluid in the lungs. Flammable in air for concentrations 15% to 28% v/v. Gas can ignite explosively if released near an active fire. The explosive range broadens 1) if hydrogen is mixed with the ammonia and 2) at higher temperatures and pressures.

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2718 2720 2722 2723 2724 2726

2728 2729 2730 2731 2733 2734 2736

2737 2738 2739 2740 2742 2744 2746

2749 2750 2751 2752 2753

Presence of oil and combustibles increases fire hazard. Ignition energy > 0.68 J. Autoignition temperature 651 C which is lowered from 842 to 651 C by the presence of iron. At atmospheric pressure, ammonia decomposes to hydrogen at temperatures > 450–500 C. Gas has explosive sensitivity to static charge. Ammonia is highly reactive with most metals, especially mercury, gold or silver compounds. Reacts violently with tellurium tetrabromide and tetrachloride, chlorine, bromine, fluorine and with acid halides, ethylene oxide and hypochlorites. Nitric acid: 2, 0, 1. $50 Sized with valve position mid-range for design flowrate. $400 Calibration not necessary. No change in operation. $3000 Same as expected. $30 Should be OK. allowed for in the Dp across control valve FV-6. Suction side OK, complete with vortex breaker. $300 Lumps persist. $1200 For all exchangers, reboilers and condensers (E 131, 107, 108, 113, 114) Process pressure is > utility pressure. For example, E131: fuel-gas pressure, 1.1 MPa > 0.65 MPa water. $400 IS: pumps and flowmeter. IS: product breaks. $60 None found. $30 000 We are not receiving the usual amount of product and what we are getting is off-spec! What’s going on? $300 No change. $5000 Responds to change. $20 Calibrate OK. $1200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. No change in swinging-loop phenomena. $200 000 38 to 42 s; based on 1.5 s/0.1 mm wall thickness. $150 Functions OK. Is left fully CLOSED. No improvement. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 Clear, no obstructions. $8000 IS: This day. IS NOT: noticed before Booster ejector plus two ejectors with direct contact interstage condensers plus well-designed hot well should provide reliable vacuum. The steam supplied for the ejectors is usually reliable; the steam pressure has never been < 95 and >120% of specifications; the steam has never been superheated by more than 10 C. $100 45 C. This is within the usual range. $120 No Fuller’s earth is conveyed. $150 000 Interesting new set of data. No improvement. $50 000 Clean, clear valve stem; trim in good condition. $5000 Warm, sunny all week, including today.

527

528

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2755 IS: 35 min after polyol addition completed. IS NOT: earlier or later. $50. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 2756 The plant worked well; met all specifications. Everything worked at the design rate. At full design rate, the conditions for the North loop were: lowest pressure in loop: 2.75 MPa g; temperature at the exit of the cooling water exchanger = 30 C; at the exit of the refrigeration exchanger = 10 C; loop-gas analysis at the inlet to the reactor: H2 = 62%; N2 = 21%; methane = 13%; ammonia = 4%; discharge temperature of the fourth and recycle stages of the compressor = 40 C; valve A= 20% open; valve B = 10% open; inlet temperature to catalyst bed in reactor = 505 C; exit temperature from catalyst bed in reactor = 550 C; Dp across catalyst bed = 75 kPa g. $300 000 2757 We’re not getting much and what we get is off-spec. $50 2758 IS: when on automatic control. IS NOT: on manual. $50 2760 Steady and lower than usual flowrate. $50. 2764 Pressure gradually rose to 45 m; (450 kPa). This corresponds to head at zero flow as given on the vendor’s head-capacity curve. $50 2765 Allowance for fouling on tube side = 0.0001 m2 K/W; shell-side steam = 0.00001. $150 2767 No change in conveying. $5000 2769 Diagram is a good representation of the plant. No surprises. Steam lines come off the top of all steam headers. Valves accessible. There is no measure of the steam flow to the reboiler on the debutanizer, E-30. $200 2771 Temperature of melt decreases to 223 C; injection time increases; cool time decreases to have a net effect of the same cycle time. Lumps persist in product. $2000 2774 Not needed. $20 000 2775 55 –1%. $2000 2777 Upstream 178 and downstream 128 C. $4000 2778 10 C. $300 2779 Why1? to produce crystals within specifications. Why2? to provide consistent and reliable product (for another unit) or for sale. Why3? so the company continues to prosper. Conclude: focus on Why1: “to produce crystals within specifications”. 2780 Temperature-sensor error/ other units sending process fluids that are colder than usual/ exit temperatures from E201 is lower than usual/ flowrate of gas from the reactor is slower/ catalyst degradation in the reformer with carryover and fouling of the shell side of the exchanger/ baffles loose and gas flow across tubes is < design/ pressure increase downstream causing less gas flow from the reactor/ inert gas blanketing the tubes on the shell side/ exchangers not vented before startup/ vapors condensing in the exchangers and causing a loss of area/ changes on the tube side. $300 2781 Mass of overhead pumped out by F1400 is 88% less than expected overhead. $50 2783 Mid-range. $200

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2784 On spec. No complaints from other customers for that batch. $350 2785 Water flowrate is slower than expected; liquid level seems to increase on tray 4. 2786 No upsets. We have had to supply more water since you started the new washing. $30 2787 Diagram does not show the water-cooled liebig condenser directly after the furnace with the vapor–liquid separator to remove the water via a barometric leg (as shown in the diagram); then a brined cooled Liebig condenser plus separator and barometric leg. After the dry reciprocating vacuum pumps the ketene vapor rises through a packed column where the ketene reacts with acetic acid to form acetic anhydride. The cracking furnace is fired by hydrogen. 2788 Nothing surprising. Indeed, the chiller operation is relatively self-contained so our interaction is minimal. 2789 No fouling. $5000 2790 Observe liquid holdup on the C2 splitter on trays 3 and above. No evidence of collapsed trays; just the top of the column seems flooded. ($5000 for scan + $1600 for time for scan + $20 000 for lost time to arrange for scan) $26 600 2791 Not that we can see on the outside. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 2793 Awkward but overhead product within spec most of the time. $500 2794 Well designed; should operate mid-position for design flowrate. $100 2795 Usual. $100 2797 OK. Recalibration not necessary. $2000 2798 See Chapter 3: valves, Section 3.1.3; pump, Section 3.2.3. $30 2799 No. The oscillations are erratic. $200 2800 Indicates when pump is operating as expected. However, no improvement in operation at this time. $500 2801 IS: for last 20 days. IS NOT: before then. $50 2802 Difficult because the water composition varies between 20 to 50%; no sampling has been done in the hot well and the system tends to fluctuate. Flowrate of live steam is not measured. 2803 We replaced this valve during the shutdown. It sparkles. $3000 2804 30 C. $300 2805 Vessel is insulated. Reactant is clear. Block valves, a drain and bypass valves are present for the steam control valve but are not shown on diagram. No drain valve is included with the steam trap. The feed flowrate of feed to the tank is measured upstream. $300 2807 Reads 3.9 MPa. $100 2808 Responds to change. $20 2809 See Chapter 3: distillation, Section 3.4.2, perhaps pertinent condensers and reboilers, Section 3.3.3; pumps, Section 3.2.3 and controllers, 3.1.1. $50 2810 Negligible fouling on the outside of the tubes.

529

530

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2811 Area 20% oversized. Coil supported along only one side of the vessel so that it won’t tear away when heated up. Thermal expansion anticipated and accounted for. $600 2812 All valves appear to function OK. $300 2813 End-suction centrifugal; 1800 rpm; sized to operate at design flowrates with appropriate allowances for Dp across control valve. $100 2814 Process pressure > atmospheric. $50 2815 Same as usual. $60 2816 Lower than expected. Lower than before the shutdown. $300 2818 Clean. $2000 2820 IS: after 2000 parts had been shipped. $50 2821 Checks that the steam flow available should be able to reach specification bottoms composition. $50 2824 Wide open. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 2825 Clear, no obstructions. $8000 2826 Should do the job. Neutralization reaction is rapid. Good mixing supplied by the rising gas from the spargers. $300 2827 See Chapter 3: adsorption-drier, Section 3.4.7; exchanger and thermosyphon boilers, Section 3.3.3; refrigeration loop, Section 3.3.4; sensors and controls, Section 3.1.3; pump. Section 3.2.3. 2830 Polymer is added to the feed to the filter to aid the filtration process. $20 2831 Steady and usual. $50 2832 Over the past few months we kept the PIC at 0.8 MPa. We did have a time when the load was small and we operated for a short time at 0.7 MPa. After that, however, we had to gradually increase the PIC settings for even the usual loads (where in the past 0.8 MPa would have handled the load well). $100 2834 Accurate; check out OK. $20 000 2836 OK. No calibration needed. $30 000 2838 IS: First operators. 2839 The diagram is reasonably correct in the overall layout. In addition, there are clean outs and flush on the pump. Isolation block valves on strainer and pump. A pressure gauge is on exit from the pump. A metering pump provides the polymer addition. Both thickeners have overflow launders, central feed and rotating central rake. There is a pressure gauge on the blowout air line upstream of the block valve. $50 2840 As measured on the gauges Dp= 243 kPa. $1000 2841 Temperature = 125 C. $600 2843 Mid-range. $300 2844 Equipment should do the job. $100 2845 Valve stem is settled in to be about 60% open and consistent with design value for 14 L/s. $30 2846 Same before and after. Bottoms flowrate = 1.38 kg/s –10%. $1200 2848 Steady. $120 2849 No change. Lumps persist. $8500

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2850 Equipment should do the job. Concern about NPSH and excessive pressure loss on the suction side. $800 2851 F/7 = 6.5 L/s and did read 7.8 L/s before the system was put on “safe-park”. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 2852 No change. Level in flocculation tank still below normal. $6000 2853 Horizontal pressure vessel; level control ensures that water is always maintained in the steam drum. Natural circulation lines feed water to the shell side of the waste-heat boiler. Traditional design. 2854 Does not sound like cavitation. 2855 The trays we can see are level; check on the clearance between downcomer bottom and tray suggests that tray should be sealed. $30 000 2857 Dp measured is ten times estimate. Suspect blockage in line from product sump to pump suction. 2859 Clear and clean. $3000 2861 IS: exit water temperature is 42 C; IS NOT: 70 C. $50 2863 Estimate seems OK. No major blockage. $300 2865 No change in swinging-loop phenomena. $100 000 2866 Negligible fouling; no condensed liquid; baffle spacing correct and baffles secure. $6000 2867 Designed to handle design flowrate; horizontal configuration; good baffle design with vertical windows; sealing strips used. Slight tilt to facilitate condensate drainage. Usual vents and drains. Fouling assumed 0.0003 m2 K/W for process fluid; 0.0002, for clean cooling water. $150 2868 Stroke same as specs. 2869 Several tubes have slight leaks. $20 000 2870 The solution leaving the reactor is 83%. This is concentrated to 99% in a downstream falling-film evaporator. This concentrated solution is then prilled to produce pellets. $50 2871 All five samples are the usual concentration. $4000. 2872 Call you, as instructed, since this is costing us megabucks! Then promptly try to close valve A in the hot loop and open valve A in the cold loop; I may have to put on the calrod startup heater if the cold loop gets too cold. $300 2873 Lumps persist. $40 000 2874 Not needed. $4000 2875 No water or ammonia in the sample. 2876 Yes. All valves appear to be closed on shell and tube side. $300 2877 Whistling sound of fluid flowing. $100 2878 Usual values; suggests that the damper is 1/3 closed. If didn’t put on safe-park as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 2879 No improvement. $500 2880 10 ppm. $4000

531

532

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2881 Wide open before “safe-park”. At “safe-park” mid-range. If didn’t put on safepark as first step, then dangerous potential fire/explosion conditions are created while you experiment. $500 000 2882 No upsets. No calls from the safety inspector. 2883 101.6 kPa $150 2885 (E114) Usual allowance for fouling on tube and shell side. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure > ethylene refrigeration pressure. $1500 2886 Higher than usual concentration. Particles are < 10 mm but cannot tell if there are many < 1 mm because of the optical limits of our analyzer. 2888 The piping is complex and consists of a range of 3-way and 4-way valves to allow any of the dryers to be on-line or regenerated. On-line is straightforward. The feed gas enters the top of the first dryer, exits the bottom, enters the top of the second dryer in series and then out the bottom to a knockout pot. Regeneration is more complex because two different sequential activities occur during regeneration. First, hot “town gas” goes through the bed and sends off the adsorbed water. When most of the water has been desorbed from the bed of alumina, the bed must now be cooled before it can be put back on-line. Town gas is used for both these functions. First, the desorption is done with “hot” town gas; then the cooling is done with “cooled” town gas. The town gas from both functions is returned to the fuel gas system. Case ’12 gives the details of a similar – but slightly different – system. Also this downtime costs more than Case ’12. $400 2889 Estimate: 7.6 m/ (1.3 m/s) = 5.75 s. $50 2890 No improvement. $980 2891 We already did this during the turnaround. Looks fine; minor or no adjustments needed. When we bring this back on line there is no change. $8000 2892 45 s; based on 1.5 s/0.1 mm wall thickness; slightly longer than usual because of the thickness of the product, especially the hub. $300 2893 Set to blow at 340 kPa. $200. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. 2894 Damaged spring; check valve partially blocking the flow. $500 2895 Agrees with design values –15%. 2896 IS: thermosyphon chiller output. IS NOT: elsewhere. 2897 Zero. $3000 2898 No change. $20 000 2899 Not needed. $30 000 2900 Not needed. $24 000 2901 No significant fouling; baffles in good condition, vortex breaker in place. $10 000 2902 Gradually increasing to 57 C from the usual 47 C. $50 2903 No. I sure wish I could. $300 2904 No improvement in operation. $8000

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2905 Concerned with the heavy maintenance caused because of the ketene dimerization. Would recommend the use of a liquid ring pump using acetic acid as the sealant. 2906 Wrong height of liquid in the seal pot/ compressor under surge or sonic conditions/ pressure sensor wrong/ control system fault/ excessive pressure in the flare line from upstream units/ kickback valve faulty/ error in matching seal-pot level with expected Dp/ plugged line from seal pot to flare/ not enough steam to the flare. $150 2907 IS: lab analysis of the overhead shows too much C4; in the bottoms is too much C3. IS NOT: anything else different. $50 2908 DT in reboiler = 198–(180 to 185) C = 13 to 18 C and hence probably nucleate boiling. 2909 Has tendency to oscillate. $50 2910 Head-capacity curve and NPSH data available. $70 2911 Yes, for this production rate provided water flow is usual and inlet water temperature < 20 C and exit water temperature is < 40 C. $50 2912 IS: independent of the operator. 2913 Temperature –5 C; 600 kPa abs. Latent heat = 438 kJ/kg. $75 2914 Sharp edged orifice facing correct direction, based on tab markings. 2915 Should work well. Baffles were in vertical. Vessel has air bleeds. $200 2916 No significant fouling; baffles in good condition, vortex breaker in place. $10 000 2917 Valve starts mid-range and gradually increases to almost wide open as the temperature drops and until the temperature reaches close to the high; then it returns to mid-range and the cycle repeats. 2918 IS: overhead pressure gauge P1 is too high, 2.7 kPa. IS NOT: 0.4 to 0.7 kPa. $30 2919 Reads 1.3 C. $500 2920 Hopper is 3/4 full. $1000 2921 No improvement. $800 2922 Reaction rate is very fast. $50 2923 No change. $20 000 2924 End suction, centrifugal pump; 1800 rpm; NPSH required at design flowrate = 1 m; mechanical seal. 2925 Yes, the leads are correct. $400 2926 Leaks in the overhead line, loss through the bottoms, incomplete condensation: condenser undersized, air flowrate too small, air too hot, air recirculation, inadequate liquid seal in condenser, dirty heat-exchanger surface, incorrect calculation. $800 2927 1725 rpm. $50 2928 Just started up. None. $50 2929 Shell and tube. Allowance for fouling on tube side = 0.0001 m2 K/W; shell-side steam= 0.00001. $930 2930 IS: pump only 2/3 design; level in reflux drum increases. IS NOT: pump design capacity and level in reflux drum constant. $50

533

534

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2931 No blockages; impeller key in place; correct clearances. No improvement in process operation. $2000 2932 Not needed. $2000 2933 Steady and at the design value. 2934 Not needed. $4000 2935 All correct. $20 2936 62 trays = 136 kPa. $200 2937 Dilution water could come in with any of the feed streams, from a leak in the cooling coil, condensate running back down from the overhead steam line, steam leak for the heater into the nitric acid in the storage tank and what about the heavy rains we have been having all week? $300 2938 Fail open. Direction agrees with flow direction. $200 2939 No change in operation. Pressures and temperatures still increasing. $250 2940 IS: control is erratic; acid flow may stop. IS NOT: steady control; acid flow as needed. $50 2941 For the naphtha pumps, the NPSH supplied in this installation >> than required. No air or water tests were done before starting up the unit. $500 2942 Controlled and steady. $200 2943 Diameter and length sized to give required residence time for mixing and reaction. OPRC sensor placed well downstream of “end of reaction”. $100 2945 No change. $200 000 2946 71 kg/s (compared with design of 100 kg/s). 2947 Level covers half of the tubes. 2948 Yes, reduce flowrates to 70 Mg/d unless you need it to run at high flowrates to get key operating data. $20 000 2949 Nothing that looks like it would vibrate or chatter. $4000 2950 OK. Recalibration not necessary. 2951 Estimate seems to agree with information from the unit. Suggests flowrates as expected for new conditions and same as previously. $600 2952 No changes over the past year. Still producing slightly superheated steam at 0.8 MPa g. $100 2953 Allowance for fouling on shell side, water-boiling = 0.0002 m2 K/W; tube-side process gas = 0.0005. An allowance of 10% more area than needed for design flowrate. Single pass on the tube side with unique baffle design for secondary part of the boiler. Simulation shows that this unit should do the job; it has done so in the past. 2956 Design method D-1a in Chin’s classic article in Hydrocarbon Process, Oct 1979, p 151. Controls the flowrate of coolant. $200 2957 End suction, centrifugal. 1800 rpm with impeller one size smaller than casing. NPSH data and head-capacity curve available. To prevent acid leakage through the packing seals, demineralized water is fed to the packing seals via lantern rings. The water is kept under pressure so that there is a Dp= 510 kPa across the seal. This means a small amount of water goes across the seal instead of a small amount of acid leaking out through the seal. $600 2958 10 C. $300

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2959 2960 2961 2962 2963 2964 2965 2966 2967

2968 2969 2970

2971 2972 2973 2974 2975 2976 2977 2978 2979

2980

Zero. $3000 Sounds as though it is working properly. Valve direction= flow direction. $200 Looks fine. $20 000 Not needed. $80 000 Non-ideal but data match theory. IS: especially on third effect but on all effects and after preheaters. IS NOT: feed inlet or steam inlet. $50 Plant operates as it should. Level in flocculation tank rises to “normal level”. $5500 Not needed. $20 000 Use the heat load gained by the steam to back calculate the process-gas temperature leaving the superheater. At design conditions this gives 600 C going to 530 C out. For actual conditions this gives 730 C going to 503 C out. These can then be used to estimate LMTD and thus UA. UA design= 21 MJ/s/ 226 = 93. UA actual= 68 MJ/s/ 232 = 293. For the higher temperatures on the actual conditions, the thermal properties will be higher; the steam-side coefficient will be 25% lower because the steam flowrate is 70% slower based on Re0.8. IS: negligible overhead product and bottoms 10% organic. IS NOT: design rate of overhead and bottoms concentration < 2% organic. See Chapter 3: vacuum, Section 3.2.2; distillation, Section 3.4.2; condensers, Section 3.3.3; reciprocating pump, Section 3.2.3. $60 Well designed based on pressure, residence time plus vortex breaker and demister pad. Simulation shows that unit was operating very close to design values for usual range of feedstocks and conditions. $150 Four months ago; routine cleaning of all heat exchangers including the wasteheat boiler and steam superheater. Takes about 2 days. Nothing to be seen. $40 000 1.7 MPa g and steady. Steady and usual value, 75 C. $50. Not needed. $20 000 No improvement in operation. Steam is still superheated. Gas temperature still very high. Warm, 20 C; humid. Rain two days ago. Warm and humid the rest of the time. $500 IS: lab analysis of the overhead shows propane concentration 212 times higher than design. IS NOT: anything else different No noise until the temperature starts to rise; then hear a high-pitched whistling that stops when the temperature hits the top of the cycle; followed by lower-pitched bubbling, then no noise. Cycle repeats. $6000 The flooded evaporator works as follows. Sufficient area is supplied so that when there is no condensate in the tubes, the area available for heat transfer can evaporate all the ethylene (for the butane temperature corresponding to the higher pressures). Usually the condensate fills the tubes so that the area is half the maximum. Thus, to increase the butane evaporation, the pressure

535

536

Appendix D Coded Answers for the Questions Posed to Solve the Cases

2981 2982 2983 2984 2985 2986

2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999

3000

PIC is increased and the valve on the condensate opens to drain the condensate out of the tubes. There is a knockout pot in the ethylene line to pipeline to remove any entrained liquid. The liquid is recycled to the storage tank. $50 “Process Design and Engineering Practice” D.R. Woods, Prentice Hall (1995) p 4-90 to 4-93 When: 6 months ago; routine checks done during the turnaround. $50 Column pressure decreases. Flare reduces. Impurities reported in pentane product. $5000 Agrees with head capacity. $200 None, this is startup of new plant. $50 4 months ago, cleanout buildup on the cooling coils in the reactor. $50. If you didn’t put plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000 When: 7 months ago. Nothing done on this section of the plant. $10 Soap tests inconclusive. $5000 Slightly more open than usual. $200 Well designed. Demister included. Globe valve on periodic drain to sewer. $150 18 C; cloudy and overcast; rain forecast. Cool and cloudy all week. $3000 See Chapter 3: distillation, Section 3.4.2; adsorption, Section 3.4.7; knockout pots, Section 3.5.1; heat exchangers, Section 3.3.3. $200 None available. In a sense this startup is the commissioning. Maybe next time! $650 Usual. $50 IS: increase flowrate above level. IS NOT: when lower flowrate used. Hot!!! $300. If you didn’t put the plant on SIS or SIS + evacuation before you asked this question, the plant explodes with loss of life. Penalty $3 000 000. Designed on condensate flowrate and for pressure differential across trap. Upstream strainer. No bypass. Discharge to header. $150 See Chapter 3: reactor, Section 3.6.2; furnace, Section 3.3.2; exchangers, Section 3.3.3; pump, Section 3.2.3; control and valves, Section 3.1.3. $200 (E113) Usual allowance for fouling on tube and shell side. Vent/blowdown lines on both the tube and shell side. Vent/blowdown lines discharge to sewer. Each vent line has a single gate valve that is normally shut. Process pressure > propylene refrigeration pressure. $1500 Oversized. This should easily do the job. $120

537

Appendix E Debrief for the Trouble-Shooting Cases

Cases ’1 and ’2 are discussed in Chapter 1. Cases ’3 to ’7 are posed in Chapter 1 with some answers in Chapter 4. Chapter 4 illustrates the process used by trouble shooters with varying degrees of skill. In this Appendix, we provide feedback about Cases ’8 to ’52. For each case is given: an estimate of the degree of difficulty, the cause, possible corrective action, and two sections related to the TS process: TS process hypotheses and TS process: possible diagnostic actions. The Estimate of difficulty is very subjective. The criteria I used in assigning the ratings were: 1) the ease with which Level 4 undergraduate engineering students had in “solving the problem’, 2) the ease industrial participants had with the cases and 3) the level of “experience with process equipment” required in solving the problem. For the latter, I used the self-test data from Chapter 1. If the experience with process equipment rating is below five, then the cases are rated as relatively easy. Ratings 7 to 9 were used where the experience with process equipment ratings were about seven to ten. In assigning the ratings I assume that the MSDS data, “More details about the process”, handbook data and guidelines for trouble shooting (from Chapter 3) are not needed by the trouble shooter. The trouble shooter knows this already. Nevertheless, this information is available for many of the cases, and should be used when working cases where you are uncertain about such information.

(courtesy of T.E. Marlin, McMaster University) Estimate of difficulty: 8/10. This gets a high rating for “experience with process equipment on process control and distillation”. Cause: No level control on the feed vessel, V29. The fact that the fault occurred because of the change in pressure was coincidental. Possible corrective action: Immediate: Re-establish the feed flow rate by: placing the TC, FC and LC controllers for the tower on manual. If enough material exists, place the column on safe-hold (total reflux) until the feed is started; place the con-

Case ’8: The depropanizer: the temperatures go crazy

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

538

Appendix E Debrief for the Trouble-Shooting Cases

troller FC-1 on manual with the valve partly opened; prime the pump; adjust FC-1 until the flows in and out of the feed drum, V29, are equal. After the flow has been re-established, the operator must closely monitor and control the level manually. In the longer term: The level L-1 should be controlled by adjusting the FC-1 setpoint. A cascade control system should be installed. Also, the feed flow rate has no low alarm. Although a SIS exists to start the backup pump, a FAL should be placed on the FC-1 measurement. In addition, the feed drum has no alarms. LAL and LAH should be placed on the LC-a measurement. TS Process: illustrative hypotheses: tray collapsed in the stripping section/ too much bottoms fed to the debutanizer/ too much overheads in the feed/ feed valve FV-1 stuck/ pump F-26 not working/ check valve on the idle pumps allows backflow/ no feed left in feed vessel V29. TS Process: possible diagnostic actions: 1666, 63, 563, 155, 78, 1436, 1765, 1502, 1947, 1610, 1834, 691, 1906, 2769, 1737, 1390.

Case ’9: The bleaching problem

Estimate of difficulty: 7/10. This is a relatively moderate rating because of the experience with process equipment required about pneumatic conveying. This was a design problem in that this mistake should not have occurred. Cause: No conveying air is getting to the inlet of the conveying line. Without air, there is nothing to transport the Fuller’s earth. Possible corrective action: Immediate action: either dump Fuller’s earth into the bleacher directly or insert air spargers into the hopper so that air is supplied to convey the solids. Longer term: redesign with fluidizing panels on the bottom cone of the hopper or install a concentric annular pipe that allows air to flow down the insert pipe to the pipe inlet. TS Process: illustrative hypotheses: insufficient vacuum/ pressure sensor wrong/ plugged conveying line/ valve not open/ no air to convey the powder/ no powder in the hopper/ powder is bridging in the hopper. TS Process: possible diagnostic actions: 2977, 2690, 2562, 59, 1368, 1097, 2431, 2920, 2630, 1206, 2767, 2334, 1856.

(based on Krishnaswamy and Parker, 1984) Estimate of difficulty: 8/10. This requires systems thinking and experience with process equipment for drying, screens and centrifuges. Many senior students have had little experience with equipment related to liquid solid separations. Cause: The washing cycles between the upstream screen and the centrifuge were not synchronized. As a result the centrifuge was getting no feed when the screen was being washed. During this time the crystal layer in the centrifuge under the scraper compacted. This resulted in a reduced filtration rate and higher crystal mois-

Case ’10: To dry or not to dry

Appendix E Debrief for the Trouble-Shooting Cases

ture content. This persisted until the wash cycle in the centrifuge. Figure E-1 (reprinted with permission from R Krishnaswamy and N.H. Parker, Chemical Engineering, April 16, 1984, p 93 to 98, copyright McGraw Hill) shows, at the top, the performance before the change when the washing cycles were synchronized. At the bottom is shown the performance after the change when the cycles were not in sync.

Figure E-1 The top shows performance conditions – before the change – for three pieces of equipment: for the screen the feedrate and the % of the feed over, as a function of time; for the centrifuge, the feedrate and discharge moisture, %, as a function of time and for the dryer, the feedrate, the % moisture in the feed and the % moisture in the discharge, as a function of time. The bottom picture shows the same data for the three pieces of equipment after the change.(Excerpted by special permission from Chemical Engineering (April 16, 1984) Copyright (2004) by Chemical Week Associates, New York, NY, 10038). This information is useful for solving Case ’10.

Possible corrective action: bring the washing cycles into sync. TS Process: illustrative hypotheses: not enough steam/ wash water carryover from the centrifuge/ cycle from screen not coordinated with cycle in centrifuge/ feed crystals too wet from the screen/ rotational speed of the dryer has increased/ more fines in the centrifuge causing the filter cycle to be too long; fines carryover to

539

540

Appendix E Debrief for the Trouble-Shooting Cases

the centrifuge causing blinding in the centrifuge/ centrifuge rpm faster than usual/ crystal size change; crystals change shape so that filtering is different/ centrifuge operates but has run out of feed/ dryer feed too cold/ vendor supplied faulty equipment. TS Process: possible diagnostic actions: 492, 4, 405, 816, 711, 1411, 603, 1009, 1041, 1804, 1971, 360, 1114, 1458, 909, 54, 1968, 1520, 196, 1705.

(courtesy W.K. Taylor, B. Eng. McMaster University, 1966) Estimate of difficulty: 5/10 Cause: The check valve on the exit line from pump B is faulty and allows liquid from pump A to recycle around when pump A is on-line and B is off-line Possible corrective action: Immediate action: close the block valve V205 on the exit line of pump B whenever pump B is idle. Longer term: replace the check valve. TS Process: illustrative hypotheses: instrument error in flowmeter/ motor fails to start when remote start button is pushed/ pump A has worn wear rings causing internal flow circulation/ on pump A the motor is turning backwards so that the impeller is turning in the wrong direction/ air lock in pump/ debris or stuff plugging the suction line of pump A/ reverse flow through check valve on pump B. TS Process: possible diagnostic actions: 930, 1470, 1998, 518, 550, 170, 102, 1482, 2027.

Case ’11: The lazy twin

(courtesy John Gates, B. Eng. McMaster University) Estimate of difficulty: 3/10. The key is to realize that the leak is from the high-pressure side to the low pressure side; identify the location where a hydrocarbon stream is on the high-pressure side. Cause: Leak in E131 from the “fuel gas” process side to the water on the tube side. The contaminated water goes to sewer. Possible corrective action: Isolate E131. Water pressure test and locate leaks. Plug the tubes that leak. TS Process: illustrative hypotheses: sampling error/ analysis error/ hazard coming from upstream plants/ leaks in vent valves from any location having “fuel gas”type materials upstream of the valve, e.g., the shell side of most exchangers/ leak in the valve on the knockout pot/ leak in the tubes from fuel gas to water in E131. TS Process: possible diagnostic actions: 2085, 2067, 1699, 2617, 1211, 2433, 119, 2424.

Case ’12: The drop boxes

Appendix E Debrief for the Trouble-Shooting Cases

(courtesy ESSO Chemicals) Estimate of difficulty: 4/10. This is relatively easy because the operator had identified the problem correctly. Cause: The temperature of the condenser effluent is controlled by varying the pitch of the fan. This is, indeed, a slow and clumsy method. The response to changes in ambient temperature meant a loss of effective control of this temperature. Possible corrective action: Immediate: run cold water onto the outside of the tubes. Longer term: this was resolved by installing a bypass control valve around the condenser to control the effluent temperature. Another option is to install a watercooled trim cooler. TS Process: illustrative hypotheses: instrument fault/ maldistribution/ hot-gas recirculation/ fouled tubes/ blade wrong pitch/ fan not working/ insufficient tube area/ buildup of non-condensibles in the bottom row of tubes/ tubes not sealed/ no vent break/ poor control system/ control system not well tuned. TS Process: possible diagnostic actions: 2179, 1613, 220, 89, 2152, 487, 367, 2793.

Case ’13: The lousy control system

Case ’14: The condenser that was just too big

Estimate of difficulty: 3/10. Cause: A vapor lock occurs in the suction line to pump F1400 because there is no vent-break line connecting the inlet to the pump to the inlet to the booster ejector. Such a vent line needs to be designed carefully to prevent liquid from going up the vent line instead of into the suction of the pump. The vertical height at the entrance to the booster ejector line must be higher than the difference in height of level of the entrance to condenser E1400 and the pressure difference between the exit from exchanger E1400 and the entrance to the booster ejector. Possible corrective action: Install vent-break line as illustrated in Figure E-2. TS Process: illustrative hypotheses: vacuum leak/ variation in steam pressure to booster or other ejectors/ liquid not subcooled enough and it is flashing in pump F1400/ wet vacuum pump not pumping at capacity/ leak in exchanger E1400 causing water to flow into vacuum system/ instrument error/ more volatile material in feed/ vapor lock in suction line because of no vent break/ leg from barometric condenser not sealed/ booster not working right and air is sucked down from the booster into the pump F1400/ dry vacuum pump F14 put into service without 30-minute warmup. TS Process: possible diagnostic actions: 400, 2273, 2377, 390, 486, 188, 871, 2511, 2781, 2580, 1317, 1020, 1594, 2201, 326, 242, 2369, 2939.

541

Figure E-2

Feed

Fatty acid

TI

to/from Dowtherm system

240ºC to 260ºC

TI

240 kPa

PI

120ºC

TI

Boiling Condenser

Cooling Water

Coil in tank

The vent-break system for Case ’14.

Cooling water exit

Cooling water

TI

60ºC

Backup Condenser

reciprocating pump

PI

0.4 kPa

Cooling water return

PI

barometric condenser

vacuum break line needed

Booster

steam

wet vacuum pump

well water

542

Appendix E Debrief for the Trouble-Shooting Cases

Appendix E Debrief for the Trouble-Shooting Cases

Cases ’15 to 18 are discussed in Appendix C.

(supplied by Mike Dudzic, B. Eng. 82, McMaster University) Estimate of difficulty: 2/10. Cause: The strainer is plugged. Possible corrective action: Clean the strainer. Long term: “ensure clean the strainer” is included in operating procedures and training. TS Process: illustrative hypotheses: inadequate water flush/ flush does not go into the correct lines/ pump stopped/ strainer plugged/ line plugged and cannot be cleared with the flush/ flush not working/ block valves leaking/ valves around pump closed/ pump clogged. TS Process: possible diagnostic actions: 1635, 1330, 2329, 2052, 1258, 322, 793, 2672, 1855, 1576.

Case ’19: The Case of the reluctant belt filter

Case ’20: The case of the fussy flocculator pump (courtesy Jonathan Yip, B. Eng. 97, McMaster University) Estimate of difficulty: 3/10 Cause: Damaged check valve that didn’t open properly combined with faulty location of the pressure gauge. Possible corrective action: replace the check valve and relocate the pressure gauge before the check valve. TS Process: illustrative hypotheses: pressure gauge reading wrong/ cavitating pump/ wear rings worn causing internal circulation/ impeller worn/ blockage in line/ change in density of liquid so that pressure gauge reading converts to a different value of “head”. This fault happened gradually so that it does not seem to be related to maintenance or other changes. Need to visit the site and see the pump. Locate the head capacity chart from the vendor. TS Process: possible diagnostic actions: 2525, 2322, 2343, 2243, 1342, 2764, 1099, 2506, 2894, and 2965. If a dominant J trouble shooter (as described in Section 6.1.3c) decided to change things and fortuitously selected diagnostic action 2965, then the “problem” would have been solved. However, a few simple tests and good use of a head-capacity curve and fundamentals will remove the fortuitousness. Test 2506 is very telling because, although the answer is relatively inaccurate, the power (and flow) is less than expected. Another clue is the slow response to action discovered by action 2764. By this time, if not earlier, the trouble shooter should realize that the pressure gauge is downstream of the check valve and thus is not reflecting the true response at the exit flange of the pump.

543

544

Appendix E Debrief for the Trouble-Shooting Cases

(courtesy Mark Argentino, B. Eng. 1981, McMaster University) Estimate of difficulty: 3/10. The solution does not require extensive knowledge about processing equipment. The key is second-year fluid statics. Cause: Off-specification “kerosene” with density 1.02 was used in the seal pot instead of the usual specification-grade kerosene of density 0.8. Possible corrective action: Use specification-grade kerosene. TS Process: illustrative hypotheses: wrong height of liquid in the seal pot/ compressor under surge or sonic conditions/ pressure sensor wrong/ control system fault/ excessive pressure in the flare line from upstream units/ kickback valve faulty/ error in matching seal-pot level with expected Dp/ plugged line from seal pot to flare/ not enough steam to the flare. TS Process: possible diagnostic actions: 2641, 1505, 1395, 1451, 9, 1497, 1842, 1974, 2610, 70, 2487, 1284, 1590. Another skilled trouble shooter used, in addition, 471, 500, 1511, 694, 410, 1240.

Case ’21: The case of the flashy flare

(courtesy of Scott Lynn, Chemical Engineering, University of California, Berkeley) Estimate of difficulty: 4/10. This is a basic design flaw in installing a pump. The control valve should never be put on the suction side, especially for a liquid with a high vapor pressure such as hydrochloric acid. Cause: The control valve is on the suction side of the pump. The pump cavitates. Possible corrective action: Move the control valve to the discharge side. Ensure that there is a vent on the storage tank and that the acid is injected into the top of the waste line to prevent waste from flowing back through the pump. Install a check valve in the pump-discharge line. TS Process: illustrative hypotheses: sensor fault/ poorly tuned control system,/ sticky control valve/ control-valve hysteresis/ caustic waste flows backwards into the pump/ electrical interference with the control system/ cavitating pump/ no vent on the storage tank; vacuum created in the storage tank/ density of acid > expected and motor overload. TS Process: possible diagnostic actions: 1425, 1191, 2509, 1290, 2002, 1068, 1835, 1682, 1150, 2129, 2467, 700.

Case ’22: The pH pump

(based on Barton and Rogers, 1997) Estimate of difficulty: 4/10. The hazardous nature of this situation should be recognized quickly. Cause: The supplier of the TDI did not have benzoyl chloride as a reaction modifier. The reaction ran away, the pressure built up to 3.4 MPa g and the reactor ejected the batch. If, for any reason, the relief system failed, the reactor would have

Case ’23: The hot TDI

Appendix E Debrief for the Trouble-Shooting Cases

exploded. Without the modifier present, the exothermic reaction ran away, solid product formed causing the stirrer to stop. The high temperature caused gas formation and pressure buildup. Possible corrective action: add reaction modifier or return to the previous supplier of TDI or redesign the coolant system and the operating procedures to prevent the run away. TS Process: illustrative hypotheses: emergency/ action to prevent explosion. Yes, there are a variety of hypotheses to explore later to discover why the temperature runaway: cooling-water failure/ coolant-inlet temperature too high/ cooling-surface fouled/ addition too fast/ operating procedures not followed/ sensors wrong/ different supplier of TDI/ contamination of TDI in the storage tanks/ different polyol/ contamination of polyol in the storage tanks or lines. TS Process: possible diagnostic actions: 2034 perhaps 1739. Later, 1195, 2289, 2893, 1201, 1422, 1986, 2188, 1743, 2824, 723, 1837, 709.

(courtesy John Gates, B. Eng. 1968, McMaster University) Estimate of difficulty: 4/10. The unit operations involved are familiar. Although there are two faults, they should be relatively easy to spot. Cause: Two causes: 1) the town gas is off-specification because it is picking up moisture from the slugs of water left over from the hydraulic testing of the line and 2) high-pressure steam is leaking into the town gas from heater 130. The town gas saturates the dryers during the regeneration stage and not sufficient moisture is removed during the cooling phase. The result is that the dryers are adding water to the process gas, instead of removing water. The result is that gas hydrates are solidifying in the column, in particular near the top of the demethanizer and the top of the C2 splitter where the pressures are high and temperatures are below 0 C. The terminology used in the industry is “icing up”. Possible corrective action: Three things to correct: 1. removal of the hydrates to keep the column going. 2. stop the steam leak and 3. insist that the town gas supplier give town gas with < 6 ppm moisture. The first two are directly under your control. Eventually, at the next turnaround the third problem can be solved by using our own fuel gas (town gas) generated on site instead of relying on an outside supplier. More specifically: 1). About two m2 of methanol was pumped into the reflux lines of each column affected. The methanol came down the column and eventually was recycled with the ethane to the cracking furnace. Fortunately, a single treatment cleared up the problem and the column could operate at 150 Mg/d with acceptable Dp. 2) The leaking tubes in heater E130 were plugged. A new bundle was prepared for the next turnaround and another spare was ordered for the next possible recur-

Case ’24: Low production on the ethylene plant

545

546

Appendix E Debrief for the Trouble-Shooting Cases

rence. 3) Samples were taken of the trends in the town gas. We met with the vendor to resolve the lack of specs on the town gas provided. TS Process: illustrative hypotheses: moisture coming in because the dryers are not working/ steam leak in the reboiler on the bottoms of the de-ethanizer/ town gas coming in too “wet”/ steam leak from E310 into the town gas/ process gas entering the site is too wet/ column was not designed to handle this high a flow/ collapsed trays. TS Process: possible diagnostic actions: 2948, 257, 2130, 2501, 280, 501, 2936, 2112, 1214, 1377, 2432, 2840, 2043, 559, 1912, 1767, 892, 527, 2050, 2545, 1874, 348.

(based on Yokel, 1983) Estimate of difficulty: 4/10. Although this is a larger system than some of the earlier cases, the cause should be relatively apparent from the typical symptoms given in Chapter 3. Cause: Under the new conditions the concentration of hydrogen in the reactor effluent is greatly reduced. Since the thermal conductivity and heat capacity of hydrogen are five to ten times higher than most other vapors, there is a significant reduction in the heat transfer because of the reduced concentration of hydrogen even though the flowrates are unchanged. Possible corrective action: recalculate the ratings of the exchangers to see if anything can be done to alter the current exchangers; if not, look for other sources of heat or install additional exchangers. TS Process: illustrative hypotheses: temperature-sensor error/ other units sending process fluids that are colder than usual/ exit temperatures from E201 is lower than usual/ flowrate of gas from the reactor is slower/ catalyst degradation in the reformer with carryover and fouling of the shell side of the exchanger/ baffles loose and gas flow across tubes is < design/ pressure increase downstream causing less gas flow from the reactor/ inert gas blanketing the tubes on the shell side/ exchangers not vented before startup/ vapors condensing in the exchangers and causing a loss of area/ changes on the tube side. TS Process: possible diagnostic actions: 2111, 2404, 356, 922, 686, 2418, 2816, 2503, 2059, 2212, 1751, 2294, 1807, 1956. Another trouble shooter used the following sequence: 473, 2111, 922, 874, 832, 951, 686, 2418, 2816, 2503, 1807, 1550, 2383, 1555.

Case ’25: The case of the delinquent exchangers

Case ’26: The case of the drooping temperatures (Used with permission from T. E. Marlin, McMaster University). Estimate of difficulty: 5/10. Cause: Insufficient air is being supplied to give the correct fuel-air mixture. Hence, the flame temperature decreases; less heat transfer and a resulting lower

Appendix E Debrief for the Trouble-Shooting Cases

temperature in the hydrocarbon stream leaving the furnace. As the temperature decreases, the temperature controller increases the fuel flow rate. This makes the situation worse and creates a hazardous situation with a lot of excess fuel at a high temperature. The excess natural gas may accumulate in the furnace and leave the furnace with the flue gas. It’s hazardous because there might be an explosion along the furnace where the fuel-oxygen mixture is high. In particular, if the furnace has a negative draft and air leaks in through the observation holes then flashes or hot spots could occur here. Possible corrective action: Immediate: either reduce the flowrate and return to safe-park conditions where complete combustion occurs or activate the SIS to stop the furnace. 2) use theoretical calculations to determine the required air flow for a furnace efficiency of 85 to 90% and 10% excess air. Daily inspection is recommended. Longer term: install a sensor to measure the percentage oxygen in the flue gas and this could be used by a feedback controller by adjusting the setpoint on the air flow controller FC-5 to adjust the damper close or open as needed. This is a very serious situation. For insurance purposes, an audit would identify this as an insecure condition, request improvement and, in the meantime, follow up closely. The insurance premium would also increase. TS Process: illustrative hypotheses: sensors at the outlet from the fired heater is wrong, T3/ sensor fault F2/ heat-transfer area too small/ heat-transfer area fouled/ poor tuning of controller/ air flow too small to support combustion/ flameout/ gas velocity on outside too small/ liquid-fluid velocity on inside too small/ decrease in thermal properties of process fluids/ process fluid flow increased. TS Process: possible diagnostic actions: 340 and then 955, 2851, 2542, 1225, 1092, 1452, 2191, 2101, 2881, 381, 2878, 1279, 171, 443, 820, 2301, 2574, 2077, 929.

Case ’27: The IPA column

Estimate of difficulty: 5/10. Cause: There is no vent break on the exit liquid header. The condensate periodically syphons out of the condenser and leaves the tubes without a liquid seal. IPA goes through the condenser uncondensed until enough condensate can build up to seal the tubes again. Possible corrective action: Put in a vent break TS Process: illustrative hypotheses: leaks in the overhead line, loss through the bottoms, incomplete condensation: condenser undersized, air flowrate too small, air too hot, air recirculation, inadequate liquid seal in condenser, dirty heat-exchanger surface, incorrect calculation. TS Process: possible diagnostic actions: 2134, 1501, 1685, 521, 1291, 18, 338, 239, 106, 1879, 1135, 152, 1259, 1517, 1916.

547

548

Appendix E Debrief for the Trouble-Shooting Cases

(based on a case supplied by P. L. Silveston, Chemical Engineering Dept., University of Waterloo) Estimate of difficulty: 5/10. Cause: Inert gas coming in with the flash steam gradually builds up in the exchanger and blankets off some of the tubes. Possible corrective action: vent the exchanger periodically or install a deaeration unit in the steam upstream at the ethyl acetate plant. TS Process: illustrative hypotheses: condensate buildup/ trap malfunction/ sensor error/ steam pressure too low/ inerts in the steam/ inert buildup in the exchanger/ fouling of the tubes on the outside/ fouling of tubes on the inside/ steam superheat too high/ steam flowrate < design/ water flowrate > design. TS Process: possible diagnostic actions: 1237, 1788, 2199, 2699, 108, 1329, 323, 177, 477, 364, 876, 617, 1632, 1533, 388. Case ’28: The boiler feed heater

Case ’29: The case of the reluctant reactor (courtesy of W.K. Taylor, B. Eng. McMaster 1966) Estimate of difficulty: 5/10 Cause: Valve A was too small. Even though it was wide open during startup, still not enough gas was bypassing the reactors on startup. Too much gas was forced through both reactors during the heating step and, under these conditions, the temperature will not rise above 325 C. Possible corrective action: In the long term the correct size of valve should be installed. In the short term, to reduce the flow of gas to the reactor loops, shut down one compressor and provide loop gas to both North and South loops from one compressor. With the gas flow to each reactor cut in half, the reactors heated up rapidly and in control. When the reaction temperature had been achieved in both reactors, the second compressor was started up and normal operations began. TS Process: hypotheses: temperature-sensor error/ power failure/ cal rod heater failure/ gas flow too high/ refrigerant condenser too cold/ cooling-water condenser too cold/ hydrogen concentration < 64%/ startup instructions not followed/ startup pressure too high. TS Process: possible diagnostic actions: 1011, 532, 1459, 1412, 1117, 1048, 1546, 1815, 1946, 1664, 2558, 1564, 1362, 1073, 1085, 660, 640.

(courtesy of ESSO Chemicals) Estimate of difficulty: 6/10. Cause: Undersized valves on both the liquid block around the pump and the control valves. No one checked the system resistance requirement when the valves were replaced; the appropriate pump was not installed.

Case ’30: The case of the reluctant reflux

Appendix E Debrief for the Trouble-Shooting Cases

Possible corrective action: Immediate: reduce the condenser duty to keep the reflux drum from filling up. This will change the top specs unless a new set of operating conditions are created. If this is not effective, shut down the plant. TS Process: illustrative hypotheses: pump cavitating/ instrument error/ poorly tuned controller/ plug in line/ blockages in condenser/ motor electrical leads crossed/ valve stuck shut or only partly open. TS Process: possible diagnostic actions: 1058, 1137, 993, 460, 752, 1864, 1606, 1513, 1985, 2071, 2213, 236, 2664, 2611.

Case ’31: The ethylene product vaporizer (courtesy of C. J. King, Chemical Engineering Dept., University of California, Berkeley) Estimate of difficulty: 6/10. Cause: Steam leak; steam then condenses, rises in the boiler and blocks off the effective area for heat transfer. Possible corrective action: Isolate the steam heater. Pull the bundle. Water pressure test to locate the leaks. Plug off the leaking tubes. TS Process: illustrative hypotheses: buildup of inerts in the butane/ butane level wrong/ leaks in the header/ faulty temperature shut-off switch/ impurities in the butane/ vapor lock on ethylene evaporation inside tubes/ sluggish flow up and down for butane vapor/ unstable flow with most of the butane going up the first opening and down the second/ ethylene coming in subcooled/ ethylene pressure > expected/ loss of butane. TS Process: possible diagnostic actions: 1359, 2427, 2248, 2952, 2528, 2603, 2807, 2549, 2109, 2407, 2498, 331, 1549, 1745, 1257, 105, 353, 2390, 2469, 2919, 2572, 881, 958, 252, 1812, 2869.

(courtesy T. E. Marlin, Chemical Engineering, McMaster University) Estimate of difficulty: 6/10. Cause: Increase in feedrate to the column. Possible corrective action: Immediate: decrease the feed flowrate. Longer term: increase the area for heat exchange for the condenser or, perhaps, increase the operating pressure in the column. TS Process: illustrative hypotheses: false alarm/ pressure sensor error. TS Process: possible diagnostic actions: 2227, 2304, 2011, 2461, 1724, 1870, 1152, 1357, 1052, 1408, 1323, 1173, 910, 802, 258, 13, 359, 2769, 858, 2103, 2054.

Case ’32: The alarming alarm

549

550

Appendix E Debrief for the Trouble-Shooting Cases

Case ’33: Chlorine feed regulation (courtesy of Scott Lynn, Chemical Engineering Department, University of California, Berkeley) Estimate of difficulty: 6/10. Cause: Excessive residence time between the redox-potential sensor and the feed point. Possible corrective action: Immediate: operate on manual. Longer term: relocate the sensor closer to the feed point. TS Process: illustrative hypotheses: fluctuation in caustic addition/ fluctuation in water flow/ plugged chlorine inlet pipe/ vibration in the chlorine inlet tube/ poorly tuned control system on chlorine addition/ poorly tuned controller for caustic addition/ pressure variation in vapor space in the chlorine storage tank/ variation in temperature outside the chlorine storage tank/ welder working close to the chlorine storage tank/ oxidation-redox sensor error/ sensor location wrong. TS Process: possible diagnostic actions: 2943, 2546, 1167, 1379, 1096, 1019, 781, 981, 2889, 225, 99, 2848, 2609, 345, 176, 1349, 1446, 1249, 2299, 2403, 1561, 1931, 267, 2261, 1209.

Case ’34: The cement plant conveyor

Estimate of difficulty: 6/10. Cause: Under the new conditions in the baghouse, the fines were being removed from the blend just before the product was packaged. Possible corrective action: Change bags or filtering conditions so that fines are not removed. TS Process: illustrative hypotheses: humidity causing particle clumping or bridging/ bridging in the feed hopper/ conveying process grinds the particles/ mixing process grinds the particles/ large-size particles missing from product/ fines missing from product/ new filter cloth lets fines escape/ fines hang up in packaging hopper/ fines fluidize in the packaging hopper. TS Process: possible diagnostic actions: 5, 516, 538, 504, 1260, 2075, 1312, 1078, 60, 2886.

Case ’35: The cycling triple effect evaporator

Estimate of difficulty: 6/10. Cause: single steam trap serving all three effects instead of a separate trap for each effect. The pressure in the condensate main= highest pressure connected to it. The condensate will only drain from this first stage; it builds up in the other stages until the heat transfer drops and the pressure increases to a level that will allow drainage. Then, the second effect drains suddenly; the pressure drops and the heat transfer increases dramatically, increasing the flowrate to the 3rd effect and dropping

Appendix E Debrief for the Trouble-Shooting Cases

the pressure in the first effect. This cycling is shown in levels, temperatures, pressures and flows. Possible corrective action: Put a separate condensate trap on each effect. TS Process: illustrative hypotheses: fouled tubes/ blockage/ instrument error/ control system poorly tuned/ flooded steam traps/ wrong type of trap/ inerts in the steam/ poorly tuned control system/ fluctuations in the concentration in the feed/ fluctuations in the feedrate/ fluctuations in the vacuum system/ steam to the ejectors is below design pressure. TS Process: possible diagnostic actions: 2746, 2489, 2344, 2147, 2450, 1028, 1228, 1278, 1328, 1378, 1476, 509, 44, 139, 1304, 281, 431, 860, 2108.

(courtesy W. K. Taylor, B. Eng. 1966, McMaster University) Estimate of difficulty: 7/10. Cause: Silica from the catalyst in the secondary reformer is vaporized and carried over to deposit on the tubes of the waste-heat boiler. Possible corrective action: Immediate: clean the tubes regularly. Longer term: lower the operating temperature in the reformer or find a new catalyst that is more robust. TS Process: illustrative hypotheses: flow blockage/ fouling of waste-heat boiler tubes/ temperature or flow instrument error/ steam generation is < design/ controller fault/ leak/ bypass not entirely closed/ insufficient water in the steam drum. TS Process: possible diagnostic actions: 2971, 37, 428, 1742, 1958, 1538, 2967, 2526, 2946, 2599, 1134, 2120, 2495, 584, 2360, 2083, 2647.

Case ’36: The really hot case

(courtesy Don F. Fox, B. Eng. McMaster University) Estimate of difficulty: 7/10. Cause: upsets upstream combined with a plugged sampler on the feed so that incorrect feed concentration is recorded (400 ppm instead the correct value of 700 ppm). The feed flowrate remains constant but the suspended solids concentration fluctuates in the feed. The rake turning rate is too slow and does not move the expectedly large concentration of solids into the central takeoff for the sludge pump. Ratholing occurs. Possible corrective action: Immediate: shut off the rake and the sludge pump. Then increase the rake speed until the speed is consistent with 0.2 kg/s solids; start the sludge pump. When the sludge is settling and the rate of raking = rate of removal the rake torque should read close to zero. The operator needs, in the interim, to watch feed conditions with low suspended solids. Then, for light loading of solids, the sludge pump should be stopped to allow the sludge to build up for a short time until the torque shows an increase. Then the sludge pump is turned on. Case ’37 The mill clarifier

551

552

Appendix E Debrief for the Trouble-Shooting Cases

Longer term: better sampling technique, so that the sampler is not plugged and so that the operators have a more accurate indication of the actual concentration of suspended solids. Need to identify and correct upsets that occur upstream. TS Process: illustrative hypotheses: upsets upstream/ rake too fast/ ratholing in thickener/ faulty design of thickener/ error in Parshall flumes/ incorrect flume design/ sampler error/ sampling at wrong location/ sampler line plugged/ velocity in sampler tube too low so that it gets a faulty sample/ excessive water flow because of rain this past week/ sludge pump takeoff nozzle poorly located/ damaged sludge rake. TS Process: possible diagnostic actions: 1656, 1959, 493, 47, 204, 403, 703, 579, 861, 1310, 1154, 1133, 1752, 2373, 1580, 2107, 2353.

(courtesy T. E. Marlin, Chemical Engineering, McMaster University) Estimate of difficulty: 7/10. Cause: The cooling water entering the battery limits is too hot in the summer. As a result, the heat transfer is reduced and the concentration of propane in the product increases. Possible corrective action: Immediate: direct water onto the shell of the exchangers using a fire hose or reduce plant production or increase the pressure in the column (if this is possible) or live with the trouble. Longer term: repipe the coldest water to this unit or increase the area in the condenser or add a heat exchanger using refrigeration. TS Process: illustrative hypotheses: sample error/ temperature or pressure sensor error/ increased concentration of very light components in the feed/ reduced tower pressure/ high feed rate/ fouled heat exchanger/ reduced flow of cooling water/ increase in inlet temperature of the cooling water. TS Process: possible diagnostic actions: 2831, 1633, 789, 513, 552, 1352, 1908, 2405, 2504, 2516, 2008, 2769, 1512, 1558, 2911, 1012, 2174, 424, 774.

Case ’38: More trouble on the deprop!

(from D. R. Winter, Universal Gravo-plast, Toronto, 2004) Estimate of difficulty: 7/10 Cause: the new batch of pearlized additive sent by the supplier was faulty. Possible corrective action: The supplier sent a fresh batch of pearlized additive. TS Process: illustrative hypotheses: Dirty machine/ dirty hopper/ moist feed/ too many volatiles in feed/ lubricant or oil on mold/ incorrect mold lubricant/ feed contaminated during material handling/ faulty raw material from supplier/ poor shutdown procedures/ screw rpm too high/ sensor error/ shutoff valve dirty or clogged/ injection pressure too high/ gates too small.

Case ’39: The case of the lumpy sunglass display

Appendix E Debrief for the Trouble-Shooting Cases

TS Process: possible diagnostic actions: 2272, 1983, 1445, 911, 136, 2738, 2521, 2310, 1934, 1545, 1857, 1183, 1084, 730, 594, 92, 425, 1945, 2087, 2724, 2575, 2771, 1662, 1718, 1624, 2359.

Case ’40: The cool refrigerant (courtesy T. E. Marlin, Chemical Engineering Department, McMaster University) Estimate of difficulty: 7/10. Cause: Over-specified control system. The heat transfer in the two exchangers depends on the areas and the refrigerant temperature. Since both of these are apparently fixed, only extreme good luck would result in both process temperatures attaining their desired values. The process operation must be changed to increase the heat transferred in E100 and to reduce the heat transferred in E101. Thus, an extra degree of freedom is required. F 1

H.P. Steam PC 1

K.O. Drum L 1

T 2

periodic flow

C.W. steam turbine

L 2

compressor

T 5

LC 3 T 6

E 100

chilled process

warm T 7 LC 4

E 101

Figure E-3

Automation of the intermediate corrective action for Case ’40.

T 8

chilled

553

554

Appendix E Debrief for the Trouble-Shooting Cases

Possible corrective action: Immediate: The set point of each level controller can be adjusted to achieve the desired process temperature. Level L4 will have to be decreased. Level L3 will be increased. If Level L3 is already at its maximum, the refrigerant pressure P1 would have to be reduced. The actions could be automated with cascade designs T6 ﬁ L3 and T8 ﬁ L4. This is shown in Figure E-3. Longer term: a design change to provide independent changes for the temperature of the refrigerant for each exchanger. Control valves could be installed in the vapor lines from each exchanger. By adjusting the valves, the pressure of the refrigerant in each of the exchangers can be set independently. Use cascade designs T6 ﬁ v1 and T8 ﬁ v2 as illustrated in Figure E- 4. F 1

H.P. Steam PC 1

K.O. Drum L 1

T 2

periodic flow

C.W. steam turbine

L 2

compressor v1

warm process T 5

LC 3 T 6

E 100

chilled process

v2

warm T 7 LC 4 T 8

E 101

Figure E-4

chilled

Design change for the long-term correction for Case ’40.

TS Process: illustrative hypotheses: exchangers fouled/ temperature sensors wrong/ exchanger design fault/ faulty control design/ not enough area/ level control fault/ buildup of inerts in E101. TS Process: possible diagnostic actions: 2708, 2157, 1734, 190, 740, 1140, 69, 362, 449, 944, 643, 713, 1369, 1046, 1809, 1932, 2338, 2783, 2843, 2583, 688, 124, 94, 2389, 130.

Appendix E Debrief for the Trouble-Shooting Cases

(courtesy T.E. Marlin, Chemical Engineering Department, McMaster University) Estimate of difficulty: 7/10. Cause: Fouling on the water side because the pressure controller has been adjusting the cooling-water flow rate, and, at times, the flow has been too low, the coolingwater temperature exceeded 50 C. The initial pressure control was poor because the control system was poorly selected: control of the pressure based on the dynamics of the condenser can be a slow process. Possible corrective action: Immediate: try to prevent too high a pressure in the column: reduce the column feed rate and column overhead rate or reduce the reflux flow rate and thus reduce the overhead vapor rate. The latter gives an increase in the impurity of the overhead product. Longer term: clean the exchanger and redesign a control system that prevents overheating of the cooling water. An example is shown in Figure E-5 where the valve on the exit of the exchanger is manipulated to change the level of condensed liquid in the condenser. This design gives relatively fast response and should result in better pressure control.

Case ’41: The ever-increasing column pressure

M.P. steam

Figure E-5

Improved control for Case ’41.

555

556

Appendix E Debrief for the Trouble-Shooting Cases

TS Process: illustrative hypotheses: sensors wrong/ cooling-water control valve malfunction/ poorly tuned pressure controller/ condenser underdesigned/ coolingwater feed too hot/ fouled condenser tubes/ inerts on shell side/ condenser not vented when started up/ restriction in condensate line to reflux drum and condensate building up in condenser/ cooling-water flow too low/ feed composition in light components increased/ overhead temperature changed. TS Process: possible diagnostic actions: 546, 2205, 1109, 1313, 1299, 1619, 1454, 1119, 1753, 2631, 1123, 1371, 231.

Case ’42: The case of the weak AN (courtesy W. K. Taylor, B. Eng. 1966, McMaster University) Estimate of difficulty: 7/10. Cause: the packing seal in the acid feed pump is worn and large amounts of demineralized water being fed to the seals enter the acid. Normally the flow of water goes to the outside of the packing seal to prevent acid from leaking out into the pump. The pressure difference between the water side of the seal and the acid side is 500 kPa. Normally the amount of water going through the seal into the acid feed is so small that there is no detectable change in the acid concentration. However, the seal is so worn that the water flow is large; the acid is diluted; the reactions are decreased. Possible corrective action: replace the packing seal and adjust the clearances. TS Process: illustrative hypotheses: temperature-sensor error/ sampling error/ analysis error for AN/ cooling water leak into the neutralizer-reactor/ ammonia vapor feed is moist/ urea plant off-gas has too much water/ rain leaked into the nitric acid storage tank and the acid in the storage tank is dilute/ steam generated in the reactor-neutralizer condenses and runs back into the reacting liquids/ poor mixing in the reactor/ increased flow of cooling water/ cooling water inlet temperature colder than usual. TS Process: possible diagnostic actions: 1082, 1440, 880, 2957, 2749, 2105, 2306, 1306, 1149, 1050, 750, 570, 448, 263, 1829, 1735, 698, 2233, 2466, 2775, 2581, 2321, 1485, 2155.

Case ’43: High pressure in the debut! (courtesy T. E. Marlin, Chemical Engineering Department, McMaster University) Estimate of difficulty: 7/10 Cause: the buildup of non-condensable propane in the debutanizer overhead condenser. The propane had come through from the upstream depropanizer during its upset. For the debutanizer overhead condenser, the area was insufficient or the pressure not high enough or the cooling-water temperatures or flowrates were insufficient.

Appendix E Debrief for the Trouble-Shooting Cases

Possible corrective action: lower the column pressure by reducing the reboiler duty and/ or ask the operator to open the “vent to fuel” manual valve on the condenser for a few minutes to purge the condenser. Long term: no action is needed. TS Process: illustrative hypotheses: faulty pressure-relief valve/ wrong setting on the relief valve/ pressure sensor wrong/ flowrate of cooling water too low/ plugs in the cooling water line/ fouled tubes/ area in the condensers too small/ control valve on condensate stuck shut and condensate floods condenser/ inert vapor in the condenser/ air not purged from the condenser during startup/ excessive boilup. TS Process: possible diagnostic actions: 365, 994, 1491, 1102, 2401, 2902, 2570, 2505, 2443, 278, 22, 458, 202, 960, 735, 853, 1301, 2769, 1605, 813, 2983.

Case ’44: Reactant storage

Estimate of difficulty: 8/10. Cause: Five faults: 1) no mixer installed to provide convection for good heat transfer; 2) contaminated steam because the steam line comes off the bottoms of the header; 3) waterlogged coil because the steam comes in the bottom of the coil, 4) flooded steam trap because the condensate line “into the header” comes into the bottom of the header and 5) cyclical condensate buildup in the coil because of poor design for the removal of condensate from the “bottom of a coil”. Possible corrective action: Immediate: the corrective changes require reconstruction. Add a mixer; relocate the steam line to the top of the steam main; relocate the condensate line into the top of the condensate header; install a condensate lift mechanism to allow steady withdrawal of the condensate from the bottom of the coil. See sketch in Figure E-6.

TO TRAP

FROM COIL FALL Figure E-6 Condensate lift mechanism showing closeup of fitting for Case ’44.

557

558

Appendix E Debrief for the Trouble-Shooting Cases

TS Process: illustrative hypotheses: no mixing/ wrong steam trap/ flooded steam trap/ condensate buildup and then cycling out of coils/ sensor fault/ sensor location wrong/ steam valve too big/ dirty steam/ corrosion products in steam/ steam trap bypass open/ cycling condensate from other units flowing back into this unit via condensate header/ fouled heat coil. TS Process: possible diagnostic actions: 2805, 2696, 2014, 1220, 1409, 1492, 1037, 1014, 727, 935, 2287, 2421, 612, 402, 2917, 2092, 2577, 2979, 1702, 1093, 1494.

(courtesy T. E. Marlin, Chemical Engineering Department, McMaster University) Estimate of difficulty: 8/10. Cause: The bottoms composition is not controlled in real time. The C-8 temperature, as measured in tray ’9 downcomer, is measured and controlled (by feedback) by adjusting the steam flow to the reboiler. This tray is not a perfect indicator of the bottoms composition. Thus, changes in the C-8 feed composition result in changes in the bottom composition. Possible corrective action: Immediate: operator adjust the set point on TC-5 based on the real-time sensor analysis A-1 on the overheads of the downstream debutanizer: if C3 is too low, decrease the TC-5 set point and vice versa. Longer term: AC-1 cascade to TC-5. The AC-1 primary will correct the deficiencies in the relationship between TC-5 and the bottoms composition. TS Process: illustrative hypotheses: sampling error/ analysis error/ variable changes in feed concentration/ fouled reboiler/ faulty reboiler design/ bottoms controller poorly tuned / controller faulty/ temperature sensor faulty/ electrical storm interfering with signals/ poor design of control system. TS Process: possible diagnostic actions: 2380, 803, 1563, 1156, 2078, 436, 2251, 1334, 219, 1303, 2989, 2769, 652, 201, 2102. Case ’45: The deprop bottoms and the ISO dilemma

Case ’46: The not so cool chiller (courtesy Scott Lynn, Chemical Engineering, University of California, Berkeley) Estimate of difficulty: 8/10. Cause: construction garbage left in thermosyphon line causing limited ethylene circulation such that the bundle in the chiller was only half covered with liquid. Possible corrective action: clean out the line. TS Process: illustrative hypotheses: temperature-sensor error/ level sensor error/ flowrate too low/ bypass open on pump/ leak from process fluid into the refrigerant loop/ boiling shifts from nucleate to film boiling along the length causing vapor instabilities/ ammonia condenser not condensing / ammonia leaking into the pro-

Appendix E Debrief for the Trouble-Shooting Cases

cess fluid/ design-problem area too small/ plug in vertical thermosyphon line/ valve partially closed in thermosyphon line. TS Process: possible diagnostic actions: 2598, 907, 863, 746, 1429, 1302, 1397, 1780, 1942, 2222, 2471, 2193, 2584, 2697, 2217, 2042, 1625, 1817, 714, 875, 1869, 1711, 1693, 2340, 1821, 2875, 1161.

Case ’47: The fluctuating production of acetic anhydride

Estimate of difficulty: 8/10. Cause: The control system was poorly designed because the sensor for the steam flowrate is a flowmeter on the vapor. This might work for well-behaved vapors; but acetic acid is not well behaved. Some acetic acid vapor dimerizes so that the molar mass of the vapor is somewhere between 60 (for the monomer) and 120 (for the dimer). For design purposes we usually used 100. However, the control is based on a signal from an orifice meter measuring the flow of the vapor. But the apparent mass flowrate (estimated from the signal from the orifice meter) is a function of the gas density and the measured Dp. Therefore, even if the mass flowrate of the vapor is constant, the apparent flowrate read by the orifice meter is erratic because of the variation in the molar mass of the vapor. Possible corrective action: Shift the controller to control the steam flow by the liquid flowrate and adjust the liquid flowrate via level control. Or better still, add a gas analyzer to determine the density of the gas. This datum is multiplied by the pressure drop given by the orifice meter, a square root function is taken of the resulting data and sent to the steam flow controller. This is illustrated in Figure E-7. PI 201

DR

TO CRACKING FURNACE

FI

A

201

201

r

FRC 202 PI 101

LC

STEAM MAIN

LIQUID ACETIC ACID TO CONDENSATE HEADER Figure E-7

Modified control system for the vaporizer for Case ’47.

559

560

Appendix E Debrief for the Trouble-Shooting Cases

TS Process: illustrative hypotheses: poorly tuned controller/ location of orifice meter too close to bends/ wrong orifice diameter/ condensate cycles because of poor trap or flooding of the exit of the trap/ cycling steam/ level cycles/ cycling in the downstream vacuum pumps/ uneven temperatures in the cracking furnace causing cycling reaction rate/ cycling brine and water to the condensers causing upstream cycling/ cycling in “barometric legs” for the condensed water from the system. TS Process: possible diagnostic actions: 2973, 2590, 2153, 1607, 1927, 1164, 2787, 1706, 2057, 2604, 2933, 560.

(courtesy T. E. Marlin, Chemical Engineering Department, McMaster University) Estimate of difficulty: 8/10. Cause: One or both of the isolation valves around the depropanizer reflux valve were not completely opened. The resistance to flow is too high. The FC-4 controller has opened the control valve 100% but the reflux flow is much below the desired value. Possible corrective action: Immediate: open the isolation valves. Longer term: check before the process is started up. TS Process: illustrative hypotheses: dry trays/ flooding/ insufficient reflux/ low feedrate/ high boilup/ feed temperature too high/ reflux pump cavitating/ sensor fault/ analysis fault/ poorly tuned controllers/ transient vapor puff from horizontal thermosyphon reboiler/ bypass open on reflux control valve. TS Process: possible diagnostic actions: 1820, 1712, 1628, 2309, 2760, 2974, 1573, 1515, 1374, 943, 974, 1766, 1913, 1695, 2277, 2127, 2925, 2593, 2366, 1655, 1747, 2769, 1217, 1015, 1246, 1647, 1604. Case ’48: The column that just wouldn’t work

(from D. R. Winter, Universal Gravo-plast, Toronto, 2004) Estimate of difficulty: 8/10 Cause: Wrong location of gate and mold temperature was too low. Possible corrective action: Use only one gate with a location at the hub; control the mold temperature at 65 C. TS Process: illustrative hypotheses: Contamination near the hub/ moist resin/ uneven temperature over the mold/ injection too slow/ injection too fast/ faulty design of the product/ faulty design of the mold/ too much cooling/ too little cooling/ faulty resin/ incorrect foaming agent/ correct foaming agent but wrong concentration/ inadequate mixing/ cooling cycle too long/ cooling cycle too short/ feed temperature too low/ mold too cold. TS Process: possible diagnostic actions: 2473, 1522, 2314, 2892, 2629, 384, 721, 904, 1076, 1419, 1939, 1980, 2447, 2007, 149, 623, 1308, 1741, 1965, 2267, 782, 965, 1159, 1126, 1754, 1597, 1783.

Case ’49: The case of the faulty stretcher pedal

Appendix E Debrief for the Trouble-Shooting Cases

(courtesy W. K. Taylor, B. Eng. 66, McMaster University) Estimate of difficulty: 9/10. Cause: Two causes: 1) the steam tracing was poorly designed and overheated the suction line to the pump and 2) the vortex breaker was poorly designed and covered most of the nozzle exit cross section leaving the column at tray ’4. Possible corrective action: Immediate: shut off the steam tracing. This helps, but the pumping is still below expectation. Long term: correct the vortex breaker. It almost plugged the exit from the column! Figure E-8 shows a photo of the vortex breaker.

Case ’50: The cleanup column

Figure E-8

Vortex breaker blocks the outlet in Case ’50.

TS Process: illustrative hypotheses: product not in the well of downcomer because of: downcomer blocked at top/ overhead leaks out of well and goes to bottoms/ tray collapsed and liquid bypasses the well/ no feed to the column/ feed bypassing to drum disposal/ insufficient steam stripping/ insufficient boilup/ reflux water off and organic goes out the top/ feed composition is primarily heavies. Product in the well of downcomer but cannot be pumped out because of: pump off/ pump not primed/ pressure relief causing liquid to recycle around the pump/ pump cavitating/ steam tracing too hot/ vacuum leak of air into system/ wrong pump selected for the job/ insufficient head/ line blocked/ product solidified in line/ junk left in line after maintenance/ physical blockage from vortex breaker/ no vortex breaker. TS Process: possible diagnostic actions: 753, 157, 2605, 2230, 1850, 1950, 1674, 847, 946, 695, 650, 2374, 2475, 2868, 797, 2030, 2413, 1612, 1854, 1525, 1336, 1039, 761, 524, 211, 393, 282, 2141, 2281, 2785, 2857, 1420.

561

562

Appendix E Debrief for the Trouble-Shooting Cases

(courtesy W. K. Taylor, B. Eng. 66, McMaster University) Estimate of difficulty: 9/10. Cause: weeping sump at Tray ’4; collapsed trays in the stripping section; weeping/ leaking trays because the sieve holes have corroded. Possible corrective action: Repair the sump; replace the trays. Long term: identify the cause of corrosion and remedy by removing the cause or selecting different materials of construction. Figure E-9 is a sketch of the sump.

Case ’51: Revisiting the cleanup column

TOWER CONNECTION 4th Tray

seal pan

25"

12½"

LEVEL GLASS

6" 13"

1" TOWER CONNECTION Product Sump

LEVEL GLASS CONNECTIONS

2'8" ID seal To 114-J

TOP VIEW not to scale

sump 4th TRAY NOZZLE

Figure E-9

Sketch of the sump in Case ’51.

TS Process: illustrative hypotheses: sampling error/ analytical error/ insufficient boilup/ stripping steam flowrate too low/ vacuum fault/ trays collapsed/ downcomers not sealed/ sieve holes corroded and trays weeping/ product sump leaking. TS Process: possible diagnostic actions: 2605, 2230, 1950, 1674, 847, 946, 2650, 1376, 2644, 2517, 2133, 2030, 2413, 2854, 2225, 1336, 761, 524, 2221, 393, 785, 404, 1120.

Appendix E Debrief for the Trouble-Shooting Cases

(courtesy W. K. Taylor, B. Eng. 66, McMaster University) Estimate of difficulty: 9/10. Cause: Trying to balance the flow to more than one downstream unit from a common header is always tricky. The pressure drop in both downstream branches must be exactly equal; otherwise the flow will go preferentially to the branch with the least resistance. A second, complicating element, is that the branches include a reactor that, if one reactor develops a hot spot, the reaction rate increases, and the flow to that reactor increases. The third complication is that the configuration of the feed gas into the manifold is not a mirror image (and fresh feed contains 1% inert whereas recycle feed contains 15% inert). Thus, if more total feed is called for by the South loop (which is usually fed by fresh feed and recycle) then the source of the “extra” feed for the South loop is “fresh feed” from the North compressor. In this way the “extra” feed to the South loop is richer in fresh reactants and has less inerts. Contrast this with what happens if more total feed is called for by the North loop. Then, the source of “extra” feed for the North loop is the “recycle feed” from the South compressor. This recycle feed contains 15% inerts so the net result is a feed to the North reactor that is slightly higher in inerts. When production in the South loop is 20% higher than the North loop, then the South loop gets all of North loop’s fresh feed (because of the 5:1 recycle to feed ratio). The North reactor gets no fresh feed, the North reactor quickly reaches equilibrium and stops reacting for lack of reactants. It is little wonder that it is usually the South loop that overheats, although occasionally the North loop does when the operators are trying to smooth out the loops. The situation is not dangerous because the temperature never exceeds 590 C (which is well within the catalyst and vessel limitations). Possible corrective action: Immediate: try to keep the South reactor from getting too hot by operating it 5 to 10 C cooler than the North one. Monitor it closely. Interim: add a high-temperature alarm. Longer term: install isolation valves so that each compressor system feeds one system. TS Process: illustrative hypotheses: fluctuating valve stem in control valve A/ poor manual control of valve B/ fluctuating kickback on one of the compressors/ fluctuating valve B on the exit of the recycle compressors/ during startup both reactors are not heated up to identical inlet temperatures/ depth of catalyst differs between the two reactors/ analysis of pressure drops in the two loops shows that Dps are not identical for the same flow/ recycle gas in one loop is cooled more than that in the other loop/ hydrogen content in one loop leaving the reactor is different so that the cooling is more in one loop than the other. TS Process: possible diagnostic actions: 1080, 1158, 2354, 1852, 1982, 1845, 1602, 920, 683, 556, 131, 369, 289, 1824, 1884, 2318, 2392, 2872, 2903, 2623, 1620, 1657, 1181, 1216, 1384, 1268, 1340, 633, 2756. Case ’52: The case of the swinging loops

563

565

Appendix F Other Tasks for the Skill-Development Activities in Chapter 5 F.1

Enrichment Tasks for Awareness Development in Section 5.1 Task 5.1-B

Tasks related to science and engineering

B.2

Bernoulli’s equation, expressed in the old units of ft lbf / lbm, is:

D2 Dp ft:lbf þ Dz ½¼ þ 2g lbm where v = velocity, ft/s p = pressure, lbf/ft2 = density, lbm/ft3 z = elevation or height, ft g = local gravitational acceleration, ft/s2 Consider the term Dz. a) b) c) d) e) f)

since the units of z are not the same as the units in the equation, it should not be included. the term should be included, the units of ft are acceptable because the correct units are understood. consistency of units does not apply to this equation; go ahead and use the term as it is. consistency of units does not apply; the equation was derived in 1738 when consistency was not an issue. Use it the way Bernoulli derived it. the term must be multiplied by g / gc. other.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

566

Appendix F Other Tasks for the Skill-Development Activities in Chapter 5

Task 5.1-C

General tasks more related to trouble shooting

C.2 I am to write a report to my supervisor telling her about what I have done this past month on my research project. Which of the following topics are most important to be included? 1) the pump for the equipment broke and only half of the experiments could be completed, 2) the mechanics could not fix the pump; the new parts will arrive in 5 weeks, 3) the results for the experiments that were completed seem to show that the most important effect is the temperature – not the concentration of the reactants. But, we cannot do a statistical analysis of the data until all of the experiments are completed. 4) I went to the library and studied the literature for another project because I could not do the experiments. 5) the technician will not be able to do the experiments in 6 weeks because he will be in Kyoto. Task 5.1-D

Engineering related

D.2 A researcher is studying the catalytic combination of two reactants A and B to form a single compound C. The reaction is first order with respect to the reactants. Ten minutes after the reaction has started, the researcher accidently adds chemical X that combines rapidly with A. Enough X is added to react with about 1/2 of A. Which of the following is most likely to occur when the clock reads 12 minutes after the reaction has started? 1) 2) 3) 4) 5)

the reaction of A and B to produce C would proceed more rapidly. the reaction of A and B to produce C would proceed at the same rate. the reaction would stop; that is, no more C would be produced. the reaction of A and B to produce C would proceed more slowly. other.

Task 5.2-C

Terry Sleuth and the Case of the Stinky Margarine

The spring rains turned the landscape into a muddy mire. What miserable weather, Pete thought. Why even the creeks by his house were overflowing the banks. He could imagine the quagmire down by the Burlington treatment plant where some of his colleagues were checking out the operation. He was glad he was indoors. Rinnnnng ... “Water Consultants” answered Pete. “Hmmm. Please slow down a bit so that I can understand what the problem is.”

F.1 Enrichment Tasks for Awareness Development in Section 5.1

Pete sipped on his coffee ... He saw Terry bustle in from the computer and waved Terry over. Motioning, he asked Terry to lift up the extension and hear the latest challenge that was besetting the engineers and scientists at Water Consultants. “So you make margarine ... ahum.. and now the margarine tastes like ammonia. But why do you think that we might help? We specialize in water and water quality! We specialize in water treatment, adsorption, ultrafiltration, and trace organic analysis but not margarine production” ... “Oh, you’ve tried everyone else and no one seems to have any answers so you thought that maybe we can help.” said Pete as he cast a puzzled look at Terry and then suggested “Maybe my colleague has some questions or suggestions for you; I’ve asked Terry Sleuth to listen in on our conversation” as Pete dodged the issue and passed the situation neatly over to Terry. “Perhaps you will refresh my memory on how you make margarine; you purchase vegetable oils, clean them up and then blend them with milk.” offered Terry Sleuth, “What do your lab analyses of the incoming oil and milk show?” “OK, so they seem to be the same as usual. How do you clean up the oils?” Terry jotted down “ deodorize under a vacuum and with live steam, adsorb undesirables on Fuller’s earth, filter out the Fuller’s earth and cool down the oil”. Pete’s ears perked up when he heard the word adsorb with Fuller’s earth because that was his specialty. Maybe they could help. Terry seemed lost in thought as Terry looked out the window at the quagmire and asked, seemingly to give more time for thought, “Do you produce your own steam from the plant with the intake from Butcher’s creek?”. “OK you do;” “And you pump the cooling water from the same source? Hmmm.” “Do you filter the cooling water before you send it to the margarine plant? It’s rather dirty these days with the terrible weather outside”. “OK so you filter it and send it to the cooling-tower circuit.” “All your coolers have no direct contact with either the milk or the oil, Right?” “Right.” Pete asked “Do you regenerate the Fuller’s earth for reuse?” Hmmm. Terry said “I think we can solve your problem; we’ll be over in an hour to take a few samples. We should have an answer for you in about 3 hours”. What samples did Terry take? what analyses were done and what advice did Water Consultants give to solve the case of the stinky margarine? Task 5.2-D

Terry Sleuth and the Case of Boorish Bob

Bob jogged by Terry’s office. The odor of locker-room socks and just plain BO wafted through the door. Terry opened the window. As Terry turned back to the desk, the perspiring face of Bob appeared in the open doorway. “Joggin’ really makes you feel good,” announced Bob proudly. “And talkin’ of feeling good, I have a meeting in half an hour with our client Marlene from Transpix. She asked us to calculate the weight per cent of zinc chloride that they should be using for the additive solution they are using. I’ve got those calculations all done. They were really simple.” With that, he threw his calculations and results down on Terry’s desk.

567

568

Appendix F Other Tasks for the Skill-Development Activities in Chapter 5

“Look em’ over. You’ll see they are great! Many of the others here in this organization don’t take the care I do to check and double check. Just as one last check, you check me out! After all, Marlene is an important client and we don’t want to make her upset, do we?” gloated Bob. Terry breathed deeply the fresh air from the window and then looked at the calculations. “What was Marlene’s question?” asked Terry. “She has a solution at 68 F (20 C) that has a density of 1.4890 g/mL that contains 0.0411 lb-mole of zinc chloride per US gallon of solution. She wants that expressed as weight%.” smugly beamed Bob. His bad breath almost made the paper curl. “I looked up the density of water at 20 C, (0.9982 g/mL) and the density of the solution to be 1.4890 Mg/m3” boasted Bob. “A US gallon is 8.337 lbm but the real thing to remember is that a gallon is a volumetric unit of measure.” Tired of Bob’s tirade, Terry looked at Bob’s calculations: 0:0411 · 136:28 ¼ 45:2 % 1:4890 8:337 · 0:998 Terry looked at Bob, frowned, and said ... What did Terry say?

569

Appendix G Selected Responses to the Activities in Chapters 6 and 7 .

From Chapter 6

6.1 a. stop the operation, open the pipe at the various fittings and look: very expensive. b. increase the flowrate and note the difference in pressure drop: ineffective activity unless one hopes to dislodge the obstruction. c. decrease the flowrate and note the difference in pressure drop: ineffective. d. estimate the pressure drop and compare with measured values: recommended early test. e. stop the operation, open one end and send a plumber’s worm through the line: expensive. f. and maybe there might be some others. 6.2 1) check the pressure difference between the refrigerant side and the process side. Easy to do. If the refrigerant pressure is greater than the process side, then the leak will be into the process and not into the refrigerant. 2) sample the refrigerant and analyze for impurities using LGC. Expensive, time consuming. 3) read the temperatures and pressures and compare these with the data on pressure–enthalpy charts for the refrigerant. Easy, pressure should be higher and temperature lower. 4) read the pressure gauge on the compressor suction. Easy, pressure higher. Select 1, 4, 3, in that order. 6.4 Typically “only card U” is chosen whereas the correct test is “cards U and 9”. If card 9 is not chosen, you might tend to have a confirmation bias in that you look only for a test that proves the case and not an additional test to check the negative, namely that card 9 does not have a vowel. 6.5 The Tom Dayton murder The answer is ’3, conclude that Tom died from self-inflicted shot based on m, p, q, hh, r, s, v, w, kk bb. If you chose ’1, perhaps you might have a bias in either selecting the evidence or reaching conclusions. Look over the characteristics given in Sections c-i and c-ii. Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

570

Appendix G Selected Responses to the Activities in Chapters 6 and 7

6.15Adiabaticity: Two simultaneously contradictory beliefs: heat is what is “in” hot things and heat is energy flowing from hot to cold (from HELP P, “bugs” session 20). .

From Chapter 7.

Activity 7.1a

Audience: Engineer to operator. The operator should be told the purpose of the test, how and why the results will be used. Is the operator permitted by the union to obtain samples, or is it the lab technician or instrument shop? Content: The instructions should have been clearly stated. Location of sample port; procedure for gathering the sample safely, how to clear the sample line and get a representative sample. What volume of sample to take? Can the samples be stored? What sample bottles to use? How to label them? Time period between samples? What analyses are to be done? When you would like the results? Organization: difficult to assess because the instructions are so short. Style: rather condescending attitude communicated. Format: should have been in writing. The Engineer should have phoned the lab, checked on the types of analyses that could be done and the time frame, the volume of samples required and obtained sample bottles from the lab. Then Jose should have verbally explained what samples, been present when the samples were taken and labeled the samples. The written instructions should have been part of the approach taken. Activity 7.1b

The Pulp Mill engineer verbally told the instrument shop to “check the flowmeters.” The results reported to Andre verbally were “Some of the instruments were away off; they forget which ones but that’s OK because they are all OK now.” This meant that all the test data taken when the instruments were “way off” we useless because a) we don’t know which values were correct and which were wrong and b) for those that were wrong, we don’t know what the correction factor should be. Thanks to Ian Shaw, B Eng. PhD McMaster for this example. Activity 7.2

Ahmed shows no listening in this situation. It is hard to identify five strengths. They might be: 1. He acknowledges the operator. 2. He goes to the control room first instead of going directly out on the plant to see the VC. 3. He calls the operator by name. 4. He acknowledges that he is new and 5. He talks aloud about his observations. Areas to work on: many! but start simply with the two most important ones: 1. Take time to introduce himself. 2 Learn to listen and seek input from the operators.

Appendix G Selected Responses to the Activities in Chapters 6 and 7

Activity 7-4

Tanya:

Marcos:

RIGHTS claimed right to have an opinion, Honored: initially tried to honor Marcos’s right to express an opinion (iii) but failed to honor Marcos’s right to have an opinion; (vi) (vii) lack of respect Destroyers: criticism (vi) (vii) RIGHTS claimed right to have an opinion. Honored: lack of respect Destroyers: defensive (ix) (x); criticism/sarcasm (xi) (xii)

Let’s try this again: Tonya: “OK, my six hypotheses are (i) 1. tray collapsed stripping section, 2. too much bottoms fed to debutanizer, 3. too much overheads in feed, 4. feed valve FV1 stuck, 5. pump F-26 not working, 6. not enough feed to the column (ii). What do you think? (iii) Marcos: “Your six hypotheses are a good start. Actually, I really support your first hypothesis I encountered something like that on the S256 plant last year. Same evidence and it was a collapsed tray. Tanya: “Maybe. But checking out a tray is more complicated than some simple tests that might rule out some of the other hypotheses. Why don’t we check out the temperatures, pressures and flows that we see on the plant. That shouldn’t take long.”

571

573

Appendix H Data about “Causes” for Selected Process Equipment Here are some example “causes” or faults that might occur for centrifugal pumps, valves, controllers and instruments, pipes and fittings, steam traps, shell and tube heat exchangers, plate heat exchangers, distillation and tray absorbers and sedimentation centrifuges. .

Centrifugal pumps

1. instrument wrong. Design. 2. discharge pressure required is greater than that provided by pump. 3. total system head required is much smaller than head provided by pump. 4. suction pressure is too low/too high. 5. available NPSH is too low or suction lift is too high. 6. undersized suction pipe. 7. excessive vapor entrainment. 8. incorrect sump design. 9. viscosity higher than expected. 10. density different from expected. 11. incorrect piping layout. 12. sump level below intake. 13. loose valve or disk in system. 14. resonance between rotating speed and foundation. 15. incorrectly designed baseplate. 16. system requirements too far out on the head-capacity curve. 17. operates at very low capacity. 18. one pump in the system affects another. Installation: 19. misalignment. 20. wrong. damaged or improperly installed bearings. 21. wrong. improperly installed packing. 22. wrong type or amount of lubricant. 23. mechanical seals exert excessive pressure. 24. glands too tight. 25. rotor not balanced. 26. shaft bent. Operation: 27. air entering or not dearated initially. 28. impeller backwards in double-suction pump. 29. impeller clogged with solids. 30. impeller damaged, vanes worn or missing. 31. impeller diameter different from expected. 32. no key between shaft and impeller. 33. speed too high or too low. 34. running at critical speed. 35. rotation in wrong direction; leads reversed on power to drive. 36. faulty casing, cracks, leaks. 37. insufficient clearance between impeller and volute tongue. 38. wrong shape for volute tongue. 39. leakage through the wear surfaces (wear rings) process fluid out/ lube oil into process/ air leak in. 40. internal waterways rough and burred. 41. plugged inlet or exit. 42. suction strainer plugged. 43. motor faulty, undersized. 44. drive connection fails, slips. 45. casing and wear rings worn. 46. defective casing gasket. 47. no cooling water to water-cooled stuffing boxes. 48. dirt and grit in sealing liquid. 49. motor windings fail. 50. motor bearings fail. 51. change in or wrong frequency of voltage to the motor. 52. wrong phases hooked Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

574

Appendix H Data about “Causes” for Selected Process Equipment

up to the motor. 53. short or fuse in power to motor. 54. motor switch not turned on. 55. power failure on site. 56. power failure at the utility. 57. faulty signal to relay to start motor. 58. faulty relay for signal from process to the pump motor. 59. not primed. .

Valves, controllers and instruments

Measurement instruments: 1. wrong sample location. 2. sample withdrawn is not representative. 3. sample mixed up with other samples; incorrectly identified. 4. error/ contamination in the sampling. 5. insufficient or no purging done. 6. no sample withdrawn because line plugged. 7. faulty measurement made on sample; instrument wrong. 8. instrument installed incorrectly. 9. instrument corrections not made to the answer. 10. calibration/ standard conversion wrong. 11. result reported incorrectly or not reported at all; or confused with other sample or signals reversed. 12. periodicity to behavior and sampling done at the wrong time. 13. instruments damaged during startup trials or between startup and current situation. 14. flashing occurs in instrument. Valve: Design: 15. valve too small. 16. valve too large. 17. valve trim wrong material. 18. valve wrong type. 19. air to open when need air to close. 20. valve positioner needed but not installed. Installation: 21. installed backwards for non-vacuum conditions. Mechanical: 22. stuffing box too tight. 23. stuffing box too loose. 24. wrong materials in stuffing box. 25. valve stem bent. Operation: 26. valve bypassed. 27. block valves closed. 28. valve seat eroded away. 29. valve stem eroded and vibrating from flow. 30. valve stuck open, closed or midway. 31. receives the reverse signal. 32. accumulation of dirt in the valve. 33. valve stem falls off. 34. faulty valve positioner. Controller: 35. wrong set point. 36. no integral action and hence no return after offset. 37. loss of instrument air or electric power. 38. signal reversed. 39. controller stuck. 40. sends out low signal; ie fails low. 41. sends out high signals; i.e. fails high. 42. set on manual. Recorder/indicator: 43. displays wrong values. 44. pens mechanically stuck and cannot cross. 45. chart motor not functioning. 46. chart pens out of ink. 47. chart started and date one-day off. .

Pipes and fittings

Design: 1. wrong diameter. 2. wrong wall thickness. 3. insufficient support. 4. no allowances for thermal expansion. 5. not insulated. 6. faulty location of lines: steam off the bottom of the header; condensate into the bottom of a header. 7. insufficient drain lines. 8. low points and undrainable sections in lines. 9. valves and instruments at wrong elevations. 10. instruments installed at wrong locations. 11. no allowance for purge. 12. item on diagram but not installed. 13. inadequate statement of specifications on the contract. Installation: 14. left garbage in the pipe. 15. installed filter, strainer, screen backwards. 16. not pressure or vacuum tight. 17. installed wrong pipe. 18. made on-site changes without recording them. 19. substitution of inferior materials without/with notification. Operation: 20. line fails/cracks. 21. corrosion products build up or con-

Appendix H Data about “Causes” for Selected Process Equipment

taminates. 22. purge operated for too short a time or not at all. 23. line sags. 24. line receives abnormal/unexpected stress and ruptures. 25. thermal expansion greater than expected; pipe tears loose from the support. 26. excessive steam tracing. 27. insufficient steam tracing. .

Steam traps

Design: 1. wrong type of trap installed. 2. wrong size of trap. 3. wrong orifice size for the Dp across the trap. 4. traps in the system interact and interfere with the operation of each other. 5. no strainer included. 6. condensate cannot get to the trap. Installation: 7. installed backwards. Operation: 8. filter plugged. 9. trap inoperative because it is frozen. 10. bypass left open. 11. mechanically stuck internals. 12. mechanically loose and unattached internals. 13. disk installed upside down. 14. exit flooded. 15. exit plugged. 16. trap internally plugged with dirt. .

Shell and tube heat exchangers, condenser, reboilers

1. Instrument fault. Design: 2. overdesign by adding more area for future. 3. overdesign by use of wrong fouling allowance. 4. overdesign “to be on the safe side”. 5. overdesign by adding large fouling factors to extend runs between cleaning. 6. underdesign. 7. inadequate allowance for thermal expansion. 8. no inlet protection baffle. 9. clearance between the shell and cross-flow baffles too large. 10. no air vents. 11. no safety relief inside blocked pipes.12. baffle spacing and baffle window area inconsistent to give turbulent flow in both. 13. liquid velocity < 1 m/s. 14. baffle loose. 15. didn’t account for hydrogen concentration in gas. 16. decrease in pH causing corrosion. 17. increase in pH causing fouling. 18. DT high; designed for nucleate but film boiling occurs. 19. water temperature exceeds 50 C. 20. temperature cross-over. 21. increase in clearance between baffles and shell. 22. no sealing strips between baffle and shell. 23. maldistribution. Installation: 24. improper fitting of gaskets in headers. 25. headers damaged during installation. 26. overpressure during trial tests. 27. garbage left in lines, tubes, shell. 28. lines installed backwards with wrong fluid on the shell side. 29. installed with cocurrent flow when design was countercurrent. 30. condenser installed vertically when design said horizontal. Operation: 31. tubes fouled inside. 32. outside tubes fouled. 33. air in tubes. 34. air in steam. 35. air not bled. 36. superheated steam. 37. change in hydrogen concentration in gas. 38. shift in DT in boiler causing film boiling. 39. wet and dirty steam. 40. steam trap fault. 41. insulation damaged or wet. 42. decrease in flowrates. 43. increase in flowrates. .

Plate exchangers

Instruments: 1. faulty readings. Design: 2. under vacuum. 3. control valves on exit lines. 4. temperature > 120 C; 5. pressure > 2.5 MPa. Operation: 6. temperature too high. 7. temperature spike. 8. cold fluid stopped but hot fluid continues. 9. superheated steam. 10. under vacuum.

575

576

Appendix H Data about “Causes” for Selected Process Equipment .

Distillation and absorption columns

Instruments: 1. faulty readings. 2. faulty samples. Control: 3. improper choice (two columns in series on level control; controlling both top and bottom on temperature). 4. pressure throttling the wrong side of the instrument. 5. faulty placement of sensor. Design: 6. trays not held down properly. 7. inadequate support-bearing area for trays. 8. trays not level. 9. insufficient provision for flashing feed. 10. pump problems (see Pumps). 11. inadequate venting of condenser/ reboiler. 12. gas pockets in liquid lines. 13. insufficient pressure equalization to pump out/ drain system. 14. reflux subcooled. 15. feed and reflux liquid incorrectly introduced to column (at wrong location or not into a downcomer). 16. access holes aligned with downcomers. 17. didn’t account for polymerization or crystallization within column. 18. series of columns with conditions shifting from oxidation to reduction. 19. wrong materials of construction (because, for example, prelim bench tests didn’t include trace contaminants). 20. incorrect allowance for parasitic reflux in uninsulated columns. 21. didn’t seal downcomers during startup. 22. didn’t allow for drainage from the seal well. 23. inadequate allowance for thermal expansion. 24. trays not designed for liquid loading, only vapor loading. 25. didn’t account for surface phenomena contributing to foaming, wetting and Marangoni instabilities. 26. no splash baffle. 27. didn’t use split flow when needed. 28. vortex breakers missing. Installation: 29. Poor-quality construction (edges rolled when too cold and cracks concealed by filling with lead, which then contaminated the system). 30. sieve holes bigger than design. 31. trays not installed correctly (not level, bent, poor seals for downcomers, not held down). 32. junk left behind in column. 33. bubble caps loose. Operation: 34. high superheat in live steam. 35. high superheat in steam to reboiler. 36. pump problems. 37. reboiler problems. 38. condenser problems. 39. plugged vapor lines, downcomers and/or lines. 40. cannot get downcomers to seal. 41. cannot get trays to drain at shutdown. 42. variable feedrate or composition. 43. fouling of trays. 44. polymerization products inside column. 45. wrong feed tray. 46. insufficient reflux. 47. cycling. 48. boilup rate below tray stability limit. 49. bumping of trays because of flashing liquid after upset. 50. flooding (plate collapse, excessive vapor rate, excessive liquid rate, low surface tension, foam stabilizing agents, particles, pH). 51. changes in utilities. 52. shift in equilibrium. 53. components in feed. 54. vapor locks in fluid lines (caused, for example, by steam tracing). 55. excessive venting of process materials. 56. excessive liquid entrainment from the top tray. 57. maldistribution on trays. 58. vortex breakers not working. 59. syphoning in down pipes. .

Sedimentation centrifuge

Installation: 1. flexible piping not used. 2. misalignment. 3. conveyor bowl not balanced. 4. not level. 5. liquid dams not set alike or set incorrectly. Operation: 6. blown fuse. 7. overload relays tripped. 8. motor overheated. 9. broken shear pin. 10. lube oil flowswitch tripped. 11. broken isolators. 12. motor on flexible mounts. 13. motor bolts loose. 14. bearing failure. 15. damaged plows.

Appendix H Data about “Causes” for Selected Process Equipment

16. damaged conveyor hub.17. solid product buildup in conveyor hub. 18. conveyor flights worn or portion of blade missing. 19. trunions cracked or broken. 20. conveyor bowl cracked or broken. 21. leaking effluent weirs. 22. plugged solids in the effluent hopper. 23. feed temperature too low. 24. feed rate too high. 25. effluent hopper not vented. 26. solids concentration too high. 27. foreign material stuck in bowl. 28. worn bowl strips. 29. loose or broken trunion bolts. 30. bowl inadequately washed. 31. clearance too large for blade tip to bowl wall. 32. bowl inside rough. 33. wrong size shear pin. 34. no strip installed. 35. interaction with other batchwise process in the system.

577

579

Appendix I Feedback about Symptoms for Selected Causes Relating cause ﬁ symptom. Some example causes are listed, for selected pieces of equipment, in Appendix H. In this section, we list a range of symptoms (readings, sounds, visuals, sample results, behaviors) and then, for selected causes, we give the symptoms. .

Centrifugal pump

Symptoms: A: no liquid delivered. B: insufficient capacity; flow lower than expected. C: intermittent operation; loses prime shortly after start. D: insufficient discharge pressure; pressure less than expected. E: short bearing life. F: short mechanical-seal life. G: vibration and noise. H. power demand excessive; higher than expected. I: pump overheats and or seizes. J: stuffing box leaks excessively, K: “crackling” noise. From Appendix H, for centrifugal pumps, the cause is listed in bold face followed by a letter code for the likely symptoms. Cause ’5 (available NPSH is too low or suction lift is too high): A, B, C, G (I) K; cause ’6: A, B, C, G, (I) K. 7: A, B, C, G, (I) K. 8: A, B, C, G, ?K. 9. B, D, H. 10: H. 12: A. 14: destruction and rupture. 15: G. 16: H, I. 17: G, I. 18: A, B, D, I. 19: E, H, F, G, I. 20: E, G, I. 21: H, F, J, cross-contamination. 22: bearings. 23: H, F. 24: H, F. 25: E, G, F, I, J. 26: E, G, F, H, J. 27: (A) B, C, D (G) K. 28: B, H. 29: B, D, G. 33: too fast H; too slow (A) B, D. 34: G. 35: (A) B, D, H. 44: A. 45: B, D, H. 46: B, D. 47: F, I. 48: F, I. 59: A, I. The above form of information is extremely useful for the creation of the hypothesis, evidence, action table used in the TS Worksheet. The information can be presented in another format; namely a symptom ‹ causes list. Thus, for centrifugal pumps, Symptom A could be caused by 5 6, 7, 8, 12 16, (27), (33), (35), 44 59. This list can be prioritized by listing the causes in the most likely sequence. This is the form of the information that is summarized in Chapter 3.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

581

Appendix J Guide for Students: How You Can Get the Most from this Book Skill in trouble shooting is needed throughout your career. Trouble shooting is the application of your problem-solving skills in a particular context where: a) safety and economics are the major drivers, b) knowledge about how equipment and systems work is crucial and c) you work in a context of someone else’s turf, namely the process operator. I hope you have a chance to develop confidence in your skills as an undergraduate student. In this section are given some guidelines, especially for students, on how to use the material in this book to develop your skill. Section J-1 suggests that you start with an overview, before you start your journey. Then, since defining the problem is often a challenging task, start by working a problem where the TS worksheet gives a reasonable definition of the problem, Section J-2. Section J-2 elaborates on the activities and questions and clarifies the style I used in posing the activities. This information is important background. Details (beyond those given in Chapter 8) that are of particular interest to you are given here. Next, I would become familiar with a variety of problems and approaches by working through, and reflecting on, the trouble-shooting processes used by Michelle, Pierre, Dave, Saadia and Frank in Chapter 4, as recommended in Section J-3. Sections J-4 and J-5 describe activities of developing your prerequisite skills and then how to use the Cases of Chapter 8. Finally, Section J-6 describes the context in which you will be applying your trouble-shooting skills in the process industry.

J-1

Getting Started: Get the Big Picture

The Preface gives you an overview of the style and goals of this book. Proceed to Chapter 1 and the first, recommended activity that is the pretest in Chapter 1, Section 1.3. This helps you identify your starting skills in the six areas. You can use this to select activities to bolster your skills in areas needing attention. You can also retest yourself after you have worked about ten cases, and then after twenty and so on.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

582

Appendix J Guide for Students: How You Can Get the Most from this Book

J-2

Try a Trouble-Shooting Case where the Problem is Reasonably Well Defined

Next, read the Trouble-Shooting approaches suggested in Chapter 2 and scan through the worksheet for Case ’8, given in Chapter 2. This is a relatively challenging problem, rated 6 out of 10. Once you have read over the worksheet, I would go to Chapter 8 and try to solve the case by selecting the actions for Case ’8. For each case the actions are organized approximately in the sequence you might select the action, although you can jump around as you see fit. Let’s take some time to briefly look at the possible actions listed with each case. For all actions, the two most important criteria are safety and economics. The safety is considered first for each case. To illustrate the economic impact of each activity, for many of the cases I have given a “cost” for the activity. Unless otherwise specified, I have used $600/h as a cost for loss in production. To this needs to be added the cost of the people and equipment needed. For example, to obtain an answer to “what’s the weather today and in the past?” costs $50, representing about a 10-minute activity. One could argue that it only takes less than a minute to look outside and get the weather, but the I use 10 minute to represent actually locating a measurement of the temperature and some reflection about the weather over the week. This might require a phone call, information from the newspaper or the web. Another example might be the sampling and analysis. The elapsed time I estimate to be about 6 hours. Arrangements have to be made to have someone (trained in gathering samples) coming to the site, locating the sampling line, flushing the line, taking the sample and returning the sample to the lab. The lab has its own routine for completing the regular analyses. This “new” sample is non-routine and will have to be fit in when possible. In addition, the laboratory will charge the unit for the “additional” analysis. Hence 6 h $600/h plus $100 per analysis = $3700. The more complex the analysis, the higher the analytical charges. Some analyses will have to be done off-site because the lab does not have the facilities. For the more complex tests and the use of off-site labs, the cost per sample could escalate to one day delay and a sample cost of $300 for a total of about $15 000/sample. For each case I tried to assign reasonable costs. Now consider the actions themselves. The actions are in four sections. The first actions relate to safety! The second set of actions help you understand the background. The third set are factual gathering of information about the process as it operates now. By this time you have information to complete the TS Worksheet. This will then guide the fourth set that are various tests and actions to test hypotheses, identify the possible fault, correct and prevent the fault from reoccurring. Throughout all actions safety and economic are the driving forces. Here are more details of the four sets of actions that you may select.

J-2 Try a Trouble-Shooting Case where the Problem is Reasonably Well Defined

The first set of actions

The first priority is safety: recognize hazards (via MSDS) and take action. .

MSDS

Usually all engineers working on any site should know this information already. That’s one of the first things you learn for the processes with which you are working. If you don’t know the hazardous properties of the materials on the plant, then check Google, the MSDS sheets for the plant or safety manuals. Here, I usually give MSDS material that I have downloaded from manufacturer’s sites. Sometimes the key information is the NFPA ratings. The National Fire Protection Association, NFPA, has assigned ratings from 0 to 4 for health, fire and spontaneous reaction/ explosion for individual chemicals. Thus an NFPA rating of 0, 4, 0 would mean that the species is not an issue for health, it is extremely flammable and it is stable. For example, hydrogen cyanide is 4, 4, 0; water is 0, 0, 0. Nitroglycerine is 2, 1, 4. The NFPA ratings apply to individual species. I tried to include information about how one chemical might react with another when this is important. .

Immediate action for safety and hazard elimination. The first decision relates to safety. The initial evidence might suggest a health, a fire or an explosive hazard. Act now! There are four options: – Continue with the trouble shooting without implementing actions related to safety: there is no hazard. – Put on safe-park: this keeps the process going but under conditions that are safe. This could mean isolating a distillation column and keeping it on total reflux; or reducing the throughput to conditions that previously did not pose a hazard. – Safety interlock shut down, SIS: this should happen automatically if the control system has been designed correctly with the four levels of response expected. However, sometimes this has to be actively initiated. This gives you a chance to reflect on the situation and decide if this should have happened. – SIS plus evacuation: the SIS should happen automatically if the control system has been designed with the four levels expected. Now, because of the hazard posed we should add evacuation.

The second set of actions: background

These are mental or office actions to try to put the problem into perspective and start the process of defining the trouble. .

More about the process. For some problems, fairly detailed knowledge about the process (beyond that given on the diagram or in the problem description) might help you feel more comfortable working on the case. I assume that you know a reasonable amount about the process described in the case. If more knowledge might be helpful, then I have included this activity to help bring everyone to a comfortable level of basic understanding of the process. For some cases, this question is not included. For those where it is included,

583

584

Appendix J Guide for Students: How You Can Get the Most from this Book

.

.

you probably don’t need this additional information; but if you need some reassurance about what the process is about, this additional information might help. I have not included key information that is needed to solve the case in this activity. IS and IS NOT: (based on given problem statement). This is an excellent way to systematically consider the starting information that is given in the problem statement. You can complete this on your own, without incurring a cost. These are included here a) to remind you to do this; b) give you feedback about how you did the task. – What? – When? – Who? – Where? – Sometimes the Who? is not that helpful; it depends on the case. Why? Why? Why?

The Why? Why? Why? may be helpful to put some of the cases in the larger context. You can do this activity on your own and use this question to give you feedback. .

Weather Today and past. All cases refer to weather conditions in Ontario where there are four seasons with snow and freezing weather December through March; hot humid summers June to August. Why is the weather important to know? Cold weather gives the potential for freezing of steam traps (float type); the need for steam tracing of lines to be turned on; colder water from the cooling tower, river or wells. Storms mean the local atmospheric pressure is low – and implies that pumps from sumps (open to the atmosphere) might experience NPSH problems. Damp weather can mean that particles in hoppers might clump together. Lightning might be giving electrical interference with the instrumentation.

The weather may have a dramatic effect on the quality of the water. As Doug Pearson says, “A major storm (local or distant) could, through excessive run-off, affect the quality of water: excess turbidity, debris, sediments and changes in temperature of pH.” In spring the water usually is turbid and may have excessive amounts of humic acid. If this water, containing high amounts of humic acid, is used to generate steam, the steam may contain ammonia. Equipment that works in January may not work in August simply because the atmospheric temperatures are so different. .

.

Maintenance: turnaround. Three conditions might apply; new plant startup, startup of an existing plant that has just been through its annual turnaround and operation after some maintenance has been done. That the problem is with a first time or startup of a new process is usually given in the statement of the problem. Therefore, no separate question is posed related to this. Startup of a new process triggers faults in design, construction waste left in the plant, pretest air and water left in the lines, faulty operating

J-2 Try a Trouble-Shooting Case where the Problem is Reasonably Well Defined

.

instructions and operators inexperienced with the subtleties of this particular process. The Maintenance: turnaround activity relates to startups after the annual turnaround. During the turnaround the minimum is usually – inspection of most pieces of equipment (such as sending someone down through the central part of a distillation column to check on the level and condition of the trays), – replacing worn parts, – the installation of changes to the process, repiping, and changing the operation to implement ideas to optimize or improve operation, – checking the calibration of sensors, – cleaning exchangers, – changing the catalyst. It is important to know “When and what done?”

.

Maintenance: routine. During routine maintenance something will have changed. It is important to know what and the extent. It could be that all the isolation valves were not correctly opened after the maintenance was completed; that the key was not correctly fitted into the drive shaft; that the pump was not primed. When and what done? The answers to this question will not admit to such mistakes. However, this will open your mind to possible things to check and look at.

.

What should be happening

The experienced engineer will have already internalized knowledge about what should be happening on the process before he/she encounters a case. However, as a student it is wise to gather this background information before jumping headlong into the case. The general sources include the simulation/design computer background; records of the design; the collection of information from the equipment vendors and any internal reports on past tests, or trials done on the plant. Some values of key properties might be useful to know. These would be available from a Handbook, from Google or from internal files. You might peruse your trouble-shooting files that give symptom ‹ cause information. Based on the information you have already, you often can do some simple checks and calculations. Here are the details of each of these actions. Design and simulation files (allowances made for fouling, overdesign and uncertainties). Here I tried to list all the possible pieces of equipment involved in the case. All equipment are designed according to the Codes and Standards; the information given here are decisions within a designer’s judgement. Vendor files. This is the practical information and some specifications from suppliers of heat exchangers, steam ejectors, pumps and equipment that would be pur-

585

586

Appendix J Guide for Students: How You Can Get the Most from this Book

chased from a vendor. The information may be slightly different from the information in the design files. Commissioning data, P&ID, internal reports. Often new plants are constructed under contract. The contract may include penalty clauses stating a financial penalty that is paid if the process does not perform up to specification within a stated period of time. For insurance and governmental regulations, certain performance standards have to be met. Usually records are kept of the performance trials. For the startup of a new plant, such records are being developed. For processes that have been operating for a while, the startup and internal reports and the P&ID diagram may still be pertinent. They may or may not be available. In some cases, the operation has changed so much that the information is not helpful. Handbook. The data given here are usually physical and thermal properties of the chemicals in the case: vapor pressure–temperatures, steam tables, thermal properties. Such information may be needed in calculations and estimates you do. Trouble-shooting files. Here I refer to specific sections of Chapter 3 that pertain to the equipment in the case. .

Calculations and estimations (that can be done in the office before special tests are done) The calculations reported here are simple and not ones that depend on numbers other than those given in the case statement or that you recall from rules of thumb. (Later you might do more improved calculations from the data you gather.) The given information varies from Case to Case. Nevertheless, some simple checks and calculations can provide neat insight. These include: – Pressure profile: in most cases the pressures on either side of a barrier are given and you can state the direction of flow that would incur if there is a leak in the barrier. For example, 1 MPa steam boiling hydrocarbon in a column at pressure of 0.6 MPa; then the steam would leak into the hydrocarbon. – Mass balance: this is an important fundamental to help us understand what is going on. However, usually at the start of a Case we do not have enough information to do the calculations. We will be able to do a mass balance later after we have gathered data. – Energy balance: sink= source; the heat lost by one fluid = heat gained by the other. Often we have enough information to allow us to do this check. – Thermodynamics: we can often use the principles of thermodynamics to predict trends, in general. – Rate: we can use DTs to estimate whether nucleate or film boiling might predominate. Similarly, we might be able to estimate rates of mass and heat transfer. – Equipment performance: we might be able to estimate the performance of equipment; usually, however, we need more information than that given in the Case statement.

J-2 Try a Trouble-Shooting Case where the Problem is Reasonably Well Defined

The third set are factual gathering of information about the process as it operates now. .

What is the current operation

Whenever you go out to the plant, you must realize that you are going onto the operator’s territory. You need to check in with the operator first. Unless you arranged over the phone to meet the operator at the piece of equipment, you will go to the control room first and check in. You might scan the data reported in the control room, discuss the operating actions with the operator and explore the operating procedures. Visit control room: control-room data: values now and from past records. Here we can obtain values that the instruments read and look at past records. Process operators; You can ask for information about what happened (from their perspective) and they might offer ideas as to what is the fault. The operators are usually a very valuable source of information. .

Operating procedures Knowledge of the usual operating procedure to be used for this condition might help.

By this time you have information to complete the TS Worksheet. If you are working on Case 3 4 5 or 8 where the TS Worksheet is published in the book, then you might start selecting actions from this point onwards. .

Check with colleagues about hypotheses

In practice, you may not have a chance to check with someone else about your hypotheses. However, in this book and series of actions, you might elect to get some feedback about your list of hypotheses that you wrote on your TS worksheet. As you develop your confidence you will no longer want to get the feedback. Indeed, I have elected not to give this feedback for cases after Case ’29. You may wish to create a symptom ‹ cause diagram similar to the one illustrated in Figure 6-2, in section 6.5.5c. The fourth set are various tests and actions selected to test hypotheses/or correct By this time, you have identified more than seven hypotheses as to the fault and probably created your symptom ‹ cause diagram. Apply a strategy of simple tests. Inexperienced trouble shooters tend to: 1.

2.

Become fixed on “it’s got to be this” and want to take a corrective action immediately. Although these tend to be dominant J stylists (described in Section 6.1.3.3, section a) inexperienced trouble shooters also often demonstrate this preference. For these people, I have included Take “corrective” action listed at the end. Sometimes this does correct the fault. Sometimes it doesn’t. I emphasize that the section called Take “corrective” action is not often the solution to the problem. Immediately open and inspect a piece of equipment. This is usually not an early step. Opening and inspecting equipment costs usually more than $4000. There are many, many very simple tests you can do to help you iden-

587

588

Appendix J Guide for Students: How You Can Get the Most from this Book

tify whether or not open and inspect is needed; and to identify which equipment to open and inspect. Work systematically. Start simply. Check for consistency as a check of the sensors Since sensors may be incorrect, one of the first checks might be to look for consistency between sensors reading the same variable on the same stream. In the format of this text, I rarely include a section called “consistency tests”. I expect you to identify neat ways to check for consistency. Methods you can use include:

1) 2) 3) 4)

to compare two sensors at the same location, agreement between composition, temperature and pressure (say at the top or bottom of a distillation column), agreement between temperature and pressure on a pressure–enthalpy diagram for pure refrigerant, to check that the conditions on a stream are the same at two locations. This latter type of information we often obtain by contacting the operators of the utilities or of the plants upstream or downstream from our process.

Other actions you can select are: . . . . . . . . . .

visit the site, check that your P&ID agrees with reality, do on-site simple tests, gather data and do calculations, checking the sensors and the control system, get information from vendors, sample and analyze, do more complicated tests, open and inspect, and take corrective action.

Here are the details. Use fundamentals to guide the selection of the information you gather. I do not include an activity called “Calculate fundamentals”. You are expected to gather the information needed and perform these. However, to make life easier, I do identify the information gathered and include the results of the calculations. Once you have zeroed in on the fault you might need to open and inspect. Or, you might want to take corrective action. Here are more details of each. Check for consistency through Contact with on-site specialists. These might include operators running the various utilities (steam, cooling water, refrigeration, power, waste water, flare, storage) and those running various processes that interact with your unit. The latter are particularly useful. They help provide “consistency” tests for data. In other words, the reading on the temperature sensor of a stream leaving your unit should agree closely with the temperature recorded for the same

J-2 Try a Trouble-Shooting Case where the Problem is Reasonably Well Defined

stream when it reaches another unit (unless additional processing occurred on that stream between units). The process-control specialists are contacted separately. Visit site, read present values, observe and sense. Walking out of the control room and visiting the actual equipment can provide great insight simply from our senses: the noises, smells and sensors. The focus here is on the senses and reading instruments whose values are not recorded in the control room. (The actions do not include actual adjusting valves, measuring temperatures, and simple tests.) The sounds to listen for include pump cavitation, pressure-relief valves “blowing”. The level of liquid in a vessel can sometimes be sensed by tapping on the side of the vessel and listening for the change in sound (consistent with the gas–liquid interface). The sights to look for include the values of the instruments, rust stains on the outside of overhead condensers (indicating that in the past, water flowed over the outside to try to improve condensation), the condition of the flare, the intermittency of the steam discharge from traps, the condition of the insulation, whether steam lines come off the top of the main (the steam tends to be wet and contain rust if it comes off the bottom of the main), whether the condensate lines entering the condensate header go in the top (bottom entry means that the traps are flooded with condensate from other locations on the site), the location of the valve stem, the signal to the valve. The sights also include checking that the motor shaft is rotating in the direction indicated on the pump casing; that the direction of flow through the valve is the same as marked on the valve, that the tab on the orifice plate suggests that the sharp edge of the orifice is on the upstream direction and the orifice is the correct diameter. The smells include anything unexpected; hot smell (suggesting an overheated motor); hydrocarbon smell suggesting a leak. Check diagram and P&ID versus what’s out on the plant. You should have done this before trouble occurred. As soon as you are assigned to a unit, spend the time creating a record of all the lines and layout on the plant. You should be able to identify what is in every line and state the direction of flow. Many P&IDs in the office are out-of-date. You need an accurate idea of the actual layout and interconnects. For the purposes of the cases, I provide that information, as it pertains to most of your potential hypotheses, in this activity. This is included here, instead of earlier under “office” actions because regardless of how up-to-date you think your information is, it is always wise to check on the plant at the time the trouble-shooting case arises. On-site simple tests. This is usually the bread and butter activity. You might use a bucket of cold water and measure the amount of condensate. You might ask for “turn and seal” tests on valves to check that they seem to be working correctly. Soap tests on flanges can be used to detect leaks from pressurized systems. Leaks in a vacuum systems may be identified by isolating the system and noting the change in vacuum over time. Couplings can be checked for wear and misalignment. For each case, I have tried to include options that might be useful to check a wide variety of hypothetical causes.

589

590

Appendix J Guide for Students: How You Can Get the Most from this Book

Trend checks. Look for variables that are linked by fundamentals and check that they behave accordingly. For example, for distillation an increase in pressure usually corresponds to an increase in temperature for constant composition. If the process is cycling, ask questions about the frequency and amplitude. Look for patterns. For each case, I have tried to include options that might be helpful. Gathering data for equipment performance calculations. The focus here is on gathering the information needed to apply the fundamentals: to complete a mass balance, to rate a heat exchanger, and to check on the pressure profile. For each case, I have tried to include options that might be useful to check a wide variety of hypothetical causes. Sensors: check response to change. Before an instrument is replaced, two simple tests can be done: 1) check the response to change and 2) check on calibration. When conditions are changed slightly, the sensors should respond accordingly. Do they? If they do not, they might be broken. Secondly, the sensor might respond to change but the numerical reading might not be accurate. Accuracy is checked via comparing with measurements from temporary instruments or via calibration (mentioned later). Sensors: use of temporary instruments. Laser or contact temperature sensors can often add confirmation to the temperatures at key locations in the plant. Since steam traps often cause problems, the distinctive noise of traps in operation (heard through a stethoscope) can often help identify the problem. Clamp-on ammeters, and power-factor meters are useful. Sensors: calibrate. If the sensor is suspect, often it can still be used if it is recalibrated. This three to four hour task might be as simple as submerging the thermocouple is ice water. Specialists are usually required. Control system. If a control problem is suspected, putting the system on manual is a good tactic to use. Sometimes retuning the control system corrects the problem. Call to vendors, licensee, or suppliers. A phone call may often provide valuable insight about a catalyst, a new source of raw material, or trouble-shooting experience from a vendor. The difficulty often is that the key person who may have the information is not available when you need them. The cost allocated to this activity varies arbitrarily from case to case to account for the uncertainty of getting answers when you need them. Samples and measurements. This sounds easy. However, for cycling or oscillating system the challenge is to obtain samples that are going to help identify the cause of the cycling. More complicated tests. More complicated tests include gamma scans; pressure testing exchanger tubes for leaks. Open and inspect. Isolate the equipment, make it safe to open. Open and inspect. This is expensive and usually a last resort. .

Take “corrective” action. Sometimes this corrects the fault. Sometimes it doesn’t. I emphasize that the section called Take “corrective” action is not often the solution to the problem.

This ends the list of possible actions. Now, what about the actual fault?

J-6 Trouble Shooting on the Job

The fault and closure for the case. Identifying the fault, correcting the fault and preventing the fault from reoccurring are left to you. There usually are no diagnostic tests listed for the case that tells you the fault. There is sufficient evidence from the actions to help you identify the fault. Write out the fault, think about how you would correct and prevent the fault and then check with the feedback in Appendix E. Next, complete the form about the TS process, given in Chapter 2, Table 2-3. Reflect on the thinking process you used, set goals for improvement for the next case.

J-3

See How Others Handle a Case

Then I would read the five cases at the end of Chapter 1, Cases ’3 to 7, and work through the stories of the five trouble shooters given in Chapter 4. Spend time with the reflective pauses given in each story. Compare how you approached the situation with the approach taken by Michelle, Pierre, Dave, Saadia and Frank.

J-4

Pause, Reflect on the Pretest, and Invest Time Polishing Specific Skills

Now that you are familiar with the process and have tried six, guided actions, reflect on the results of the pretest, Section J-1, identify any specific prerequisite skills you want to polish and invest time developing your confidence with that skill before tackling the structured set of Cases in Chapter 8.

J-5

Work your First Cases Starting with Case ’19

The simplest cases start with Case ’19. Use the coding related to difficulty to guide you. Select processes with which you are more familiar. Start simply.

J-6

Trouble Shooting on the Job

In this book the actions are selected and answers are given. However, it is never stated who actually does the tests. On the job the engineer rarely does any of the tests. The engineer asks questions, requests that tests be done, interprets the results and proposes actions, but rarely does the engineer do such things as turn valves or change set points. The cardinal rule is that the process site is the responsibility of the process operator. This is their turf. Engineers are expected to check in with the

591

592

Appendix J Guide for Students: How You Can Get the Most from this Book

process operators before they go out and look at equipment. Engineers do not alter valves; the operators do. Building and developing trust with the process operators is crucial for successful trouble shooting.

J-7

Summary

Here is provided a simplified guide, for students, on how to use this book. Start by getting the big picture and self-assessing your prerequisite skills. Then start simply; try trouble shooting Case ’8 for which a TS Worksheet has been completed. Details are given of the rationale and structure used in developing the cases and for presenting the optional actions available for each case. Then, reflect on how your troubleshooting processes compare with approaches of others by working through the five case stories in Chapter 4. Polish the prerequisite skills and then start simply and work through the cases in Chapter 8. Finally, in working the cases in this book details were not given as to who performed the tests. Indeed, it is incorrectly inferred that the engineering trouble shooter did the tests. Rarely do process engineers actually do the tests; tests are done by the process operators or by specialists.

Trouble-Shooter’s Worksheet 2-1: Succinct summary (D. R. Woods and T. E. Marlin)

1. Engage: Gather initial information. .

. . .

Establish if emergency priority: safety? damage? shut down or safe-park or continue? Describe what’s going on. Manage panic: “I want to and I can.” Monitor: Have you finished this stage? Can you check? What next?

2. Define the stated problem: based on given information. If the information is not known at the stage, gather it later. IS

IS NOT

WHAT

(should be happening but it is not)

WHEN

?& ?&

WHERE

?& ?&

WHO

?& ?&

Identify situation as 1) startup new process; 2) startup after maintenance or change, 3) usual operation. Monitor: Have you finished this stage? Can you check? What next? 3. Explore: Exercise? or problem? Strategy for change or basics? Useful to broaden with Why? Why? Why? Gather information. Perspectives: customers? suppliers? weather? changed economics? politics? environment? . . . . .

Prioritize: product quality, production rate or profit? Goal: safe-park? short term? long term? SMARTS$ Data consistency? Pertinent fundamentals? Likelihood of problem type. Explore with What if? List changes made and/or list trouble-shooting experience: root causes based on symptom. (Chapter 3)

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

. .

Brainstorm hypotheses Hypotheses and evidence of symptoms:

Evidence of symptoms: a. b. c. d. e.

Working Hypotheses

Initial Evidence a

b

c

Diagnostic Actions d

e

1 2 3 4 5

S = supports; D = disproves, N = neutral

Diagnostic actions: A. B. C. D. 4. Plan 5. Do it 6. Look back

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

A

B

C

D

Worksheet 2-2: Summary observation form for feedback about Trouble Shooting.

TS name _____________________ Case _____ Initials ES _____ Obs ___ Rough work area: Process: how Data/analysis: what Monitoring _____________________ Data resolution _______________ Checking _______________________ Fundamentals? _______________ Systematic ______________________ Reasoning ___________________ Subs and perspective _____________ Completeness ________________ Decision making: how Priorities ______________________ Bias ___________________________

Synthesis: what Hypotheses __________________ Flexibility ____________________

Rating and Feedback Clarity of Communication None

Some

Most

All

Some

Most

All

Some

Most

All

Some

Most

All

Some

Most

All

Process used: None

Data collections and analysis: None

Synthesis: None

Decision-making: None

Five Strengths: ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ Two areas for improvement ________________________________________________________________ ________________________________________________________________ Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Trouble-Shooter’s Worksheet 2-3: Case ’8: The depropanizer: the temperatures go crazy (Donald R. Woods and Thomas E. Marlin)

1. Engage: Write down what is said; what you sense, smell, hear. If someone is telling you, then use skilled reflective statements to ensure you accurately obtain the information. .

.

.

.

Emergency priority: Safety? Hazard? Equipment damage? shut down &; safe park &. If not &, then: Draw a sketch of the process and mark on values. Provide a description in words of what is going on. This is a simple distillation column but all the piping and instrumentation details make it look complex. On the LHS, feed from upstream processing enters drum V29. This feed is pumped (via either a steam driven or motor driven centrifugal pump, F25, 26) through a preheater, E24, and into the depropanizer at tray 18. As the name suggests, the purpose of this column is to take overhead “propane and all lighter species”. Let’s follow the overhead. The overhead is condensed in two condensers in series, E25, collected in overhead drum V-30 with the non-condensibles (such as methane and hydrogen) removed from the drum and vented to the fuel-gas system. The pressure on the column, C8, is controlled by the valve on the vent system, PV 10. Condensed propane is pumped, F-27, from the drum V-30 forward as product, through product cooler E-26, and returned to the column as reflux. The reflux is flow controlled. Following the bottoms: a thermosyphon reboiler is steam heated. The bottoms flows forward to the next column, the debutanizer. No pump is needed because of the pressure difference between the depropanizer, 1.7 MPa, and the debutanizer, 0.48 MPa. I’m not sure at this stage if this is a control “problem” so I won’t elaborate further on the system at this time. I also will focus on the depropanizer, and not explore the debutanizer at this time. Manage any panic you might feel by saying “I want to and I can. I have a strategy that works. Let’s systematically follow it.” Monitor: Have you finished this stage? Can you check? What next? I’ve systematically followed my way around the flow diagram. I think I understand enough for now.

2. Define the stated problem: Systematically classify the given information using IS and IS NOT. If the information is not known at the stage check? & to remind you to gather this information

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

IS

IS NOT

(should be happening but it’s not) tray temperatures steady; bottoms level steady. ? & before the pressure increase; running well for several shifts. First plant startup.

WHERE

tray temperatures “go crazy”; bottoms level decreases. ? & 10 minutes after the pressure in column C8 increased by 0.1 MPa. ? & depropanizer, C8.

WHO

? & new operator.

WHAT WHEN

?& maybe upstream; no information about downstream debutanizer, yet! ? & not with previous operators.

– startup new process & suggest use Basics – startup after maintenance or change & suggest use Change – usual operation but changes made in operation but not in equipment & suggest use Basics – usual operation & suggest use Basics. .

Monitor: Have you finished this stage? Can you check? What next? Yes. I think I’ve finished.

3. Explore: Gather information to be gathered ? & in Define stage &. Perhaps questions about downstream effects. The decrease in liquid level could be because the flow has increased to the debutanizer (because the Dp) has increased. Exercise? & or a problem? & . I haven’t seen anything like this before Strategy: change & or basics & . Perspectives. Why? Why? Why? no information at this time that suggests this might be useful. ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________ Why? › Start ﬁ _________________________________________________________

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

. .

Prioritize: product quality & ; production rate &; profit & Goal: safe-park? & : short term now with long term later & ; long term now &

Action to be achieved: Specific terms and Measurable: level out the temperatures and the bottoms level Attainable? total reflux is a start but I hope it’s attainable Reliable? depends on my short-term solution will work on solving it quickly Timely? cannot think of major hazard now Safe? $ .

.

.

Check consistency of data/symptoms: inter-data consistency? OK & no & data consistent with fundamentals? OK & no & Type of problem: startup new process & maybe mechanical electrical failure usual operation & : ambient temp? & maybe fluids problems; high temperature? & then maybe materials problems System? failure of heat exchanger & > rotating equipment & > vessels & > towers & Identify key and What if?

What if? then What if? then What if? then What if? then

.

temperatures “going crazy” = temperature cycling focus on cycling symptoms / causes only temperature “cycling” and no decrease in level bottoms and tops temperature and pressures should be cycling too only bottoms level drop and no temperature “going crazy” root cause related to bottoms level drop column pressure increases condensation temperature at top increases; DT condenser increases and condensation should be easier; boiling temperature at bottoms increases; DT reboiler decreases so might shift from film to nucleate boiling giving higher heat flux, causing increased boilup or if nucleate to start with then insufficient area and boilup decreases.

List changes made & and/or trouble-shooting causes based on symptom & .

“Crazy temperatures” and decreasing bottoms level sounds like a control problem. From Chapter 3, no symptoms listed for bottoms level dropping, but symptoms related to “oversized condenser” and “undersized reboiler” are:

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

“Insufficient boilup”: [ fouling on process side]*/ condensate flooding, see steam trap malfunction, Section 3.5 including higher pressure in the condensate header/ inadequate heat supply, steam valve closed, superheated steam/ boiling point elevation of the bottoms/ inert blanketing/ film boiling/ increase in pressure for the process side/ feed richer in the higher boiling components/ undersized reboiler/ control system fault/ for distillation, overdesigned condenser. But if there was “insufficient boilup”, then the bottoms level should be increasing and not decreasing. This doesn’t make sense?? For: “Cycling of column temperatures:” controller fault/ [ jet flooding]*/ [downcomer flooding]*/ [ foaming]*/ [dry trays]* with each of the []* items listed as separate symptoms with their own root causes. [ jet flooding]*: excess loading/ fouled trays/ plugged holes in tray/ restricted transfer area/ poor vapor distribution/ wrong introduction of feed fluid/ [ foaming]*/ feed temperature too low/ high boilup/ entrainment of liquid because of excessive vapor velocity through the trays/water in a hydrocarbon column. [downcomer flooding]*: excessive liquid load/ restrictions/ inward leaking of vapor into downcomer/ wrong feed introduction/ poor design of downcomers on bottom trays/ unsealed downcomers/ [ foaming]* [ foaming]*: surfactants present/ surface tension positive system/ operating too close to the critical temperature and pressure of the species/ dirt and corrosion solids. [Dry trays]*: flooded above/ insufficient reflux/ low feedrate/ high boilup / feed temperature too high. .

Brainstorm root causes: summary of major ideas generated:

change in feed, too much overheads in feed, not enough feed, tray collapsed in stripping section, too much vaporized feed, pump F26 failure, pump F26 cavitates, increased pressure and DT increases, dry tray, flooded in rectification, insufficient reflux, low feedrate, feed temperature too high, boilup too high, too much feed to debutanizer, leak in bottoms, vaporizer flashes 90% (instead of 67%), failure of check-valve on idle pump outlet, boilup controller fault. Some of these are symptoms and not root cause, e.g. “not enough feed” “dry trays” .

Hypotheses: list in Chart; Symptoms: code and list in chart; Analyze with S supports; D disproves and N neutral or can’t tell.

Symptom a. 10 min after column pressure increased, column temperatures go crazy b. 10 min after column pressure increased, bottoms level decreases c. d. e.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Working Hypotheses

1. tray collapsed stripping section 2. too much bottoms fed to debutanizer 3. too much overheads in feed 4. feed valve FV1 stuck 5. pump F-26 not working 6. check valve on idle pump allows backflow 7.

Initial Evidence a S N N S S S

b S S S S S S

c

d

Diagnostic Actions e

A 4 4 4

B

C

D

4 4 4

Diagnostic actions: A. B. C. D.

readings of instruments on column visit site and listen to pump for cavitation visit site and see location of valve stem on FV-1 shut isolation valves on idle pump

4. Plan Select “read instruments” as the first task because it is inexpensive and should help test many of the hypotheses. Many of the key variables are displayed in the control room. 5. Do it Go to the control room, notebook in hand. 6. Look back

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Trouble-Shooter’s Worksheet 2-4: (Donald R. Woods and Thomas E. Marlin)

1. Engage: Write down what is said; what you sense, smell, hear. If someone is telling you, then use skilled reflective statements to ensure you accurately obtain the information. .

.

.

.

Emergency priority: Safety? Hazard? Equipment damage? shut down &; safe park &. If not, & then: Draw a sketch of the process and mark on values. Provide a description in words of what is going on. Manage any panic you might feel by saying “I want to and I can. I have a strategy that works. Let’s systematically follow it.” Monitor: Have you finished this stage? Can you check? What next?

2. Define the stated problem: Systematically classify the given information using IS and IS NOT. If the information is not known at the stage check ? & to remind you to gather this information

IS

IS NOT

WHAT

(should be happening but it is not)

WHEN

?& ?&

WHERE

?& ?&

WHO

?& ?&

– startup new process & suggest use Basics – startup after maintenance or change & suggest use Change – usual operation but changes made in operation but not in equipment & suggest use Basics – usual operation & suggest use Basics. .

Monitor: Have you finished this stage? Can you check? What next?

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

3. Explore: Gather information to be gathered ? & in Define stage & Exercise? & or a problem? &. Strategy: change & or basics &. Perspectives. Why? Why? Why? ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________ Why? › ____________________________________________________________ Why? › Start ﬁ _________________________________________________________ . .

Prioritize: product quality & ; production rate &; profit & Goal: safe-park? &: short term now with long term later &; long term now &

Action to be achieved: Specific terms and Measurable: __________________ Attainable? ______________________________________________________ Reliable? ________________________________________________________ Timely? _________________________________________________________ Safe? ___________________________________________________________ $ _______________________________________________________________ .

.

.

Check consistency of data/symptoms: inter-data consistency? OK & no & data consistent with fundamentals? OK & no & Likelihood of problem: startup new process & maybe mechanical electrical failure usual operation &: ambient temp? & maybe fluids problems; high temperature? & then maybe materials problems System? failure of heat exchanger &> rotating equipment & > vessels & > towers & Identify key and What if?

What if? then What if? then

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

What if? then .

. .

List changes made & and/or trouble-shooting causes based on symptom &. Brainstorm root causes: Hypotheses: list in Chart; Symptoms: code and list in chart; Analyze with S supports; D disproves and N neutral or can’t tell.

Symptom a. b. c. d. e. Working Hypotheses

Initial Evidence a

b

c

Diagnostic Actions d

e

1 2 3 4 5 6 7

Diagnostic actions: A. B. C. D.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

A

B

C

D

Worksheet 3-1: Rating form for teams

Assessment of your team and team meeting

Name: _________________ Date: __________________ Purpose of team: ______________________________________ unclear & Purpose of this meeting ________________________________ unclear & Agenda for this meeting: detailed, clear and circulated ahead of time bare minimum circulated ahead of time none

& & &

Three-minute team task to seek consensus about the rating of the Task and Morale: .

Teamwork: Task all members clear about and committed to goals; all assume roles willingly; all influence the decisions; know when to disband for individual activity; all provide their unique skills; share information openly; the team is open in seeking input; frank; reflection and building on each other’s information; team believe they can do the impossible; all are seen as pulling their fair share of the load.

The degree to which these descriptors describe your team’s performance (as substantiated by evidence: meetings, engineering journal, interim report, presentations). None of Few of these these behaviors but behaviors major omissions & 1

& 2 .

& 3

& 4

Most features demonstrated

All of these behaviors

& 5

& 7

& 6

Teamwork: Morale: Trust high, written communication about any individual difficulties in meeting commitments; cohesive group; pride in membership; high esprit de corps; team welcomes conflict and uses methodology to resolve conflicts and disagreements; able to flexibly relieve tension; sense of pride; we attitude; mutual respect for the seven fundamental rights of all team members; Absence of contempt, criticism, defensiveness and withdrawal.

The degree to which these descriptors describe your team’s performance (as substantiated by evidence: meetings, engineering journal, interim report, presentations).

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

None of Few of these these behaviors but behaviors major omissions & 1

& 2

& 3

& 4

Most features demonstrated

All of these behaviors

& 5

& 7

& 6

Each, in turn, gives a 30-second summary of his/her perception of his/her contribution. This is presented without discussion. Individual, 30 second reporting of his/her contribution to this meeting: ___________________________________________________________________ ___________________________________________________________________ ___________________________________________________________________ ___________________________________________________________________ Four-minute team task to reach consensus about the five strengths and the two areas for growth. Strengths of your team

Areas to work on for growth

D.R. Woods (2005)

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Worksheet 5-1: Reflections: A place for you to record your ideas about the TAPPS Method:

Being the talker: What did you enjoy most? What was most difficult about the task? What did you discover by being the talker? What did you discover from interacting with a listener? What were your strengths? Focus on accuracy? Very few silent periods? Being active? Good communication? ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ Being the listener: What did you enjoy most? What was most difficult about the task? What did you discover by being the listener? What did you discover about problem solving by comparing the talker’s approach to yours? What were your strengths as listener? Quality of your prompts and degree of interaction? Tone of interaction? Good communication? Non-intrusiveness? ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ ________________________________________________________________

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Worksheet 5-5: Evidence for listening: Feedback to listener & the listener will encourage verbalization, an emphasis on accuracy, active thinking and encourage the problem solver to move the marker correctly on the strategy board. Your interventions will be judged by the problem solver to be helpful, and not judged to be disruptive.

Activity 1: Talker ____________

Case ___________

encourage verbalization: encourage emphasis on accuracy: encourage active thinking interventions:

not needed not needed not needed not needed

Listener _____________

interruptive interruptive interruptive interruptive

OK OK OK OK

really helped really helped really helped really helped

Comments: ___________________________________________________________________ ___________________________________________________________________

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Worksheet 5-4: Record of the talker’s strategy with . for monitoring statements.

Talker __________________ Case _________ Listener ____________________ Stage Engage: “I want to and I can!” Define-the-stated problem: Sort the given problem statement Explore the problem to discover what the problem really is Plan Do it Look back: elaborate, check

0

2

4

6

8

Time, minutes

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

10

12

Worksheet 5-3: Feedback to the listener:

problem _______________________

listener __________________________

I found the listener: .

–10

The quality of the comments: –8

–6

–4

–2

J

–2

–4

–6

–8

–10

too passive little too passive about right a little interruptive too interruptive .

–10

The attitude displayed: –8

–6

–4

–2

J

–2

–4

–6

–8

–10

too passive little too passive about right a little interruptive too interruptive .

The listener’s emphasis was listening to me &; helping me verbalize &; helping me solve the problem &; solving the problem for me &.

validated by talker __________________________________________________

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Worksheet 5-2: Feedback from the listener to the talker about the process used.

Awareness

____________________________ ___________________________ problem listener

Number of silent periods 0 1 2 3 4 5 >5 Number of checks, double checks >5 5 4 3 2 1 0 Amount of writing/ charting >5 5 4 3 2 1 0 Comments: ___________________________________________________________________ ___________________________________________________________________ Validated by: __________________________________

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Worksheet 6-1: The Tom Dayton murder (adapted from Sherlock Holmes).

Tom Dayton had many enemies. (a) He was a scalawag and a prankster who never passed up an opportunity to embarrass someone through a practical joke (b). It was Tom who invented the joy buzzer and the whoopee cushion, and some even credit him with having originated fake vomit. It is well known that Tom’s favorite target was his old headmaster, Stanley Bosworth, (c) at Bromley School. Stanley Bosworth was the victim of some of Tom’s most elaborate pranks. Tom’s eventual marriage to Bosworth’s daughter, Melissa, was considered by many to be Tom’s ultimate joke (d) on the respected headmaster. Among the more prominent victims of Tom Dayton’s past pranks were Judge Walter Brighton (e), Lord and Lady Morton (f) of Westchester, banker Mortimer Fawcett (g), Doctor Fabian Peerpoint (h), and tobacco merchant Dawes Flescher (i). All of the above were present at the dinner party held on the Bosworth yacht in honor of Stanley Bosworth’s 60th birthday (j). Following an uneventful dinner, most of the guests retired to their staterooms to freshen up. The clock in the dining room struck 10 pm (k) when a shot rang out (l). Most later claimed they heard a second shot (m). All aboard the yacht, including the yacht’s captain, Jonas Fenton, (n) and cook, the curvaceous Mildred Weekson (o), arrived at Tom Dayton’s stateroom to find him dead – shot in the forehead (p). A smoking revolver lay near the doorway (q); Tom’s body lay on the floor across the room (r), just below an open porthole (s). Scrawled in the dust near Dayton’s body were the initials SB (t). In the corner of the room was a suitcase filled with $500,000 (u). No bullets were found in the walls, ceiling or floor of Tom’s room (v). Who killed Dayton? During the investigation the following factual evidence was produced: w. Although just about everyone claim they heard two shots, Tom Dayton had one bullet in his head. x. Lady Morton was being blackmailed. y. The clock in the dining room was 15 minutes slow. z. The $500 000 in the suitcase was counterfeit and was accompanied by a withdrawal slip for $1 000 000 from Mortimer Fawcett’s bank. aa. Bosworth angrily stated at dinner that he felt Tom was mistreating his daughter. bb. The actual murder weapon was found under water below the porthole. cc. Dr Fabian Peerpoint advocates mercy killing. dd. The smoking revolver in the room belonged to Stanley Bosworth. ee. Tobacco merchant Dawes Flescher is a talented mountain climber. ff. Jonas Fenton, the yacht’s captain, went to school with Tom Dayton.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

gg. Dr Fabian Peerpoint revealed that Tom Dayton was terminally ill with only a few months to live. hh. The bullet in Tom’s head did not come from the smoking revolver on the floor in Tom’s room. ii. Melissa says that she was visiting her father in his stateroom from 9:45 pm until she heard the shot. jj. Mildred Weekson was having a secret affair with Tom Dayton for the past two years; and with Lord Morton. kk. The shot that killed Tom Dayton was fired from outside the porthole. Directly below the porthole is water. Based on the evidence so far would you: 1. 2. 3.

accuse ____________ of murder based on evidence (list the letters of the evidence supporting your conclusion) ____________________________. accuse Dr. Peerpoint of mercy killing based on evidence (list the letters) _____________________________________________________________ conclude that Tom died from __________ based on evidence. (list the letters) ________________________________________________

or... . 4. require that the following information is needed before any conclusion can be drawn: ll) what pranks had Tom played on _________________________ mm) check for poison in Tom’s body nn) other _______________________________________________

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Worksheet 7-1: Trust.

Trust is having the confidence that you can mutually reveal aspects of yourselfand your work without fear of reprisals, embarrassment or publicity. Trust works both ways: you trust them and they trust you. Trust is not developed overnight, trust takes time to develop. Trust can be destroyed by one incorrect act. Check your current status Building your trustworthiness getting them to trust you

already do this

needs some work

need lots unsure of work if this is for me

1. Do what you say you will do. 2. Be willing to self-disclose: don’t hide your shortcomings; share yourself-honestly. 3. Listen carefully to others and reflect to validate your interpretation. 4. Understand what really matters to others; do your best to look out for their best interests. 5. Ask for feedback. 6. Don’t push others to trust you more than you trust them. 7. Don’t confuse “Being a buddy” with trustworthiness. 8. Tell the truth. 9. Keep confidences. 10. Honor and claim the 7 RIGHTS. 11. Don’t embarrass them.

& &

& &

& &

& &

&

&

&

&

&

&

&

&

& &

& &

& &

& &

&

&

&

&

& & & &

& & & &

& & & &

& & & &

Checking your trustworthiness do they trust you? always

most times

sometimes

don’t think applies

1.

&

&

&

&

&

&

&

&

& &

& &

& &

& &

& &

& &

& &

& &

2. 3. 4. 5. 6.

Do they disclose confidential information trusting that you will keep it confidential? Do they assign you challenging tasks to do without frequently checking up on you? Do they honor your RIGHTS? Do they seem to look out for your best interests? Honest and forthright. Do not leave you feeling that they haven’t told you everything about the situation; they seem to be holding back. copyright 1999 , Donald R. Woods

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Worksheet 7-3: Feedback about your environment.

To what extent do you agree with the following descriptors of your environment where you usually “trouble shoot”. People are willing to admit error: “The people that I work with are very unwilling to admit errors; they blame others, they pass the buck and, if necessary, would purposely mislead me rather than to admit error.” Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________ Encourage risk taking: “Risking is rewarded. We are expected to take risks about 10 times a day. Risks should be wisely, not indiscriminately selected. But, nevertheless, we are not only encouraged but we are rewarded for risk taking.” Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________ General stress at work they are under: “Their environment is very stressful. People have many deadlines and interruptions. The consequences of making mistakes is very high. The issues are complex. The environment changes often and includes a lot of uncertainty.” Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________ General stress at work you are under: “My environment is not stressful. I do not have many deadlines or interruptions. The consequences of making mistakes are low. The issues are straightforward. The environment is safe, stable and secure.” Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

People’s listening and responding: “The people are open, communicate well, can clearly identify facts, will offer opinion when asked for, and are very competent but are not aggressive “know-it-alls””. Strongly Moderately Slightly Slightly Moderately Strongly Disagree Disagree Disagree Agree Agree Agree 1 2 3 4 5 6 ________________________________________________________________

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Worksheet 7-2: The problem given in Case ’12 arises at 8:30 am. The following background and resources are available. Identify which ones you think are directly under your control. If you think you have control over the item, circle Y; if you do not have control over it, circle N.

1. The safety officer, Hack, is overbearing, not liked and gets carried away about simple things. 2. The laboratory can analyze liquid samples with their equipment but gaseous samples cannot be analyzed because their instrument is broken. 3. The lab schedule is busy with top priority analyses. Samples could not be analyzed until after 3:00 pm. 4. The union prevents you from analyzing any samples; if you do, there will be a strike. 5. The upstream styrene plant is operating at 80% capacity. 6. The upstream ethylene cracking furnace is operating at 95% capacity. 7. The upstream propylene plant is shut down. 8. The operator on the ethylene plant is cooperative. 9. Samples can be taken at any of the sewer gates within the Battery Limits and at A, B, and C. 10. No blueprints are available for the sewer system. 11. Your performance review to establish your salary is being done next Thursday at 4:30 pm. 12. You have to prepare your “record of progress” record for the performance review on Thursday.

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

Y

N

Y

N

Y

N

Y Y

N N

Y Y Y

N N N

Y Y

N N

Y

N

Y

N

Index

Index 12/1 rule 48 20 min rule 48 80/20 rule 22 85/15 rule 22

a Absorption, gas – cases involving 383, 387 – causes 576 – fundamentals 67 – symptoms-cause 73–78 Accuracy, need to focus on 20 – example check and double check – in Case ’5 141, 142, 143 – in Case ’6 147, 151 Activated sludge, symptoms-cause 106 Active, importance of being mentally active 20 – Case ’7 158–160 – example Case ’6 145 – to aid in TS 25 AAI, Adhesion angle index, defined 108 Adsorption, gas, fundamentals 67 – Case ’6 12–13, 144ff – cases involving 12, 247, 299, 370 – symptoms-cause 77 Adsorption, liquid, fundamentals 67 – symptoms-cause 77 – cases involving 181, 247–248 Aerobic bioreactors, symptoms-cause 96 – foaming 129 AI, Arching Index – cited 56, 57 – defined 108 Alkylation, decanters, symptoms-cause 83 – reactors, symptoms-cause 99 Ambiguity, – in terminology 226 – in words or instructions 218

Amine absorbers, causes of corrosion in 46 – symptoms-cause 73–75, 129 Amine flash drum, symptoms-cause 85 Ammonia process, cases involving 1, 171, 317, 339, 357, 383, 387, 391 Anaerobic digesters, symptoms-cause 97 Analysis (as part of the Define stage) see also Classification – example use in TS Worksheet for Case ’8 36 – second stage in a generic problem solving process 20, 21 Anderson, R. B., cited 9 Anxiety, and the Engage stage 21 Argentino, Mark, cited 291 Assertiveness, how to 48 Assessment, self, see Self assessment and Feedback Assumptions, – and reasoning 224 – definition 224, 230 – diagram 229 – formulate 230 Attitude toward problem solving 22 Awareness of problem solving process, importance of, 20 – activity to improve your skill, 166–172 – feedback form, 172–173 – target skills, 166

b Bag filters, symptoms-cause 81 – cases involving 222, 333 Bagging machine, cases involving 222, 333 Barton, J., cited 297, 407 Basadur, M., cited 180, 407 Basics strategy for TS: – contrast with change strategy 23 – description 23

Successful Trouble Shooting for Process Engineers. Don Woods Copyright 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-31163-7

I 1

I 2

Index – diagnostic actions to help select 197 – example questions to ask 24 – example use, Case ’4 139 – Case ’5 141 – Case ’6 145 – how to apply 24 – use of fundamentals in 23 – when to use 4, 23 Bayes theorem, use 25 BDI, Bin density index – cited 56, 58 – defined 108 Bernoulli’s equation, use 51, 58, 419, 565 Bias, personal 25, 205 – in collecting evidence 205 – in reaching conclusions 205 – self test 206 – types: – anchoring 25, 205 – availability 205 – confirmation 25 – inadequate synthesis 206 – misinterpretation 206 – need to avoid 195, 205 – overinterpretation 25 – premature closure 25, 205 – pseudodiagnosticity 25, 205 – representative 25, 205 – test 206 – underinterpretation 206 Bins, symptoms-cause 126 – cases involving 182, 222, 347 Blenders, solid, symptoms-cause 108–109 – cases involving 333 Bloch, H.P. 44, 403, 408 Blowers, symptoms-cause 52 Boiling film 63 – nucleate 63 198 267 Brainstorm, see Creativity Bubble formation, symptoms-cause 110 Bubble reactors symptoms-cause 98 Bucket elevators, symptoms-cause 56

–

–

–

–

–

–

–

–

c Carry out the plan see Do it Cases, overall list of, with ratings 268–271 – Case ’1 1 – answer 2 – Case ’2 1 – answer 2 – Case ’3 10 – answer 133–138 – critique 212

–

–

–

– reasoning 224–232 – evidence 227, 228 – meaning 226 – diagram of cause-effect 228, 229 – classification 224 – conclusion 225 – context 225 – worksheet 423 Case ’4 10 – answer 138–140 – cited 180 – worksheet 427 Case ’5 11 – answer 141–144 – patterns 223 – worksheet 430 Case ’6 12 – answer 144–156 – worksheet 202, 217, 217–218 Case ’7 13 – answer 157–162 – cyclical 223 – Why? 180 Case ’8 35 – actions 271 – answer 36–39, 537 – destroyers activity 240 – listening activity 239 Case ’9 181 – actions 274 – answer 538 – brainstorm 190 – identify symptoms 220 Case ’10 186–87 – actions 278 – answer 538 – cyclical 223 – example approaches 204 – patterns 223 Case ’11 214 – actions 282 – answer 540 – hypotheses 216 – root cause 217 Case ’12 247 – actions 280 – answer 540 Case ’13 251 – actions 282 – answer 541 Case ’14 252 – actions 285 – answer 541

Index – Case ’15 – answer – Case ’16 – answer – Case ’17 – answer – Case ’18 – answer – Case ’19 – answer – Case ’20 – answer – Case ’21 – answer – Case ’22 – answer – Case ’23 – answer – Case ’24 – answer – Case ’25 – answer – Case ’26 – answer – Case ’27 – answer – Case ’28 – answer – Case ’29 – answer – Case ’30 – answer – Case ’31 – answer – Case ’32 – answer – Case ’33 – answer – Case ’34 – answer – Case ’35 – answer – Case ’36 – answer – Case ’37 – answer – Case ’38 – answer – Case ’39 – answer – Case ’40 – answer – Case ’41

255 33 256 433 256 433 257 434 286 543 289 543 291 544 294 544 297 544 299 545 304 546 308 546 312 547 314 548 317 548 321 548 323 549 327 549 330 550 333 550 335 550 339 551 342 551 345 552 347 552 350 553 354

– answer 555 – Case ’42 357 – answer 556 – Case ’43 362 – answer 556 – Case ’44 364 – answer 557 – Case ’45 367 – answer 558 – Case ’46 370 – answer 559 – Case ’47 374 – answer 559 – Case ’48 377 – answer 560 – Case ’49 380 – answer 561 – Case ’50 383 – answer 561 – Case ’51 387 – answer 562 – Case ’52 391 – answer 563 – Other cases 403 – Feedback about hypotheses in cases 267 – Rating of cases 262, 263, 537 – Selecting cases 263 Cause (of trouble) see also Hypothesis and related terms Cues and Symptoms – actual cause in case 264, 591 – relating cause to magnitude of the extent of symptoms 43 – relating to when: startup of new process, startup after shutdown, usual operations 44 Cause-effect information, see also specific equipment for details – activity to check consistency 213 – activity to improve skill 213 – consistency between 213 – use of 195 Cellular PE, extrusion 120 Centrifuges, filtering, symptoms-cause 87 – brainstorming 187 – cases involving 186 Centrifuges, sedimentation, symptomscause 87 – causes 576 Change strategy for TS: – contrast with basics strategy 23 – description 23 – example questions to ask 23 – when to use 4, 23, 197

I 3

I 4

Index Check, see Look back stage in generic strategy Check and double check, see Accuracy Chin, T.G. 409, 534 CI, Chute index – cited 56, 57 – defined 108 Classifying information 219 related term Analysis – activities to improve skill 219–220 – as first step in reasoning 224 – example 224 – definition 219 – how to 219, 399 – need for 20, 21, 195, 219, 399 – of ideas from brainstorming 220 – of starting information 219 – of triggering events 220, 399 Coagulation, symptoms-cause 111 Coalescers, fundamentals 110 – symptoms-cause 110–111 Coating, symptoms-cause 126 Commissioning 266 – diagnostic actions 586 Communication: – activities to improve skill 237–238 – criteria for effective 237 – definition 237 – preference to interpret 249 – self test of skill 8 Compressors: – cases involving 1, 291, 304, 317, 350, 391 – centrifugal, symptoms-cause 52 – frequency of faults 44 – MTBF 44 – reciprocating, symptoms-cause 52 Conclusion (inference, proof or disproof of hypothesis): – example 228 – reasoning process to test validity 223–232 – write down 224 Condensers, shell and tube, symptoms-cause 62 – cases involving 251, 252, 255–256, 312, 317, 321, 323, 370, 374 – causes 575 Consistency (among evidence): – activities to improve skill 209–228, 227, 399 – diagnostic actions to check 267–268, 588–589 – examples 199, 227 – how to check with evidence 227 – importance of, for reasoning 209

– need for 195, 209 – to assess reasoning 230 – types of – cause-effects 212, 399 – evidence 227, 399 – experience with equipment 228, 399 – rules of Mathematics and English 217, 399 – words/definitions 210–212, 399 Contamination (impurities) and crystallization 72 Context – definition 224 – example 224 – how to identify 225 – of reasoning 224 Control, process: – activities to improve skill 418, 421 – four elements of 3 – self test about knowledge of 8 – symptoms-cause 47 Control room: need to visit 198, 267, 589 Controller, process, causes 574 Conveying, solids, see also Pneumatic conveying, and feeders – cases involving 182, 186, 222, 333 – symptoms-cause 34 Corrosion: – general reasons why corrosion becomes a cause 45 – symptom-case data 45 – types of corrosion and their frequency 45 Costs, see also Financial penalty of diagnostic action 264 and 582 Covey, S., cited 22, 406 Cox chart, use 418 Cracking, catalytic, reactor, symptoms-cause 101 Creativity, and Brainstorming and Hypothesis generation: – activity to improve your skill 190 – and skill in classification of ideas 219 – case ’7 171 – checklist of triggers 183–186 – example Case ’10 187–190 – feedback form 191 – need for 20,183 – target skills 183 – to identify root cause, example on TS Worksheet for Case ’8 38 Criteria: – and goals 401, 193 – need for must and want 22

Index Critical thinking, see also Reasoning, and Diagnostic actions: – activities to improve skill in consistency 209, 399 – list of subskills 195 – need for 20 – reflections about 400 Crystallization, solution: fundamentals 67 – cases involving 13–15, 157ff, 186 – symptoms-cause 72 – vacuum and circulating system, symptomscause 72 Cues (for storing experience in Long Term memory) 20 Cues/evidence (for trouble) see also Diagnostic actions – how to critique of evidence 227, 231: – check for consistency 227 – diagram 228–229 – test pertinence of 228 – number of cues used per case 34 – suggestions for weighting 25 – suggestions to prevent overlooking 25 Cycling, related term Surging: – batch cycles 221, 222 – symptoms-cause 130, 63–65, 71, 101 Cyclone, GS, symptoms-cause 81–82 – cases involving 333

d Data handling, see also Diagnostic actions: – activities to improve skill 399 – and cycling 222 – elements of the TS process 33, 398 – errors in data 208 – interpreting data 208, 249–250 – list of detracting and enriching behaviors 30 – options for gathering data 196ff – personal bias in 204–208 – reflections about your approach 400 – self test of skill 5 Decanters – cases involving 179 – symptoms-cause 82–84 Decision making: – elements of the TS process 22, 33 – list of detracting and enriching behaviors 32 – self test of skill 6 Define: second stage in a generic problem solving process: – and skill in classification 219

– – – –

context, how to identify 225 critique use of 163 description 20, 21, 28 example use in TS Worksheet for Case ’8 36 – for Case ’5 141 – for Case ’6 145 – prompts for use in the TS Worksheet 26, 40 Dehydration, glycol system, symptoms-cause 73, 74, 75, 76, 129 Dehydrogenation, safety 96, 98, 99 Demisters – fundamentals 110 – symptoms-cause 110 Depropanizer: – Case ’8 description 35 – cases involving 35, 213, 327, 345, 362, 367, 377 – P&ID, 34 – TS Worksheet for 35 Design fault, occurrence 44 Design data fault, occurrence 44 Desorption, gas, fundamentals 67 – cases involving 181, 383, 387 – symptoms-cause 75 Diagnostic actions – activity to improve skill 201, 202 – and cycling 222 – chart relating to hypotheses and evidence on TS Worksheet, see Hypothesis, chart – consistency tests 267–268, 588 – costs incurred 264, 582 – criteria for selecting 196, 264 – equipment to help 203–204 – errors in evidence 208 – examples for TS Case ’8 39, 201 – how to select 24, 196, 264, 587 – interpreting data 208 – options 196: – recent events 197, 265, 583 – safety 196, 264, 583 – to correct 200, 264, 268 – to help select TS strategy 197 – to understand what’s going on 198, 267, 585, 587 – to test hypotheses 198, 267, 587 – personal style in selecting 204 – reflection about 400 – relating to hypotheses 24 – site visit 589 – time required for selected 204 – trends and patterns 590

I 5

I 6

Index – using senses 589 Differences in concentration for mass transfer 67 – exchange equilibrium for separations 67 – molecular geometry for separations 67 – partition coefficient for separations 67 – pressure for fluid flow 51 – solubility for separations 67 – temperature for heat transfer 58 – vapor pressure for separations 67 Distillation column – activities to improve skill 418 – cases involving 170, 10, 35, 211, 222, 247, 251, 252, 256, 299, 321, 327, 345, 354, 362, 367, 372, 377, 383, 387 – causes 576 – frequency of failures of 44, 201 – good practice 69 – self test of knowledge about 6 – symptoms-cause 69–72 Do it, fifth stage in a generic problem solving process, – description 20, 21, 28 – example on TS Worksheet for Case ’8 39 Doig, L 403, 408 Drives, see Engines, Motors and Turbines – for extruders 120, 121 Dryers, symptoms-cause 85 – cases involving 186, 347 Dudzic, Mike, cited 286, 543 Dutta, S., cited 101

e Edgar, M.D. 9 Ejectors, steam: – cases involving 13, 181, 335 – symptoms-cause 53–54 Elaboration, as part of the Look back stage 20 – importance of in problem solving 17 Electrical failure, occurrence 44 Elonka, S.M. 403, 408 Elstein A.S. 24, 405 Emulsion, formation of stable, cause 129 Energy, conservation of 58 Energy, mechanical, fundamentals 58 Energy, thermal, see Heat exchangers, and furnaces Energy exchange, fundamentals 58 Engage, first stage in a generic problem solving process – critique use of 163 – description 20, 21, 28 – example use in the TS Worksheet for

– Case ’5 141 – Case ’6 145 – Case ’8 35 – prompts for use in the TS Worksheet 26, 40 Engines, symptoms-cause 58 Entrainment (GL), symptoms-cause 74, 84 Entrainment (LL), symptoms-cause 74, 85 Environment in which you TS: – impact on your approach 253 – self test 254–255 Environmental impact of process: self test about knowledge of 7 Equilibrium, phase, use of 67 Esso Chemicals, cited 251, 321, 541, 548 Esterification, safety 96, 98, 99 Ethylene 247, 299 Evacuation 196, 265 – as diagnostic action 265 Evaporators, general, fundamentals 67 – cases involving 257, 335 – forced circulation, symptoms-cause 68 – multiple effect, symptoms-cause 68–69 – symptoms-cause 67 – vapor recompression, symptoms-cause 68 – vertical falling film, symptoms-cause 65, 68 Evidence, see Cues Exercise solving: – contrast with problem solving 17 – definition 17 – diagram of 19 – example use, Case ’7 157 – frequency of, for experienced trouble shooters 17 Experience, past, and use in problem solving, trouble shooting 17 Experience with process equipment, see specific equipment for details – and consistency 218 – keeping up-to-date 402–403 – need for and importance of 4, 397, 402 – reflections about your data base 398 – self test of 6, 411 – stuff left in lines 433, 584 Explore, the third stage in a generic problem solving process, – and skill in classification of information 219 – critique use of 163 – description 20, 21, 28 – example for Case ’4 139

Index – example use in TS Worksheet for Case ’8 36 – example use of Why? Why? Why? 180 – prompts for use in the TS Worksheet 26, 41 Extruder, symptoms-caus see also Reactive extrusion 120–125 – blown film, symptoms-cause 121, 122, 124 – cast, symptoms-cause 122 – coating, symptoms-cause 122 – coating wire and cable, good practice 120 – filament, symptoms-cause 122, 124 – pipe, symptoms-cause 122 – sheet, symptoms-cause 122 – single screw, good practice 120 – symptoms-cause 123 – twin screw, good practice 120 – vented twin screw, good practice 120

f Fabrication fault, occurrence 44 Facts, related to Opinions and Opinionated facts: – activities to improve skill 211, 250–253 – definition 210 – examples 211 – facts versus opinions 208 – opinionated facts – definition 210 – examples 211 – opinions – definition 210 – examples 211 – preference to interpret 249 Fans, symptoms-cause 52 – cases involving 308, 312, 321 Farrell, R. J., cited 256 Faults, see Causes FCCU, fluid catalytic cracking unit, see Cracking, catalytic FDI, Feed density index – cited 56, 57, 58 – defined 108 Feedback: – self tests: – experience with process equipment 411 – of critical thinking skills 233 – of PS skill 194 – of trust 242 – personal style, confirmation bias 206 – personal style in TS 206 – skill in five elements of TS 5–8

– forms: – brainstorming 191 – criteria identification 193 – example data 177 – goal setting 193 – listener 180 – PS strategy 178 – reflections about critical thinking skills 233 – reflections about PS skills 194 – reflections for TAPPS 172 – self assessment 193 – talker-listener in TAPPS 173 – teamwork 50–51 – trust 242 – TS environment 254–255 Feeder, belt, symptoms-cause 57 – cases involving 186, 347 – from bottom of hopper, symptoms-cause 57 – screw conveyor, symptoms-cause 57 – solids volumetric, symptoms-cause 57 Filter (LS), symptoms-cause 88 – cases involving 286, 342 Financial penalties 1, 4, 266 – cases involving 321 Fisher, K., cited 48, 406 Flash drum, three phase separator (GLL), symptoms-cause 84–85 Flexibility, need for 22 – as part of the synthesis elements of TS 21, 22 Flocculation, symptoms-cause 111 – case involving 289 Fluid dynamics: – faults, occurrence 44 – fundamentals 51 – types of faults 51–52 Fluoroplastics, extrusion 120 Foaming, symptoms-cause 62, 67, 69, 71, 74, 80, 93–98, 110, 128 – reboiler selection 63 Fouling, symptoms-causes 63, 64, 67, 68, 69, 128 Fox, Don F., cited 342,-551 Francis, D., cited 48, 406 Freon, foaming 129 FRI, Flow ratio index – cited 56 – defined 108 Fundamentals: – and consistency 218 – energy exchange 58

I 7

I 8

Index – example in Case ’4 139 – fluid flow 51 – separations of homogeneous phases 67 – use of, for the basics strategy for TS 23 Furnaces: – cases involving 171, 256–257, 304, 308, 374 – fundamentals 58 – symptoms-cause 60

g Gans, M., cited 43, 133, 405–408 Gates, John, cited 247, 299, 540, 545 Garbage left in process, see Experience with process equipment Gas breakup, symptoms-cause 110 Gauly, R., cited 91, 101 Geitner, F. K., cited 403, 408 Goals, express as results 22 – need for 401 – setting goals for improvement 42, 402 Goyal, O.P., cited 411 Grit chamber, symptoms-cause 86

h Halpern, D., cited 229, 407 Handbook data 266 Hazard see Health and safety HAZOP, use of 3 HDPE, high density polyethylene, extrusion 120 Health and safety: – and Case ’4 140 – combinations of chemicals 131–132 – dust explosions, example data 131 – flammability, example data 131 – health, example data 130 – impact on diagnostic action 3, 196, 264 – importance in TS 2 – need for knowledge about 5 – reactions 131 – self test about knowledge of 6–7 – shutdown to avoid hazard 3, 264 – stability, example data 131 – symptom-cause 130–132 – symptoms of hazard in STR reactors 96 Heat exchangers: – cases involving 10, 12, 171, 247, 299, 304, 308, 314, 335, 339, 350, 357, 364, 391 – frequency of faults 44, 201 – fundamentals 58 – types: – air cooled, symptoms-cause 55

– trim coolers 65 – causes 575 – plate, symptoms-cause 65 – shell and tube: – causes 575 – fundamentals 58 – good practice 61 – self test experience 6 – symptoms-cause 61 – spiral plate, symptoms-cause 65 Heat of reaction, and hazards 131 Heat transfer fluids, high temperature, symptoms-cause 66–67 – cases involving 252 HI, Hopper index – cited 56, 57 – defined 108 Holmes-Rahe scale for stress 22, 406 Hoppers, see Bins Hydrocyclone, LL, symptoms-cause 84 Hydrocyclone, LS, symptoms-cause 86 – cases involving 186 Hydrolysis, safety 96, 98 Hydrotreating, reactors, symptoms-cause 95 Hypothesis, – and cues 25 – and reasoning process to validate 223 – and skill in classification 219 – chart relating to evidence and diagnostic actions 24 – Case ’3 426 – Case ’4 429 – Case ’5 430 – Case ’6 148, 202–203 – Case ’7 158, 159, 171 – Case ’10 171 – Case ’11 216 – example for TS Case ’8 39 – example on TS Worksheet 42 – check with colleagues 587 – feedback for all cases 537ff – feedback for some cases 267 – generate early 24 – generating multiple, as part of the synthesis element of the TS process 21, 22 – need to generate 195 – number of active 24 – self rate of skill 5

i “I want to and I can”, use of 20 – Case ’5 141

Index – example use in TS worksheet for Case ’8 36 Ice formation: in steam ejectors 53 “If...then...” logical statements, characteristics of 214 Impact sensitivity 131 Indicator, see Instruments Inference, see Conclusion Injection molding, symptoms-cause 112–119 – cases involving 347, 380 Instruments, sensors, indicators, recorders – causes 574 – frequency of failure 201 – simple consistency tests 267–268, 588 – symptom-cause data 46 Insurance, and data 266, 586 – case involving 321 Interpersonal skills, see People skills Ion exchange IX: – cases involving 256 – fundamentals 67 – symptoms-cause 77–78 IS and IS NOT: – as diagnostic action 265, 584 – example use in TS worksheet – Case ’3 135, 136 – Case ’4 139 – Case ’5 141 – Case ’6 145 – Case ’7 158 – Case ’8 36 – use in strategy 23, 197

j Johanson, J. R. – cited 56, 108 – description 20, 21, 29 Johnson, D. A., cited 135 Johnson, D. W., cited 210 Johnson, D. W., and Johnson, F. P., cited 243 Johnson-Laird, P. N., cited 206, 405 Jungian typology, (MBTI) and style 204, 234, 243, 399

k Kepner-Tregoe 23, 406 King, C. J., cited 12,323 Kirton, M. J., cited 243, 406 Kister, H., cited 135, 403, 405, 406, 408 Kletz, T., cited 245, 253, 407–409 Knock out pots, symptoms-cause 80 – cases involving 247, 291, 299, 350

Knowledge about process equipment, see Experience with process equipment Koros, R. M., cited 95 Krishnaswamy, R., cited 173, 407, 538

l Lapp, S. A., cited 245, 407, 409 Latex crumb, flocculation, symptoms-cause 111 Lieberman, N. P., cited 101, 133, 403, 405–409 Limitations, personal, need to learn and identify 22 Listening skills: activities to improve skill 239, 250–252 – characteristics of 238 – feedback via TAPPS 173 – need for skill 249 – self rate 8 – self rate the environment 255 – the SIER model 238 – three skills 238 – activities to improve skill 239 – attending, characteristics of 238 – how to 239 – reflecting, characteristics of 239 – tracking, characteristics of 238 LLDPE, extrusion 120 LM ID, log mean temperature difference 58, 61 Look back description 20, 21, 29 Luckenbach, E. C., cited 101 Lynn, Scott, cited 294, 330, 370, 544, 559, 550

m Maintenance, as diagnostic action 265, 584, 585 Marangoni effects, causes 76, 83 Marlin, T. E., cited 35, 327, 309, 345, 350, 354, 357, 362, 367, 377, 537, 546, 549,552, 553, 555, 556, 558, 560 Mass balance 266 Material failure fault, occurrence 44 MBTI, see Jungian typology McNally Institute, cited 403 Mean time between failure, MTBF, use of data 44 Mechanical failure, occurrence 44 Membranes – fundamentals 67 – symptoms-cause 79

I 9

I 10

Index Mental representation of the problem: – as part of the Explore stage 21 – importance in problem solving 17 Methyl ethyl ketone, and foaming 129 Microfiltration, see Membranes Mixers: mechanical agitators, L. Mixing, solids, see Blending – cases involving 181, 297 – frequency of faults 44 – symptoms-cause 107–108 MM, molar mass 103, 124 Monitoring: – example application in TS Worksheet for – Case ’5 141, 142 – Case ’6 145 – Case ’8 36 – example data of 177 – importance of in problem solving 20, 260 Motors, electric: – MTBF 44 – symptoms-cause 59 MPI, main plant items 7 MSDS, rating 130, 263, 264, 583 MTBF see Mean Time between failures

n Nano filtration, see Membranes Neutralization, safety 96, 98, 99 NFPA, National Fire Protection Agency, rating 130, 264, 583 NIPR, net inlet pressure required 54 Nitration, safety 96, 98, 99 Nominal group 49 NPSH, net positive suction head 54, 64, 201 – activities to improve skill 420 – and Case ’5 141 – definition 420

o On-going process, typical causes for 44 Operators see also People – example interaction with, – by Ahmed 239 – by David 143 – by Frank 158 – by Jose 238 – importance of talking to 198, 591, 587 – on other plants 268, 588 Opinionated facts, see Facts Opinions, see Facts Oxidation, safety 96, 98, 99

p P&ID (process and instrumentation diagram): – as diagnostic action 268, 589 – for depropanizer 34 – for ethylene 300 – information from 266 – symbols on 415 Pareto’s principle, use of 22 Parker, N. H., cited 186, 407, 538 Pattern recognition: – look for as diagnostic action 590 – need for skill in 195 – types of patterns: – in cues/evidence 223 – in symptoms 221 Pearson, Doug, cited 584 Pelleting, strand type, symptoms-cause 111–112 – water ring, symptoms-cause 112 Penalties 266, 586 – cases involving 321 People: – activities to improve skill 240 – destroyers (four) of relationships 240 – fundamental interpersonal RIGHTS 239 – definition 239–240 – list of nine skills needed 47 – performance: – factors affecting 244–252 – impact of alienation 249 – impact of “I know best” 249 – impact of listening skills, see Listening – impact of motivation 249 – impact of personal style, see Personal style – impact of pride 244–245 – impact of stress on 22, 221, 245–246 – activities to improve skill 246–247 – self rate environment 254 – impact of the environment on 253–255 – self rate 254–255 – impact of unwillingness to admit error 244–245 – self rate environment 254 – impact of willingness to risk, self rate environment 254 – types of errors made 244 People problems, see Problems involving people performance review/assessment 249, see also Self assessment problem involving 2 – five elements of 8 – reflections about 400 – self rate 8

Index – skill with, need for 2 – trust, see Trust Personal style: – bias 25, 205–206 – effect on selecting actions 200, 204–208 – identify via Jungian typology (MBT1) 204–205, 243 – Saadia’s dominant P 156 – Frank’s dominant J 161 – example data 243 – impact on selection of diagnostic actions 204–208 – via Johnson and Johnson inventory 243 – via Kirton inventory 243 – need to identify 22 – preference to infer 249ff – self rate 8 – self tests about bias 206 – skill in listening 238 – tendency to interpret 249–250 – activities to improve skill 250–252 – feedback 253 – self rate environment 255 PET, extrusion 120 pH, acidity, as cause of corrosion 45, 46 – as cause of stability 69, 71, 72, 74, 76, 78, 79, 82, 83, 87, 88 p-H, pressure enthalpy diagram, use for TS refrigeration 65 Pinch analysis, use of 61 Pipes, causes 574 Pistorius, J. T., cited 95 Plan, fourth stage in a generic problem solving process – description 20, 21, 28 – example on the TS Worksheet for Case ’8 39 Platforming, reactors, symptoms-cause 91, 93 – cases involving 10 Pneumatic conveying, symptoms-cause 56–57 – cases involving 182 Polymerization: – cases involving 178, 297, 370 – reactors, STR, symptoms-cause 98 – safety 96, 98, 99 Powers, G., cited 245, 409 Pressure profile: – activities to improve skill 419 – self rate skill 7 – use of 266

Problems involving people, impact of rules and regulations 22 Problems, example “problem” in TS Worksheet – Case ’3 137 – Case ’4 139 – Case ’5 141 – Case ’6 145 – Case ’8 36 – need to define the real problem 20, 21, 181 – versus exercises 17 Problems, TS see also Cases – as related to different types of equipment 44 – four common characteristics of 2 – high temperature 44 – types: change versus on-going process 23 – for on-going process 3, 4, 44 – posing safety and health hazards 3, 264 – when startup after maintenance shutdown 4, 44, 265 – usual causes 4, 265 – when startup new process 3, 44, 265 – usual causes 3 Problem solving, generic See also Trouble Shooting, mental process; See also subset skills of Analysis, Awareness, Creativity, Critical thinking, Strategies for problem solving – activities to improve skill 399 – and type of TS problem: – as related to different types of equipment 44 – high temperature 44 – on-going process 44 – startup after shutdown 44 – startup of new process 44 – contrast with exercise solving 17 – definition and description of mental process 17 – diagram of process 18 – list of characteristics 19, 20 – list of detracting and enriching behaviors 29 – reflections about your skill 399 – self test 194 – stages in a strategy 20, 21 – strategy for general problem solving 20, 21 Process and Instrumentation diagram, see P&ID Process control, see Control, process

I 11

I 12

Index Process equipment, knowledge about, see Experience with process equipment Pumps, centrifugal: – activities to improve skill 420 – cases involving 11, 35, 186, 214, 251, 255–256, 286, 289, 294, 304, 308, 321, 330, 336, 342, 350, 357, 370, 383, 387 – causes 573–574 – fundamentals 51, 420 – frequency of faults 44, 201 – MTBF 44 – self rate experience 6 – symptoms-cause 54, 579 Pumps, dry vacuum, symptoms-cause 53 – gear, symptoms-cause 55 – liquid piston vacuum, symptoms-cause 53 – mono, symptoms-cause 56 – reciprocating, symptoms-cause 54 – rotary, symptoms-cause 54 – rotary screw, symptoms-cause 56 PVC, polyvinyl chloride, extrusion 120

q Questions to ask, see Diagnostic actions

r Rag at interface for SX or decanters – causes of 76 RAS, Rough wall angle slide – cited 56, 57 – defined 108 Reactors – cases involving 1, 10, 171, 178, 182, 297, 304, 308, 317, 330, 339, 357, 374, 391 – frequency of faults, 44, 201 – introduction 88–89 – safety, 96–99 – types – CSTR, mechanical mixer, symptomscause 99–100 – CSTR-PFTR with recycle, symptomscause 106 – PFTR, bubble reactors, tray columns, symptoms-cause 94 – PFTR, fixed bed, adiabatic, symptomscause 91–93 – PFTR, thin film, symptoms-cause 96 – PFTR, trickle bed, symptoms-cause 94–96 – PFTR multitube fixed bed, non adiabatic, symptoms-cause 89–91

– Reactive extrusion, symptoms-cause 106–107, 123, 125 – STR, batch, symptoms-cause 96–98 – STR, fluidized bed, symptoms-cause 101–106 – STR, semibatch, agitated bubble, symptoms-cause 98 – STR, semibatch, bioreactor, symptomscause 98 – STR, semibatch, symptoms-cause 98 – processes: – alkylation 83, 99 – platforming 91, 93 – polymerization 178 – symptom cause 98 Reasoning related term Critical thinking – activities to improve skill 223–232 – common biases in 205 – nine-step process 223–232, 299 Reboilers, general, fundamentals 63 – cases involving 10, 35, 323, 370 – causes 575 – symptoms-cause 63 – types: forced circulation, symptoms-cause 64 – kettle: – fundamentals 63 – symptoms-cause 63–64 – thermosyphon: – fundamentals 63 – symptoms-cause 64 – when used 64 Reflections – activity 5, 172, 194, 233, 251, 262 – need for 401, 403 Reforming, symptoms-cause 89, 91 – cases involving 171, 339 Refrigeration – symptoms-cause 65–66 – cases involving 202, 247, 299, 323, 350 Reverse osmosis, RO, see Membranes RI, Ratholing index – cited 56, 57 – defined 108 Riance, X. P., cited 403, 409 Rights, personal: – list of 47, 239–240 – self rate 8 Risk, willingness to 21, 22 Rogers, R., cited 297, 407, 544 Root cause, see Symptom Rotating equipment, frequency of faults 44 – MTBF 44

Index RTD, residence time distribution 107, 124, 125 Rules of thumb: – importance of 218 – listing of by process equipment name, Chapter 3, 43ff

s Safe park, use of 1, 3, 134, 141, 195, 197, 265 Safety, see Health and safety Safety interlock shutdown, SIS 265 Saletan, D., cited 403, 405, 408 Samples, collecting 200 SBI, Spring back index – cited 56, 57 – defined 108 SCAMPER, trigger for creativity 184 Screens, dewatering – cases involving 186, 286 – for Liquid Solid separation, symptomscause 85–86 – for Solid Solid separation, symptomscause 88 Scriven, M., cited 209, 229, 407 Self assessment, related topics Feedback – activity to develop skill 192 – feedback form 193 – how to develop skill in 192, 401 – need for 403 – of bias 206 – of creativity 191 – of critical thinking skills 233 – of experience with process equipment 411 – of five component skills in TS 5–8 – of personal style 206 – of personal style, confirmation bias 206 – of PS skills 194 – of TAPPS 172 – of trust 242 – target skills 192 – using to improve 401 Sensors, see Instruments Separation of species in heterogeneous phases, introduction to 79 – process equipment, symptoms-cause 80ff Separation of species in homogeneous phase, fundamentals 67 – process equipment, symptoms-cause 67ff Separator, gas-liquid, cases involving 304, 317, 321, 326, 350, 370, 391 Shaw, I. D., cited 570 Short term Memory, STM 25

Shut down (the process), and safety, see also Health and safety 1 – troubles after, for maintenance 4, 44, 265 Silveston, P. L., cited 314 SIS, safety interlock shutdown 3, 265 Size enlargement, fundamentals 110 – process equipment for 110ff SMARTS: example use in TS Worksheet for – Case ’3 136, 423 – Case ’4 428 – Case ’5 141, 431 – Case ’8 37 Solids conveying, see Conveying, solids Solvent extraction, SX, fundamentals 67 – and coalescers, symptoms-cause 110 – SX, column extractors, symptoms-cause 76 – SX, centrifugal, symptoms-cause 77 – symptoms-cause 76 Sonic velocity, and compressors 52 Sour water scrubbers, SWS, causes of corrosion of 46 – symptoms-cause 75, 76 So What?, use 232 Startup, of new process: – as diagnostic actions 265 – typical causes of trouble 44, 265 – use of basic strategy for 23 Startup after shutdown: – as diagnostic actions 265 – typical causes of trouble 44, 265 – use of change strategy for 23 Steam, usage 198 Steam systems, causes of corrosion in 36 – cases involving steam generation 339 – good practice 58 – symptoms-cause 66 Steam traps, symptoms-cause 80 – causes 575 Stiction 47 STM, see Short Term Memory Strategies for problem solving, – activities to improve skill 173 – description 20, 21 Strategies for TS, related entries Basics and Change – selecting basics versus change 197 Stress: – activities to improve skill 246–247 – effect of high distress on propensity to make mistakes 22, 221, 245–246 – need for some stress 22 – need to manage 20, 22

I 13

I 14

Index – self rate environment 254 – suggestions on how to manage 22 Stripping, see Desorption, gas Style, preferred, see Personal style Subproblems, identify 20 Sulfinol 383, 387 Sulfolane, and foaming 129 Sulfonation, safety 96, 98, 99 Sulfuric acid systems, causes of corrosion in 46 Surfactants (causing foaming), examples 74 – causing emulsion stability, examples 84 Surface phenomena, importance 67 Surface tension: – and absorption 73 – and coalescers 111 – and wetting 94 – critical, use in absorption 73 – impact 69, 71, 73, 83, 110, 129 – in packed columns 109 – in reactors 94 – negative, definition 69 – positive, definition 69 Symptoms, (one type of cue): – characteristics of 2 – definition 212 – how to identify symptom 220 – root cause versus symptom 24, 43, 195 – activities to improve skill in identifying root cause 217, 225 – examples, Case ’8 38, 216 Symptom-cause data, see specific pieces of equipment – diagraming 228, 229 – symptom-cause data for – Case ’3 228–229 – Case ’8 38 – symptom-hypotheses-diagnostic action chart, see Hypothesis, chart symptoms, list of for – Case ’6 147 – Case ’8 39 Synthesis – element in the TS process 21, 22, 33 – list of detracting and enriching behaviors 31 – self rate skill 5 Systems thinking: – activities to improve skill 418 – definition 5, 7, 127 – elements in 7 – need for 2 – self rate skill 7

t Tabletting, symptoms-cause 111 TAPPS (talk aloud pairs problem solving) – use of, to improve awareness 165–168 – use of, to improve the application of strategies 174–177 Tanks, frequency of faults 44 Taylor, W., cited 13, 214, 255, 317, 339, 357, 383, 387, 391, 540 Teamwork: – feedback form for meetings 50–51 – symptoms-cause 48–49 Temperature, high, fiult in systems operating at 44 Tests, selecting and designing, see Diagnostic actions Thermodynamics 267 Thickener, symptoms-cause 86–87 – cases involving 286, 342 Three phase separators, GLL, symptomscause 84 Time management: – how to 22 – need for good 22 Transmitters, symptom-cause data 47 Trends in evidence, see Patterns Trim coolers, use of 61 Trouble shooting, mental process: – activities to improve skill, triad 259–262 – as a problem solving process 17ff – assessment of skill 33, 262 – classification of components: – data and analysis elements 19, 398 – Case ’4 140 – Case ’5 144 – Case ’6 156 – Case ’7 161–162 – example rating form 33 – example critique for Case ’3 138 – list of detracting and enriching behaviors 30 – decision making elements 19, 22, 398 – Case ’4 140 – Case ’5 144 – Case ’6 156 – Case ’7 161–162 – example rating form 33 – example critique for Case ’3 138 – list of detracting and enriching behaviors 32 – problem solving elements 19, 20, 21, 398 – Case ’4 140

Index Case ’5 144 Case ’6 156 Case ’7 161–162 example rating form 33 example critique for Case ’3 137 list of detracting and enriching behaviors 29 – synthesis elements 19, 22, 398 – Case ’4 140 – Case ’5 144 – Case ’6 156 – Case ’7 161–162 – example rating form 33 – example critique for Case ’3 138 – list of detracting and enriching behaviors 31 – definition 1, 2 – five key skills: – experience with process equipment 4, 398 – self test 5 – people skills 5, 397, 400 – self test 8 – problem solving 4, 397, 398 – self test 5 – process safety and properties of materials 5, 397 – self test 6 – systems thinking 5, 397 – self test 7 – individual 262 – possible immediate corrective options 1 – reflection about activities 401 – reflection about skill 398 – self rate five strengths 402 – strategies, see Basics strategy for TS and Change strategy for TS – worksheet, description 27–29, 587 – succinct version 26–27 – detailed version 40–42 – examples – Case ’3 423 – Case ’4 427 – Case ’5 141, 430 – Case ’6 145 – Case ’8 35–39 Trouble shooting problems, see Problems, TS Trust, building and developing: – activities to improve skill 241, 242 – components of 240 – definition 242 – need for 47, 240 – self rate skill 9, 242 – – – – – –

Turbine, MTBF 44, 201 – cases involving 35, 171, 255, 339, 350 – steam, symptoms-cause 59 Turnaround, as diagnostic action 265, 584, 585 – trouble after 4, 44 Turner, J., cited 80 Tyler, Ted, cited 222 Types of trouble shooting problems, see Problems, TS

u Ultrafiltration, see Membranes Uniqueness, see Personal style

v Vacuum, equipment to produce, symptomscause 53–54 – and crystallizers, symptoms-cause 72 – and evaporators, symptoms-cause 68 – cases involving 13, 181, 252, 255–256, 335, 374, 383, 387 Valid conclusions, how to reach, see Reasoning Valves, causes 574 – block, symptom-cause data 47 – check, symptom-cause data 47 – control, symptom-cause data 47 – rotary star, symptoms-cause 56 Varadi, T., cited 94 Vendors, vendor files, use of 197, 266 – call 268 Vessels, pressure, frequency of faults 44

w Wasan, W.C., cited 405 Water, makeup contains corrosion, products, causes 46 Water treatment, flocculation, symptomscause 111 Weather, importance of 197, 584 – as diagnostic action 265, 584 – example impact 221 Wetting – and absorbers, symptoms-cause 73 – and decanters, symptoms-cause 83 – and demisters 110 – and packed columns 109 – and reactors 94 – and SX, symptoms-cause 76 What if? – example use in TS Worksheet for – Case ’4 429

I 15

I 16

Index – Case ’5 142, 432 – Case ’8 37 – to create counterarguments 232 – trigger for brainstorming 185 – when to use 198 Why? Why? Why? – activity to improve skill 181 – as diagnostic action 265 – example use on TS Worksheet – Case ’4 139 – Case ’7 180 – Case ’8 36 – prompt to use on TS Worksheet 26, 41 – use of during the explore stage of problem solving 20

Winter, D. R., cited 347, 380, 552, 561 Woods, D. R., cited 130 Worksheet, see Trouble shooting, worksheet

y Yip, Jonathan, cited 256, 289, 543 Yokell, S., cited 546, 409 Young, D., cited 48, 406

z zpc, zero point of charge (isoelectric point), and stability 69, 71, 74, 76, 78–80, 83, 93, 111