2004 New and Renewable
Energy Technologies for Sustainable Development
This page intentionally left blank
2004 New and Renewable
Energy Technologies for Sustainable Development 28 June - 1 July 2004
Evora, Portugal
Editsrs
Maria da Graca Carvalho Naim Hamdia Afgan Institute superior Tecnico Portugal
world scientific tific NEW JERSEY * L O N D O N * S I N G A P O R E
BElJlNG * S H A N G H A I * HONG K O N G * TAIPEI
*
CIiEMNAI
Published by World Scientific Publishing Co. Pte. Ltd. 5 Toh Tuck Link, Singapore 596224 USA office; 27 Warren Street, Suite 401-402, Hackensack, NJ 07601
U K ofice: 57 Shelton Street, Covent Garden, London WC2H 9HE
British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.
2004 NEW AND RENEWABLE ENERGY TECHNOLOGIES FOR SUSTAINABLE DEVELOPMENT Copyright 0 2007 by World Scientific Publishing Co. Re. Ltd All rights reserved. This book, orparts thereof. may not be reproduced in anyform or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.
ISBN-13 978-981-270-505-1 ISBN40 98 1-270-505-8
Printed in Singapore by World Scientific Printers (S) Pte Ltd
CONTENTS Foreword ....................................................................................
ix
Chapter 1 - Renewable Energy Sources Promoting Renewable Energies and Energy Efficiency through the CDM Funding Options K.D. Patlitzianas, A . Flamos, H. Doukas, A.G. Kagiannas and J. Psarras ...........
1
H2RES,Energy Planning Tool for Increasing the Penetration of Renewable Energy Sources in Island Energy Supply M. Lerer, N. Duic, L.M. Alves and M. G. Cawalho.. .....................................
15
The Shape of Complete Renewable Energy Systems in the World H. Yamamoto and K. Yamaji ...............................................................
31
Experimental Investigation and Modelling of Thermoelectric Generators for Use in Hydrogen Based Energy Systems J. Eriksen, R. Glocknev, V.A. Yartys, B. T. Hafsteinsson and T.I. Sidusson ...........
41
Intelligent Operation Management of Fuel Cells and Micro-Turbines Using Genetic Algorithms and Neural Networks A. M Azmy and I. Erlich ...................................................................
53
Chapter 2 - Biomass Energy Municipal Solid Waste Valorisation as Energy for Isolated Communities G. Tavares, Z. Zsigraiova, V. Semiao and M G . Cawalho ..............................
67
Production of Sustainable Hydrogen Using Thermochemical Gasification of Biomass J. Andries, W. de Jong and H. Spliethoff.. ................................................
77
Characterization of Kenaf Potential in Portugal as an Industrial and Energy Feedstock A.L. Fernando, M P . Duarte, J. Morais, A . Catroga, G. Serras, B. Mendes andJ.F.S. Oliveira.. ........................................................................
87
Implementation of Anaerobic Digestion Plants in Breweries - Difficulties and Benefits G. Pesta and R. Meyer-PittrofJ:............................................................
95
V
vi
Contents
Application of Biofuels to Compression Ignition Engines A. Kowalewicz ...............................................................................
105
Kinetic Study of Froth Flotation for PET-PVC Separation E. Agante, T. Cawalho, F. Durao, A. Pinto and T. Mariano.. ..........................
117
Use of Biomass in Small Direct Fired Systems C. Syred, W. Fick, N. Syred andA.J. GrifJhs ............................................
127
Emissions Reduction by Co-Firing Biomass or Waste with Coal in a Pressurized Fluidised Bed Combustion Combined Cycle Power Plant Y. Huang, J. T. McMullan, D. McIlveen- Wright and S. McCahey .......................
153
Production of Biocoal from Cassava Stalk T. Puthikitakawiwong, R. Boonsu and 0.Joompha ......................................
165
A Comparison of Power Generation from Biomass in a Small CFBC Plant with Biomass Co-fired with Coal in a Large CFBC D.R. McIlveen- Wright, B. C. Williams and J. T. McMullan. ..............................
171
Characterization of Sweet, Fibre and Biomass Sorghum Potential in Portugal as an Industrial and Energy Feedstock A.L. Fernando, M.P. Duarte, J. Morais and J.F.S. Oliveira.. ...........................
183
Chapter 3 - Solar, Wind, Wave and Geothermal Energy Process Dynamics of Fossil Steam Power Plants Induced by the Integration of Transient Solar Heat V. Scherer, K. Roth and M Eck.. ..........................................................
193
Directions for Renewable Energy in Canada’s Smallest Province A. Tvivett.....................................................................................
205
Externalities Affecting the Viability of Wind Power for Hydrogen Production N. Kassem ...................................................................................
21 1
Non-Technical Barriers to Large-Scale Wave Energy Utilisation A.J.N.A. Sarmento, F. Neumann andA. Brito-Melo.. ....................................
225
Solar Power Agriculture: A New Paradigm for Energy Production U. Bardi .....................................................................................
235
Contents
Possibility Assessment of Wind Energy Utilization in Bosnia and Herzegovina F, Begic, A. Kuzagic and N.H. Afgun.. ....................................................
vii
249
Chapter 4 - Greenhouse Gas Emissions Earth Tube Ventilation System: A Project Pre-Feasibility Analysis Tool MA.A. Costa, A4S.A. Oliveira and N. Martins ...........................................
259
Greenhouse Gases Emission Reduction in an Urban Bus Fleet: Application to a Case Study in Funchal, Madeira Island A.M Simoes, P. Lages, T.L. Furius, C.M Silvu and M N . Aguas .......................
273
Environmental Impact of Hydrogen in Urban Transports K.R. Calhau, G.A. Goncalves and T.L. Furius.. ..........................................
285
Chapter 5 - Gas Production and Utilization Devolatilisation of Coal-Biomass Blends and Combustion Behaviour of their Chars P. Ciambelli, V. Palma, P. Russo, S. Vuccaro and V. Vaiano ...........................
295
Dual Fuel Combustion as a Way of Increasing Sustainability of Power Production D.R. Schneider and Z. Bogdan. ............................................................
3 11
Hydrogen Production by Allothermal Reforming of Ethanol for Fuel Cell Applications. Prototype Development A.J. Marin Neto, E.P. Silva, J.C. Cumargo, N.P. Neves Jr. and C.S. Pinto ............ 325 Hydrogen Production from Steam or Autothermal Reforming of Ethanol for Low- and High-Temperature Fuel Cell V. Galvita ....................................................................................
335
Hydrogen Fuel Cell Urban Buses Operating in the City of Porto G.A. Gonculves, T.L. Furias, R. Teixeiru undA. Silva.. .................................
343
Chapter 6 - Sustainability Development Energy in Slovenia and Croatia: Challenges and Possibilities for Sustainability A . Zidunsek, R. Blinc, I. Sluus andD. Nujdovski .........................................
353
Sustainability Assessment as a Basis for the Decision Making in Selection of Energy System F. Begic and N.H. Afgan ...................................................................
363
This page intentionally left blank
FOREWORD “Since 1990, the EU has been engaged in an ambitious and successful plan to become world leader in renewable energy. To take one example, the EU has now installed wind energy capacity equivalent to 50 coal fired power stations, with costs halved in the past 15 years. The EU’s renewable energy market has an annual turnover of 15 billion Euros (half the world market), employs some 300,000 people and is a major exporter. Renewable energy is now starting to compete on price with fossil fuels. In 2001, the EU agreed that the share of electricity from renewable energy sources in the EU consumption should reach 21% by 2010. In 2003, it agreed that at least 5.75 % of all petrol and diesel should be bio-fuels by 2010” GREEN PAPER European Strategy for Sustainable, Competitive and Secure Energy, 2006. The United Nations World Summit on Sustainable Development (UNWSSD) 2002 (Rio+lO) produced an Action Plan to guide increased efforts towards meeting Sustainable Development targets as set out in the Millenium Development Goals. In this document scientific and engineering communities are invited to take the leading role in the promotion and organisation of actions that are designed to deliver sustainable development. Modem technologies are a key part of achieving this objective. In this respect the utilisation of new and renewable energy technologies provide milestones for the improvement of the economic, social and environmental quality of our life. Due to the heavy reliance of modem civilisation on intensive energy consumption, the accelerated development of renewable technologies will play a crucial role in the realisation of the shared vision of sustainable development. It will impose a new dimension in the new and renewable energy technology development. Increasing efficiency is an ever-lasting demand of modem technologies. The miniaturisation of the systems for energy conversion has opened a new sphere of activity which is of particular importance to renewable energy technologies. The Evora International Conference on New and Renewable Energy Technologies for Sustainable Development marks a new development in modem technologies. It reflects on the progress towards meeting the goals set out in the Johannesburg Conference Action Plan. This Conference followed the objectives of the Lisbon (1998), Madeira (2000) and Azores (2002) Conferences “to bring together engineering practitioners, product developers and researchers with economists, political scientist and government administrators to explore the multifaceted relationship between renewable technologies and sustainable development”. Key lectures frame the technical and policy issues confronting the sustainable development movement and enrich the dialog between the various segments of the community. This dialog provided the context for more detailed technical presentations and panel discussions on energy systems, renewable resource exploitation, and the engineering design and optimisation for minimum consumption of water, with sanitation management, including agriculture productivity and biodiversity and ecosystem management. ix
X
Foreword
The Proceedings are organised in the following chapters: Renewable Energy Sources; Biomass Energy; Solar, Wind, Wave and Geothermal Energy; Greenhouse Gas Emissions; Gas Production and Utilisation; and Sustainable Development. The Editors would like to express their appreciation to Maria Fernanda Afonso for her excellent work on the technical editing of the manuscript. It is our duty to recognise also the contribution of Ana Mascarenhas for typing and Jorge Coelho for the preparation of the figures.
The Editors May, 2006
PROMOTING RENEWABLE ENERGIES AND ENERGY EFFICIENCY THROUGH THE CDM FUNDING OPTIONS K.D. PATLITZIANAS, A. FLAMOS, H. DOUKAS, A.G. KAGIANNAS, J. PSARRAS Management & Decision Support Systems Laboratory, Department ofElectrica1 and Computer Engineering National Technical University of Athens, Heroon Polytechniou 9, 157 73, Athens, Greece The promotion of Renewable Energy Sources (RES) and Energy Efficiency (EE) is a central aim of the world energy policy in order to contribute in reducing greenhouse emissions. Moreover, Clean Development Mechanism (CDM) allows Annex I countries to achieve their emission targets. Financial assistance to developing countries is needed in order to avoid high increase of their greenhouse emissions associated with their development. Today, a large number of Annex I countries are also setting up programmes to fund RES and EE investment through CDM. These include Denmark, Great Britain, Italy, Japan, Norway, Canada and Austria. These programmes are being applied by international organisations and numerous non-governmental organisations and they facilitate projects throughout the developing world. Initial experiences have also been gained from the “World Bank Prototype Carbon Fund - PCF” and the “Dutch CDM tender-programme CERUPT”. The most important result of these initial experiences is the creation of appropriate organisational structures and the use of qualified personnel by both the host and investor countries and the respective participants in the project, thus, building the capacity to facilitate the CDM. This paper describes the main role of CDM funding programmes and their potential role in promoting RES and EE projects.
1. Introduction Energy is essential for economic and social development. About ninety per cent of the world energy supplies are provided by fossil fuels, with the associated emissions causing local, regional and global environmental problems. Most energy projections show that current and expected hture global energy demand patterns are not sustainable. Even when assuming massive improvements in energy intensity, long term projections indicate that world energy demand may increase dramatically, with most of this increase, taking place in developing countries. These trends indicate that, in order to comply with the necessary conditions for the three dimensions of sustainability (economic, environmental and social) with respect to energy production and consumption, a decoupling of economic activity from fossil primary energy consumption should be achieved. Increasing the share of Renewable Energy Sources (RES) in the energy balance enhances sustainability and helps to improve the security of energy supply by reducing dependence on imported energy sources. In addition, the development of Energy Efficiency (EE) is a central aim of the world energy policy in order to contribute in reducing greenhouse emissions [ 11. The interest in RES and EE increased when the oil crises of the 1970’s made everyone aware of the fact that fossil resources would run out one day - but since there is some uncertainty about when that will actually happen the efforts made in this area remained rather tentative. 1
2
K.D. Patlitzianas et al.
Nowadays, growing environmental concerns and limitations in the exploitation of conventional energy resources have given new impulse in modem RES technologies. Beyond 2020, new technologies, such as hydrogen-based fuel cells and carbon sequestration, hold out promising prospects of plentiful, clean energy supplies for the world. So, RES projects and EE will need to play a greater role in the future energy mix in order to achieve low-carbon intensive energy systems [2]. Despite the fact that RES and EE provide a series of benefits in national and private scale, the development and dissemination process of such technologies has been slower than expected. The state cannot always place sufficient amounts for the modernisation of the energy sector through the promotion and penetration of these technologies. In addition, the users of energy often face the implementation of RES and EE projects hesitantly. A pool of inhibits that justify this attitude, includes [2]: The high initial cost of RES and EE projects in relation with the long time period of depreciation of the investment; The lack of available funds of the enterprises for the implementation of these projects. In most cases the enterprises cannot allocate sufficient amounts of their budget in such projects, since they have to overcome more demanding priorities, such as the improvement of their competitiveness and the identification of new markets; The financial, technological and performance risks of these projects are often high for an enterprise related to the expected results; The lack of awareness regarding the performance of modern and innovative renewable technologies. Besides the barriers mentioned above, a project investment in GHG abatement technologies might encounter several other barriers, especially when the investment takes place in developing countries. For instance, one barrier may be the flow credit worthiness that often developing countries have. The Kyoto Protocol through the Clean Development Mechanism (CDM) can exert a leverage effect to increase the attractiveness of new and renewable technologies [3]. The purpose of the CDM, as defined in Article 12 of the Kyoto Protocol, is to assist Parties not included in Annex I in achieving sustainable development and to assist Parties included in Annex I in achieving compliance with their quantified emission limitation and reduction commitments [4-61. The additional funding channeled through the CDM could assist countries in reaching some of their environmental and sustainable development objectives, such as C 0 2 emission reductions, cleaner air and water and reduced dependence on imported fossil fuels. CDM projects are comprised of two parallel income flows [7-91. The first flow refers to the base project and is typical of any traditional investment in an energy project; The second flow is the “carbon flow”. The products of this flow are tonnes of carbon dioxide avoided or reduced in the case of energy projects. Through the CDM project cycle, these reductions are converted into certified emission reductions (CERs).
Enera EfJiciency through the CDM Funding Options
3
These CERs can be purchased by greenhouse gas emitting sources, such as thermal power generation companies in industrialized countries, to meet domestic emissions reductions requirements. The market price of CERs will fall in between the cost of production (including all transaction costs incurred in the CDM project cycle) and the maximum purchaser price. This price is not anticipated to exceed the cost of domestic reduction measures in industrialized countries. The sale of CERs can significantly increase the internal rate of return (IRR) of an entire project. In the case of projects with a healthy IRR on the traditional flow, the CER is an incentive to implement a more greenhouse gas efficient project. In the case of projects that are not quite commercially viable, the added input of financing for the resulting CERs could make them viable. In this context, the objective of this paper is to present the main role of CDM funding programmes and their potential role in promoting RES and EE projects. The paper is structured along 3 parts, as follows: The first part is the introduction of the paper; The second part is devoted to the review of the objectives and the procedures of the main CDM - funding programmes (PCF, CERUPT, CDM Programme of Canada, Rabobank and Finland Programme) presenting the progress so far. This part is deovoted to the interpretation of the results of CDM funding programmes and concerns actions and measures to be undertaken for maximum RES and EE penetration; The last part presents the conclusions of this study. 2. CDM - Funding Options
Governments and private companies from Annex I parties are the main end-users of CERs. Funds to implement CDM project in a developing country can flow from three different trading models: Multi-lateral model - financial resources from Annex I investors flow through a centralized investment fund and channeled toward project activities in host developing countries; Bilateral model - this involves collaboration among investors from Annex I countries, project developers, and host country governments. These parties agree on project selection, funding and credit sharing arrangements; Unilateral model - the host country invests in a project, and sells or banks CERs. The carbon funds of the World Bank (Prototype Carbon Fund, Community Development Fund, BioCarbon Fund) are examples of multi-lateral funds using the World Bank as fund manager. The Dutch Government adopts several means of producing emission reduction credits, through multi-lateral organizations such as World Bank and European Bank for Reconstruction and Development (EBRD), through banks (e.g. Rabobank) or through bilateral contracts and via its own tender (CERUPT). Similarly, the Government of Japan uses the Japanese Bank of Industrial Development to manage its CDM funds. More recently, the Canadian and Netherlands governments have initiated
4
K.D. Patlitzianas et al.
bilateral transactions with several Latin American countries while the Danish government has recently come to bilateral agreements with Malaysia and China [lo]. The World Bank’s PCF and the Dutch Government’s CERUPT tender are the current main buyers of CERs through direct purchase transactions. As of 2002, the PCF Fund Management Committee and the Participants Committee approved a total of 16 CDM projects with emission reduction purchase agreement of almost 24 Mtn COz. The CERUPT tender, on the other hand, approved 18 projects in 2003, to generate emissions reductions of 16, 7 Mtn COZ. A number of PCF projects have been commissioned in 2002. Most of the PCF and CERUPT’s projects would be commissioned between 20032007. Furthermore, the Finnish and Swedish government launched CDM tenders in 2002 while the Austrian government is currently in the process of preparing CDM/ JI tenders. Today, CDM funds and investment approaches are growing and expanding. The Prototype Carbon Fund (1 80 million $), the Community Development Carbon Fund (100 million $) and the World Bank Bio-Carbon Fund (100 million $) are the main multilateral funds. The main government funds are categorized as follows [lo]: Own Tender: Dutch Government C-ERUPT Program; /Finnish CDWJI Pilot Program (20 million $); J Sweden International Climate Investment Program - CDM; J Austria J K D M Procurement Program; CommerciaVDevelopmentBanks: JRaboBank (Dutch Government); Japanese Bank of Industrial Cooperation (Japan CDM Fund - 4 billion yen); JDevelopment Bank of Japan (Japan CDM Fund - 3 billion yen); Multilateral Institutions: World Bank (The Netherlands Clean Development Facility - € 70 million); JIFC (IFC-Netherlands Carbon Facility - € 44 million); Bilateral Transactions: Canadian Government with Colombia and Chile; J Dutch Government with Bolivia, Colombia, Uruguay and Ecuador; /Danish CDM Portfolio -bilateral agreements with Malaysia and China. In the above framework, the most known CDM funding schemes are analysed as follows:
2.1. Prototype Carbon Fund (PCF) 2.1.I . Objective
Recognizing that climate change will have a significant impact on the world, on July 20th, 1999, the Executive Directors of the World Bank approved the establishment of the Prototype Carbon Fund (PCF). The PCF, with the operational objective of combating climate change, aspires to promote the Bank’s tenet of sustainable development,
Energy Eficiency through the CDM Funding Options
5
demonstrate the possibilities of public/ private partnerships, and offer a “learning-bydoing” opportunity to its stakeholders [ 1I]. The PCF has three primary strategic objectives [12]: Show how project-based greenhouse gas emission reduction (ER) investment can contribute to sustainable development; Provide the Parties of the UNFCCC, the private sector, and other interested parties with an opportunity to “learn by doing” in the development of policies, rules, and business processes for the achievement of emission reductions under JI and the CDM. 2.1.2. Progress so far
PCF has prepared and successfully negotiated an Emission Reduction Purchase Agreement with 12 Host Governments. The Prototype Carbon Fund signed emission reductions purchase agreements for most of its projects in 2005. More specifically, the following RES developing projects have been selected [12]: Brazil: Plantar Sequestration and Biomass Use; Bulgaria: 2 projects: J 13,4 MW biomass-based boiler to utilize wood waste produced at the Svilosa pulp and cellulose plant to replace coal; /District heating system upgrades for the city of Sofia; Chile: Mid size run-of river hydropower plant with capacity of 25 MW and 175 GWh displacing the dispatch of coal thermal power generation; Colombia: Wind farm in the northern part of Colombia to displace a mix of coal-and gas-based power generation; Costa Rica: 2 projects: Wind farm to displace thermal power capacity addition; J Small hydro to replace thermal power generation; Czech Republic: EE measures and RES through the Czech Energy Agency (CEA); Guatemala: Peaking run-of-river hydro-electric plant in the west coast of Guatemala to displace thermal power plants; Hungary: Conversion of Pecs Power plant’s existing coal-fired boilers to biomass, with annual generation 162 TJ heat and 334,3 GWh electricity; Indonesia: Implement energy efficiency measures including reducing clinker contents in the produced cement, burning alternative fuels for clinker formation, utilizing heat power generation; Latvia: Methane capture from waste management and carbon dioxide reduction from power generation; Moldova: Soil conservation afforestation of 14,394 hectares of degraded and eroded state-owned and communal agricultural lands throughout Moldova; Romania: Afforestation of 6,852 hectares of public land; South Africa: Durban Municipal Solid Waste 10 MW landfill gas-fired generator to produce electricity from landfill-collected methane;
K.D. Patlitzianas et al.
Uganda: Off-grid hydropower development in the West Nile region of Uganda. In addition several projects are under negotiations as of September 30, 2004, such Brazil: 3 projects: JAlta Mogiana Bagasse Cogeneration; JGuarani Cruz Alta Bagasse Cogeneration; JLages Wood Waste Cogeneration Facility; China: 2 projects: J Jincheng Coal Mine Methane Recovery; 4Xiaogushan Run-of River Hydropower; Costa Rica: 2 projects: J Rio General; JVara Blanca Wind Farm; India: 2 projects: JMunicipal Solid Wastes (ABIL); JNitrous Oxide Removal Project; Mexico: 2 projects: J INELEC Hydros; /Umbrella Waste Management; Philippines: North Wind Bangui Bay Wind Farm; Poland: Stargard Geothermal Heating; Vietnam: Grontmij Landfill in Ho Chi Minh City. A broad balance has been achieved in the number of projects undertaken in economies in transition and in developing countries. Furthermore, major emphasis has been laid on the development of projects in the area of RES technology such as, but not limited to, geothermal, wind, solar and small-scale hydro energy projects.
2.2. Certified Emission Reduction Unit Procurement Tender (CERUPT) 2.2. I . Objective
Through CERUPT (Certified Emission Reduction Unit Procurement Tender), the Netherlands is aiming at attracting CDM projects by providing funds for the acquisition of CERs. The responsibility for CDM in the Netherlands is with the Minister of Housing, Spatial Planning and the Environment (VROM). The Minister has appointed Senter International as tendering authority for CERUPT [ 131. The main objective of CERUPT is to buy CERs at a low price and acceptable risk. Senter International, the organisation that manages the CERUPT for the Netherlands’ Government, therefore assesses primarily price and project risk and does not have a preference for specific project types, which means that the submitted project proposals most likely reflect the actual supply on the market. CERs are generated and delivered as follows: CERs are generated in CDM projects. To generate CERs a CDM project has to realise a GHG emissions reduction compared to the baseline scenario. The CDM
Energy Eficiency through the CDM Funding Options
7
project’s feasibility should be proven at the moment of submission to Senter. The CDM project should be operational in the contract period. To ensure that all these requirements have been met, a business plan for the project must be included in the proposal. CDM projects have to be part of one of the following project categories: JRES (e.g. solar, wind, biomass, hydro); JEE (e.g. CHP, lighting, insulation, process optimisation); /The replacement of C 0 2- intensive fuels (e.g. oil to gas, coal to gas); J Waste processing (e.g. land fill gas extraction, waste incineration); CDM projects have to assist host countries in achieving sustainable development goals. The host country will have to judge impacts on the environment of the CDM project. This may require an Environmental Impact Assessment on request of the host country; Public participation is essential in CDM projects. Local stakeholders are to be invited to comment the draft project design. The final project design should take these comments into account. Further general requirements are that: Greenhouse gase (GHG) emission reductions should be additional to any that would occur in the absence of the CDM project. This must be demonstrated through in a baseline study; CDM projects are validated by Designated Operational Entities (DOE) and registered by the Executive Board (EB); Emission reductions are verified and certified by DOES, CERs are issued by the EB; CDM projects are subject to levies by the EB. The EB will withhold part of the CERs (2%) to cover administrative expenses and to assist least developed countries in meeting costs of adaptation to the adverse effects of climate. CDM projects in one of the least developed countries are exempt from the adaptation levy [14]. Furthermore, the EB will charge an administration fee on each registered CDM project. The volume of this fee has not been specified yet. 2.2.2. Progress so far The program has been launched on 1 November 2001 and resulted in 78 expressions of interests covering all relevant regions and technologies. Among these proposals, 26 qualified for the awarding stage and 18 proposals have been already awarded (13 March 2003) [ 141: 11 projects from Central and Latin America: JTotal emission reductions achieved: 7,7 Mton; /Type of projects: Biomass, geothermal, hydro, landfill, wind and EE projects; 0 7 projects from South-East Asia: /Total emission reductions achieved: 8,8 Mton; /Type of projects: Biomass, geothermal and wind projects. It is unclear what the future of CERUPT will be. There has only been one tender period which officially closed on January 3 1, 2002 and is therefore no longer open for
8
K.D. Patlitzianas et al.
submission of projects. Since CDM projects can often take a long time to process, it is likely that the Dutch Government is hoping to see first if the CDM Executive Board approves the 18 projects under CERUPT and then evaluate. As of yet none of the 18 projects have passed through the CDM approval process, though the Wigton Wind Farm project in Jamaica which CDM Watch describes as 'business as usual' has been approved by the CDM Executive Board 'subject to required changes'.
2.3. CDM Programme of Canada's CDM and JI Office 2.3.1. Objective
Canada's Clean Development Mechanism and Joint Implementation (CDM & JI) Office was established within the Climate Change and Energy Division of the Department of Foreign Affairs and International Trade (DFAIT) in 1998. The Office is the federal government's focal point for CDM and JI activities. It was created to enhance Canada's capacity to take advantage of the opportunities offered by the CDM and JI and is guided by an Interdepartmental Steering Committee comprised of representatives from Industry Canada, Natural Resources Canada, Agriculture Canada, Environment Canada, the Canadian International Development Agency and Canada's Climate Change Secretariat. Under Action Plan 2000, the CDM & JI Office received funds for the period 200 1-2005 in order to pursue the following three objectives: To strengthen Canada's capacity to take maximum advantage of the Kyoto Mechanisms. These include the Clean Development Mechanism, Joint Implementation and Emission Trades backed up by emission reduction projects; To encourage and facilitate Canadian participation in the Kyoto Mechanisms by building awareness, promoting cost-effective opportunities and lowering transactions costs, while also engaging developing countries and countries-in-transition in such activities; To assist Canadian entities in obtaining emissions reductions credits from CDM and JI type projects, this can assist Canada in meeting its Kyoto target, according to international rules and guidelines. With the elaboration of international rules and guidelines at Marrakech in November 200 1, the Office is primarily focused on project facilitation including: Guiding companies on the technical requirements for CDM activities and issues related to ownership of emission reduction credits; Providing financial support for market identification studies, feasibility assessments, baselines and monitoring plans, risk assessments and environmental impact studies; Preparing memoranda of understanding and project-specific agreements with host country governments. Project proposals submitted to Canada's CDM & JI Office will be assessed in reference to the following key criteria:
Energy Eflciency through the CDM Funding Options
9
Promotion of the objectives of the UNFCCC. The UNFCC seeks to stabilize GHG concentrations in the atmosphere at a level that does not create dangerous anthropogenic interference with the climate system; Contribution to the objectives of Canada's CDM & JI Office; Consistency with Canada's International Strategy on Climate Change Proposals must respond to at least one of the Strategy's objectives: JMaximise Canada's ability to meet its UNFCCC commitments and Kyoto Protocol targets at the lowest cost; J Contribute to the achievement of global climate change objectives; JMaximize business opportunities for Canadian business interests in international projects and initiatives on climate change [15-161.
2.3.2. Progress so far Indicative recent events of the Canada's CDM & JI Office in order to strengthen Canada's capacity for CDM are the following: Introductory Workshops on the Kyoto Mechanisms, SPRING 2006; Trade Team Canada Environment (TTCE) Mission to China, February 27 to March 10, 2006; CDM Seminars in India, March 1-11,2006; Canada's CDM & JI Office. DNA and Focal Point Workshop, March 27th-28th, 2006, Vancouver, BC; Carbon Expo, Cologne, Germany, (May 10-12,2006); Trade Mission to Carbon Expo 2005. May 11-13. Cologne, Germany; Business Development Mission to India Partnering Together. February 2005; AMERICANA 2005, Pan-American Environmental. Technology Trade Show and Conference; COP- 10, Buenos Aires, Argentina, (December 6th-l7th, 2004); Green Power Conference. Metro Toronto Convention Centre, (November 24 & 25, 2004); DNA and Focal Point Conference Ottawa, (September 30th- October lst, 2004); Canada's CDM & JI Office at the Emissions Marketing Association (EMA), Fall Meeting. Toronto, (September 2004); Canada's CDM & JI Office at Globe 2004.Vancouver, British Columbia, (March 3 1April 2); Globe 2004 Conference Program; CDM & JI Workshop at Globe 2004 (March 29,2004); Salon des technologies environnementales du QuCbec 2004 (March 17-19,2004) ; CDM & JI National Technical Workshop Ottawa (March 16-17,2004); Canada-Cuba Seminar on Clean Development Mechanism - Havana, (February 2324,2004).
10
K.D. Patlitzianas et al.
In addition, 65 CDM and JI proposals have been awarded from the following regions: Africa and Middle East (14 projects); Latin and South America (19 projects); Central Eastern and Europe (I 5 projects); Asia (17 projects). 2.4. Rabobank Fund 2.4.1. Objective
In January 2003, Rabobank lnternational has established the Carbon Procurement Department for the sale and purchase of project-related emission reductions from projects in developing countries More specifically, the Rabobank International concluded a contract with the Ministry of Housing, Spatial Planning and the Environment for the Dutch government to purchase ten million tons of certified greenhouse gas reductions from projects in developing countries in Asia, Latin America and Africa. These projects should have a sustainable character. In these projects, Rabobank may participate as financier via its international network.
2.4.2. Progress so far During the next eighteen months Rabobank will conclude contracts for the purchase of emission reductions with the owners of CDM projects. These will involve bilateral contracts and project-related reductions such as: A wind farm and a hydroelectric power station in China; A hydroelectric power station in Chile; A wind farm and a biomass project in India; A manure management project in Brazil. In some cases the extra proceeds generated by the sale of emission reductions will enable projects to be carried out. Otherwise, these projects would be difficult to be financed. In addition to reducing harmful emissions, these projects often also serve to strengthen the local economy, in part by improving the infrastructure and creating employment [ 171.
2.5. Finland Programme 2.5.I . Objective Finland has a commitment of 8% reduction of greenhouse gases according to the burden sharing agreement under the EU as a result of the Kyoto Protocol. The Ministry of Foreign Affairs (Development Cooperation) is currently exploring the possibilities of
Energy Eficiency through the CDM Funding Options
11
purchasing certified emission reductions via small-scale CDM projects. The guidelines for small-scale projects are simplified, reducing transaction costs, and small-scale projects are in general more beneficial for sustainable development than large-scale CDM projects. The government of Finland has published an invitation to submit project proposals [15-161.
2.5.2. Progress so far Currently, the Pilot Programme has seven on-going CDM and five on-going JI projects. The CDM projects are located in Viet Nam, Honduras, El Salvador, Costa Rica, India and Zambia. The JI projects include four projects in Estonia (Tamsalu, Kadrina, Paide and Pakri). The projects are at different stages. Some are already validated or determined and some are only in their early stages. Most of them are RES projects and include among others Hydro Power; Biogas; Landfill; Gas Recovery; Wind Power. 3. Conclusions
RES and EE constitute really important factors to sustainable development. However, the development and dissemination process of such technologies has been slower than expected because the use and promotion of RES and EE is inhibited by a number of factors. The most important factors are the high initial costs, the financial, technological and performance risks, the scarcity of investment capital and the investing priorities of private companies. The new climate change regime also offers an opportunity for RES and EE as they meet the two basic conditions to be eligible for assistance under the UNFCCC implementing mechanisms: They contribute to global sustainability through GHG mitigation; They conform to national priorities by leading to the development of local capacities and infrastructure. The application of CDM funding programmes could facilitate the development of RES and EE projects both in national and private level. RES, as indigenous sources can enable local development since they can secure the environmental integrity through the minimisation of C02 emissions. In addition, RES projects can contribute to the technology transfer through the CDM of the Kyoto Protocol [ 18-19]. CDM projects require upfront investments that are normally obtained from different sources such as loans, equity and grants. As in conventional projects, the funding of CDM projects can be arranged either through corporate or project funding. Additional project revenues (i.e. CERs) could leverage debt financing.
12
K.D. Patlitzianas et al.
In addition, the net financial gain derived from the sale of CERs is the difference between the project CER value and the transaction costs. There are three elements that influence the net impact of CERs on project profitability [20]: Value of CERs (low CER value implies low net benefits); Overall transaction costs (high transaction costs yield low net benefits); Up-front transaction costs (high upfront payments could also result in low benefits). Project developers generally expect up-front transaction costs within the range of 5 to 7% of the net present value of the revenue or total transaction costs around 10 to 12% of the net present value of revenue. A positive net financial gain means that CER revenues improve the financial viability of the project [lo]. Nowadays, the decisions adopted at the COP 9 and the recent adaptation of the directive linking the EU Emissions Trading Scheme to RES and EE projects, made clear that CDM will play a significant role as an effective mechanism for achieving the binding targets of the Kyoto Protocol. Within this framework it is expected that the EU Governments as well as the developing countries will make further use of the CDM funding programmes, which will play a more and more important role in funding RES and EE projects [21,22]. References
[ 13 European Commission (2003) Renewable Energy Technologies and Kyoto Protocol Mechanisms - Joint Implementation in Central and Eastern Europe - Clean Development in the Mediterranean Area.
[2] Second European Climate Change Programme (ECCP) Progress Report. (2003) Can we meet our Kyoto targets? [3] Geres R, Michaelowa A. (2002) A qualitative method to consider leakage effects from CDM and JI projects. Energy Policy 30 (6): 461-463. [4] The Pembina Institute for Appropriate Development. (2002) A User’s Guide to the Clean Development Mechanism (CDM). Second Edition. Canada. [5] Wees M. (2002) Current developments in CDM implementation: from theory to th th practice. National Policy Seminar on Clean Development Mechanism 5 -6 August, Kuala Lumpur. [6] Friedman S . (2002) The use of benchmarks to determine emission additionality in the clean development mechanism. US Environmental Protection Agency. [7] Anagnostopoulos K, Flamos A, Kagiannas AG, Psarras J. (2003) The impact of clean development mechanism in achieving sustainable development. International Journal of Environment and Pollution 21 (1): 1-23. [S] Flamos A, Anagnostopoulos K, Askounis D, Psarras J, Butzengeiger S, Gaast W. (2003) E-serem - A web-based manual for the estimation of emission reductions from JI and CDMprojects. Special Issue Mitigation and Adaptation Strategies for Global Change. Kluwer Academic Publishers. [9] Flamos A, Anagnostopoulos K, Askounis D, Psarras J. (2003) The multiple benchmark system application to Indonesia, Russia and Panama. Special Issue
Energy Eflciency through the CDM Funding Options
13
Mitigation and Adaptation Strategies for Global Change. Kluwer Academic Publishers. I101 Myung-Kyoon L. (2004) CDM Information and Guidebook. United Nations Environment Programme (UNEP) - Netherlands Ministry of Foreign Affairs. Unep Riso Centre on Energy, Climate and Sustainable Development Riso National Laboratory Roskilde, Denmark. [ 113 Word Bank (2006) Carbon Finance at the World Bank (http://carbonfinance.org). [12] Prototype Carbon Fund (PCF) - The World Bank. (2005) Annual Report - 2004. Washington. [13] Ministry of Housing, Spatial Planning and Environment of the Netherlands International Environmental Affairs Directorate - Clean Development Mechanism Division. (2006) CERUPT Guideline, (http://www.senter.nV). [14] Ministry of Housing, Spatial Planning and Environment of the Netherlands International Environmental Affairs Directorate - Clean Development Mechanism Division. (2006) CERUPT Tender Document. (http://www.senter.n1). [15] Coninck HC, Linden NH. (2003) An overview of carbon transactions - General characteristics and specific peculiarities, ECN, Peten ECN-C-03-022, March 2003. [16] Davidson 0. & Sparks D. (Eds) (2002) Developing energy solutions for climate change: South African research at EDRC - Energy & Development Research Centre, University of Cape Town, pp. 83-103. [17] Maruyama A. - UNEP, Aalders E. - IETA, Lee MK - UNEP (2005) Carbon Market Update for CDM Host Countries. Newsletter 1. [ 181 World Bank Carbon Finance Business, International Energy Agency, International Emissions Trading Association - PCFplus. (2004) Estimating the market potential for the clean development mechanism: Review of models and lessons learned. Washington DC. [19] Barcena A. (1994) What is capacity in environment? A developing country perspective in capacity development in environment. Proceedings of a Workshop held in Costa Rica, 9-1 1 November 1993, Paris: OECD: 46-53. [20] Wohlgemuth N. (2001) Directing investment to cleaner energy technologies: The role of financial institutions. In: Sustainable banking, the greening of finance. Greenleaf Publishing, Sheffield, 401-4 11. [21] Patlitzianas K, Kagiannas A, Askounis D, Psarras J. (2003) The policy perspective for RES development in the new member states of the EU. Renewable Energy 30 (4): 477-492. [22] Doukas H, Patlitzianas K, Kagiannas A, Psarras, J. (2005) Renewable energy sources and rationale use of energy development in the gcc region: Myth or reality? Renewable Energy 31: 755-770.
This page intentionally left blank
HJWS, ENERGY PLANNING TOOL FOR INCREASING THE PENETRATION OF RENEWABLE ENERGY SOURCES IN ISLAND ENERGY SUPPLY M. LERER, N. DUIC, L.M. ALVES, M.G. CARVALHO Instituto Superior Tecnico, Av. Rovisco Pais, 1049-001, Lisbon, Portugal The islands’ high energy costs make them excellent test beds for introduction of new technologies. Some islands are trying to become so called renewable islands, to satisfy their energy demand mainly or entirely from indigenous and renewable sources, thus increasing the security of supply, and employment opportunities, without necessarily increasing the costs. Advanced energy planning must consider different intermittent and constant resources, to satisfy the demand for energy, water and fuel for transport, taking into account security of supply on different time scales and necessary supply storage. Such studies cannot be done with traditional methods (LDC and Weibul curves) therefore time series analysis methodology is a requirement. The most promising energy storing technologies are reversible hydro (where geography allows) and storing hydrogen, making possible in the first case integration of the water system, and in the second case integration of the future hydrogen based power and transport systems. This paper describes the new developments in H2RES model, and its application on the example of an isolated island in Madeira archipelago, Porto Santo. HzRES is based on time series analysis that includes wind, precipitation and solar as renewable resources and presents reversible hydro, batteries and hydrogen as energy storages. It also allows for deferrable and hydrogen loads. Considering the present situation and the expected energy needs in Porto Santo by 2010, 7 possible cases of evolution of renewable energies in the island are presented, considering the power and road transport systems. The renewable sources available are wind and solar. In some of the cases, storage is introduced: the excess renewable energy can be used to electrolyze water and the produced hydrogen is stored for later use by fuel cells, either for supplying the power system or for transport needs. Renewable sources can be limited to 30% or to 100% of load, on hourly basis. It is seen that it is possible to significantly increase the penetration of renewable energy sources, albeit at a relatively high cost, with hydrogen storage technology. The H21ZESmodel can serve as a valuable tool for island energy planning.
1. Introduction
As discussed in the European Commission’s White Paper on Renewable Energy Sources [l], the United Nations Conference on Islands and Small Island States (Barbados 94, [ 2 ] and the lStEuropean Conference on Island Sustainable Development, there is a need to create conditions for developing the renewable energies, specifically regarding islands. The European Island Agenda [3] considers “the non-renewable energy sources as provisional solutions, inadequate to solve in the long term the energy problems of the islands”. Porto Santo, in Madeira archipelago, Portugal, is one of the European islands that aim at becoming 100% renewable. Environmental reasons are not the only good reasons to turn to renewable islands. Insularity, in general, means isolation andor dispersion, with the associated supply related obstacles. For these small local markets, the costs of transports, communications and energy are high when compared to the closest continental regions. Regarding electricity generation, most of islands depend on fossil fuel imports, whose costs are increased by transportation and that may suffer from security of supply problems. From the economic point of view, the higher the costs of conventional fossil solutions, the more attractive the renewable energy technologies become. 15
16
M. Lerer et al.
In some regions of Europe, intermittent renewable energies supply already about 10% of the demand (one example is Denmark’s wind electricity). In special cases of very good resources (for instance, steady wind) this fraction can be increased to about 20%, but for higher penetrations energy storage is necessary. Islands arise as excellent small scale systems to access the technical fitness of such schemes. As common tourist destinations, islands shall take special care of their environmental conditions, so use of fossil fuels should be decreased as possible. The increasing awareness of the public relating environmental issues makes it care about the impact of tourism in the local environment and “green” turns out to be appealing for marketing. As a consequence, Eco-tourism is a new market for islands, and allows highlight the sustainability of the community and tourist project. Dimensioning the renewable capacity of an island’s power system is a complex task. In some cases, the insular electrical grid may be connected to the mainland, but if the island is remote the grid will be isolated. If the island is connected, the excess of renewable energy can usually be taken by the larger system of the mainland without disturbing it. In case of isolated grids with important intermittent renewable energies supplies, such as solar or wind, the stability of the grid imposes limiting the amount of power taken and also requests control of frequency and voltage. The maximum intermittent power that enters the grid must be limited, and the rest of it should be stored or rejected. A very effective way of storing surplus energy is reversible hydro, but it may be inadequate for some geographical and climatic conditions. A new and very promising storage technology is based on hydrogen. The hydrogen is obtained by electrolysing water with the excess renewable energy, then it is stored, and its energy can be used later by a fuel cell, for stationary or mobile uses. For conventional energy planning, it is usual to use the statistic approach based on load duration curves (LDC), where the load is sorted by magnitude instead of time, and Weibull curves. This statistical approach cannot be used well when there are intermittent renewable sources, especially if the sources show patterns different from the intermittence of the load and when both are of the same order of magnitude, as desired for the renewable islands. In such case it is necessary to model the system on an hour per hour basis, for a representative year regarding resources. On islands that show a relevant tourism activity the load is likely to show not only the usual daily intermittence pattern but also an important seasonal fluctuation, and the same applies to water demand. For energy planning should be considered the expected evolution for the convenient time horizon (10-30 years). The conventional planning tools, like ENPEP, cannot be used in such situations, and several new energy-planning tools are being developed. EnergyPlan [4]for example is well adjusted for decentralised power generation, and it also integrates heat demand into the model, enabling the optimisation of combined heat and power generation, which already delivers nearly 50% of power to Danish system. It also integrates other intermittent resources, and optimises different strategies to treat the power excess. On the other hand, it does not treat hydro resource, water demand, hydrogen demand, reversible hydro, hydrogen storage, batteries and other specifics of the isolated power systems.
Energy Planning for the Penetration of RES
17
HOMER on the other hand is made particularly for small isolated power systems, although allows for grid connection, and has some of the needed technologies, but still lacks reversible hydro and water demand treatment, that is the cheapest way to store energy in those islands where there is potential [5].There are also models that contains more precise physical models of some technologies, like Hydrogems [6], but it also lacks hydro resource, reversible hydro storage and water demand, among others, and for the purpose of energy planning it is not necessary to go into conversion detail, beside the necessary level, although when it comes to component dimensioning, it might be beneficial. 2. H2RESModel
The H2RES [7] model is based on hourly time series analysis of demand (water, electricity and hydrogen), and of resources (wind speed, solar radiation and precipitation). It is composed of several modules that consider specific user data on system components (power, nominal efficiency, output curves, storage capacities, etc.) and control characteristics (maximum renewable penetration on hourly basis, security of supply, etc.). The wind module uses the wind velocity data, typically from the meteorological station, measured at 10 m height and, for a given choice of wind turbines, adjusts the velocity data to the wind turbines hub level and considers the efficiency curve to convert the velocities to the output. The solar module converts the total radiation on the horizontal surface, obtained typically from the meteorological station, to the optimum inclined surface considering geographical location, and then to electrical output. The hydro module takes into account precipitation data, typically from the nearest meteorological station, water collection area, and evaporation data, based on the reservoir free surface, to predict the water net inflow into the reservoir. Load module, based on a given criteria for the maximum acceptable renewable electricity in the power system, puts a part or all of wind and solar output into the system and discards the rest of the renewable output. The excess renewable electricity is then either stored as hydrogen, pumped water or electricity in batteries, consumed for some non-time critical use, or rejected. The energy that is stored can be retrieved later, and supplied to the system as electricity. The rest of demand is covered from Diesel blocks. The model can also optimise the supply of water and hydrogen demand. There are two abilities of H2RES that make it specifically suited for islands: it can manage the water supply and consider hydrogen load for other than power supply. The hydro module allows managing potable water demand and excess energy storage at the same time, using pumping as a deferrable load and considering pumped potable water as a storage for reversible hydro power. Hydrogen has been seen as a new energy vector and can be specially suited for islands and remote regions, when produced in a renewable way, because it increases the security of supply and opens the renewable market for other areas, as for instance transport. For islands lacking fresh water, there is the possibility of developing a new module based on desalination to be associated to the hydro module. In the current version
18
M. Lerer ef al
~esa~ination c m be treated as deferrable load, but there is no link between it and water storage module.
3. Port0 Santo Popto Santo i s a 42 Ian2island in Madeira Archipelago, situated at about 16"W 32% on p h e rover a ~ 800 km away from any cont~ne~tal the Atlantic Ocean? an ~ ~ ~ ~ ~ p e r iregion land. The main economic activity is towism, with a clear predom~ancein summer, which is ~ ~ s ~ o nfor s ~anbincrease ~e from about 5500 occupants in winter to about 15000 in summer. The island has a mild temperature all year (average of 18.4"C), showing a temperate, oceanic, semi-arid climate. Regarding final energy use, transport arises as the most important sector (Fig. I), also indirectly related to tourism activity.
'industry
Hotel Sector
Agriculture, Cattleraising, Fishing -.
I
Fig. 1. Port0 Santo final energy distribution in 2000, by sector
3.2. Resou*ces
The solar resources of Port0 Santo are quite good, presenting about 1700 k W m 2 year. Tbe annual distribution of solar radiation is shown in Fig. 2. Currently there are some localked ~ns~a~lations of thermal solar panels for heating water at ~ w e ~ ~ but ~ ~ g § , photovoltaic technology has not been used except for eventual stand alone systems and do not provide electricity to the grid.
Energy Planningfor the Penetration ofRES
I
19
1,200,
I
0
1000
2000
3000
5000
4000
6000
7000
8000
hours
Fig. 2. Hourly average total solar radiation on horizontal surface, meteorological station, Porto Santo, 2000
The 30 years averaged wind velocity, as measured at the meteorological station on Porto Santo airport, at 10 m height, is only 4.2 d s , what is not particularly high. The lowest monthly average is 3.4 m/s corresponding to September, while the highest average corresponds to April, 4.5 mh, which makes it fairly constant throughout the year. The results from exploitation of installed wind turbines are giving 35% higher average wind velocity at 10 m, at the location. The wind velocity at the airport meteorological station is given for year 2000 in Fig. 3:
0
1000
2000
3000
4000
5000
6000
7000
8000
hours
Fig. 3. Hourly average wind velocities at 10 m, meteorological station, Porto Santo, 2000.
The island lacks water resources, both from springs, wells and precipitation, so hydro energy is not considered an option, unless the water system was re-designed and water storage could be integrated into a reversible hydro facility. The biomass is scarce. The wave conversion technology is not yet feasible. It was not considered technically practical usage of any other renewable resources. Among the fossil fuels, only oil derivatives are imported to Porto Santo: petrol, Diesel, fuel oil and LPG. In Madeira Archipelago there has been discussion on the potential for LNG. The technology of sea transport of liquefied natural gas is well established on large scale. It is expected that with maturing of the technology, smaller scale applications will become
M Lerer et al.
20
viable, which will show a new way for islands, starting from bigger ones, like Cyprus or Madeira, but if the scaling down does not reach some unsurpassable obstacle, it is probable that it will also be applicable to small scale islands, like Port0 Santo, in the same way LPG distribution is made today. 3.2. Electricity Demand and Power System
The annual load profile for electricity is given in Fig. 4.
1000
'
I
0 , 0
1000
2000
3000
4000
7000
6000
5000
8000
hours
Fig. 4. Hourly average electricity system load, 2000.
The system had in 2000 a peak load of about 5.5 MW and a low load of about 1.1 MW. The seasonal fluctuation is considerable and mainly related to tourism activity. The increase in electricity peak and annual demand for recent years is shown in Figs. 5 and 6:
1.:
-
..
..
..
. " " .... .
,.
. .. . .
...
.,
,
__
.
1
2,
1
Fig. 5. Peak Demand, 1996-2000 [8].
..
,
.
......
.
..
Energy Planningfov the Penetration ofRES
Annual Electricity Production
1
1991
1992
1993
1994
1995
1996
1997
21
IHThermal .Wind)
1998
1999
2000
2001
2002
Fig. 6. Annual electricity production 1991-2002 [8].
Regarding the production of electricity, there is a wind park with two 225 kW Vestas and one 660 kW Vestas, and a thermal plant that has 3 fuel-fired units (installed from 1998 on) of 3.41 MW each and 2 old diesel-fired groups (3.5 MW each) that are used mainly for backup and rare peak times. Since the installation of the 660 kW Vestas in the end of 2000, the wind has been providing about 8% of the yearly electricity demand. For 2000 the total demand was 25.2 GWh, increasing to 29.8 GWh in 2002: Table 1 , Electricity generation history, Porto Santo, 1991-2002.
The lack of water, mainly during summer, led to the installation of a reverse osmosis desalination plant in 1990, which is currently the bigger electricity client of the island. As the desalination plant doesn’t have storage for fresh water, the plant functions mainly during the day, contributing to increase the load at peak times. Installing a water tank to store water produced during the night would allow a better management of electricity consume, both reducing the load at peak time and increasing the load at night, thus admitting the penetration of intermittent sources to increase. 3.3. Road Transport Sector
The number of vehicles has been growing fast in the last years in the whole archipelago, especially concerning private owned vehicles, thanks to the increase in wealth of local communities and the great development in road accesses. Porto Santo has 5000 yearlong inhabitants and the penetration rate of private owned vehicles is 522%0(the greatest of all
22
M. Lerer et al.
the Municipalities in the archipelago) or about 2250 vehicles [9]. In 2000, Porto Santo road transport consumed 2274 toe of fuel, which corresponds to 75% of the total spent in transport in the island (3028 toe), and almost 40% of all primary energy consumed (5761 toe). The land transport has doubled its fuel consumption between 1991 and 2000.
3.4. Energy Demand Expected Evolutionfor 2000-2010 Based on hypothesis of regional development, national and European orientations, international markets of energy products, technological progress and environmental aspects, three scenaria (high, low and alternative) were constructed for the energetic sector development in the Madeira Region, as described in [9]. The values presented here for the demand are based on the high development scenarium forecasts, and it is assumed that Porto Santo Island will show the same evolution as Madeira Region as a whole, The forecasts can then be used for medium-range planning on the energy sector. The yearly demand growth rate is expected to be 8% during 2000-2005 period and 7% during 2005-2010 period, both concerning electricity and fuel for road transport. This means Porto Santo’s electricity production will increase from 25.2 GWh in 2000 to 37.0 GWh in 2005 and to 51.9 GWh in 2010. If the yearly peak follows the same growth rate, it will increase from 5.4 MW in 2000 to 8 MW in 2005 and to 1 1 . 1 MW in 2010. The fuel consumption for land transport will grow to 3341 toe in 2005 and 4686 toe in 2010.
4. Becoming Renewable Small power systems usually have their power frequency controlled by a single block. Small amounts of power coming from other sources will easily adjust to synchronous operation. It is safe to say that, at any single hour, the maximum power that can come from sources without frequency control is around 30%. That allows for even higher values during smaller periods of time, but at any moment will in principle stay under 50% of the load. With the present wind turbines technology, it would not be possible to increase the wind capacity in Porto Santo because it is already higher than 50% of the low load value (1.1 MW). This means that, if a low load situation occurs in a time when wind availability is good, the momentary penetration of wind may be higher than 50%, and the quality of the electricity delivered will be unacceptably low. Such a limit placed on intermittent sources will, typically for wind, allow only 10-15% of the total yearly electricity produced to be renewable. According to [lo] for a 5 MW system, as Porto Santo, one could possibly expect, at the current level of technology, to get less than 20% of electricity from wind, on yearly basis. That would mean either accepting more than 30% of wind electricity in some intervals, with consequences to the quality of electricity delivered, or installing active power controlled variable pitch wind turbines that can easily adjust the output to the load, andfor would condition installation of frequency and voltage control for all wind turbines and other renewable sources, and some kind of energy storage [ 1 1 - 161.
Energy Planning for the Penetration of RES
23
In this work we consider that all new wind turbines to be installed should be of variable pitch and be indirectly connected to the grid, with asynchronous AC-DC-AC conversion and frequency and voltage active control (Enercon 300). For the photovoltaic and the fuel cell systems, the connection is DC-AC with frequency and voltage active control. We consider 7 different cases that allow increasing the renewable share, especially Concerning the electrical demand:
4.1. Case I : Wind 30% Baseline Scenarium This scenarium is used as baseline scenarium, the one that could be considered a best conservative economically viable scenarium, while in the same maximising the renewable penetration. Wind energy now can be considered economically viable on islands, when not surpassing certain limit of penetration. For the purpose of this scenarium that limit is established at 30% renewable penetration on the hourly basis, meaning that no more than 30% of load during one hour can be covered by power coming from wind turbines. The system is optimised so that yearly wind penetration was maximised while keeping the rejected wind electricity close to 10%.
4.2. Case 2: Wind 100% Scenarium This scenarium allows the hourly penetration rate to reach loo%, meaning that when there is enough wind, Diesel generators will be completely shut down. The number of wind turbines was optimised so that yearly wind penetration was maximised while keeping the rejected wind electricity close to 30%.
4.3. Case 3: FCWind 30% Scenarium This scenarium is similar to case 1 in limiting the wind penetration to 30% on hourly basis, and optimising it so that less then 10% of wind potential is rejected. The difference is that a peak shaving fuel cell is added, together with electrolyser and hydrogen storage, in order to store the excess of wind generated electricity. Water is electrolysed to produce Hydrogen using the renewable excess that cannot be taken directly by the power system. The case defines the peak periods when load is more than 80% of the weekly moving hourly peak. The system is optimised so that fuel cell covers 1% of the yearly load.
4.4. Case 4: FCWind 100% Scenarium FCWind 100% scenarium enables that all the electricity in Port0 Santo is supplied from wind turbines, either directly, or through storage oE the excess production into hydrogen. Direct wind power is maximised (up to 100% hourly load), so that no more than 30% is rejected, and the rest is supplied via fuel cell. This is a 100% renewable scenarium concerning electricity.
24
M Lerer et al.
4.5. Case 5: FCSolar Wind 30% Scenarium This scenarium is similar to case 3, just that wind turbines are kept exactly as in case 1 and solar PV panels supply the rest of renewable electricity, with penetration from intermittent sources limited to 30% on hourly basis, and optimised so that less then 10% of renewable potential is rejected. The fuel cell is used for peak shaving, supplying approximately 1% of yearly load.
4.6. Case 6: FCSolar Wind 100% Scenarium This scenarium is similar to case 4, just that instead of only wind, such a combination of solar and wind is chosen, in which for the same results (30% rejected, 0% Diesel) one try to minimise the sum of installed wind + solar power. This is a 100% renewable scenarium for electricity generation.
4.7. Case 7: FCWind Transport 30% Scenarium This is very similar to case 3, just there is additional hydrogen load, represented by three shuttle vans operating on fuel cell, and doing 40,000 km per year per shuttle. With an average car economy of 0.05 kgH2h-1, that makes 2000 kgH2 per shuttle per year. For simplicity we can consider the hourly load of Hydrogen for the shuttles as constant trough the year, which does not cause much error as the load is quite small. The hydrogen for stationery fuel cell as well as for shuttle vans is produced by electrolyser using the excess wind electricity and stored, In order to have reasonable security of supply of hydrogen for shuttle, there has to be one month transport operation reserve (since the stationery use can be backed up with a Diesel engine). When the hydrogen falls under that level, stationery fuel cell will not be used. For these simulations we assumed the following efficiencies for the equipment: PV: 5.78% total (8.5% nominal for PV panels, 80% for converter, 15% losses); Electrolyser: 60%; Fuel Cell: 50%; Diesel blocks: 30%. The losses associated with compressing and storing the Hydrogen are included in the electrolyser efficiency. The wind turbines efficiency curve is given by the producer. Next we present the results for year 2010.
4.8. Results for 30% Hourly Penetration: Cases 1 , 3 , 5 and 7 These cases refer to a limit of 30% on hourly penetration of intermittent renewable sources but differ in technologies installed, regarding both generation and possibility of storage of surplus energy. Fig. 7 shows the equipment power and storage capacity that should be installed for each scenarium, in 2010:
Energy Planningfor the Penetration of RES
60
t
I -
50
40 P
30 0
a! m
E
20
f
10
0 Case 1 Wind 30%
Case 3 FCWind 30% Case 5 FCSolarWind 30%
Case 7 FCWindTransport
Fig. 7. The comparison of components needed to be installed in different cases of 30% intermittent limit, year 2010.
The fraction of electricity supplied by renewable sources (Fig. 8) is about 11% in Case 1, 16% in Case 3, 17% in Case 5 and 17% in Case 7. This means that, by applying the peak shaving Fuel Cell and the Hydrogen based storage cycle (Cases 3, 5 and 7) we could increase renewable penetration in about 5-6%: I
Case 1 Wind 30%
Case 3 FCWlnd 30%
Case 5 FCSolarWind 30%
Case 7 FCWindTransport
Fig. 8. The demand supply per technology, in different cases of 30% intermittent limit, year 2010.
On comparing the generation technology chosen, we see that by installing a convenient power of photovoltaic panels we can reduce the Hydrogen storage to 115 (electrolyser
M Lever et al.
26
power also decreased, although not as significantly), while keeping the fuel cell at the same value, and still increase a little the overall renewable penetration. This is because the intermittence of the solar potential is smaller than the intermittence of wind potential, and so the storage is filled up more regularly and can cover the same supply by the end of the year. The increase in storage and electrolyser from Case 3 to Case 7 is related to the additional Hydrogen load for shuttles. The significant increase in hydrogen storage comes not from a great increase in hydrogen load, but from the increase in security of supply condition for transport (a reserve of 30 days of hydrogen for transport is considered as a minimum for allowing the hydrogen to be used for power system). From the point of view of power system, those cases are identical. As result from the introduction of the peak shaving fuel cell, the use and wear of the Diesel engines will be reduced, as can be seen (Fig. 9) from the Load Duration Curves for cases 1, 3 , 5 and 7. The curve for case 3 is very similar for those of cases 5 and 7, showing that there is not much difference when we introduce the solar panels and confirming that case 5 and 7 are approximately equal concerning electricity.
0
1000
2000
3000
4000
5000
6000
Diesel output [kWhlh]
Fig. 9. Load duration curves for diesel generators.
4.9. Results for 100% Hourly Penetration: Cases 2 , 4 and 6 Cases 2 , 4 and 6 differ in technologies installed, regarding both generation and possibility of storage of surplus energy, but they all allow for 100% penetration of intermittent renewable sources. Also, Cases 4 and 6 represent 100% Renewable Scenaria for electricity generation. Fig. 10 shows the equipment power and storage capacity that should be installed for each scenarium, in 2010:
Energy Planning for the Penetration ofRES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 .~~~~~
1
~~.~~ . . . . . . . . . . . . . . . . . . . . . . . .
21
~ ~ . . . . . ~ . ~ 3000. . ~
2500
Case 2 Wind 100%
Case 4 FCWlnd 100%
2000
F
1000
t
Case 6 FCSolarWind 100%
Fig. 10. The comparison of components needed to be installed in different cases of 100% intermittent limit, year 2010.
It is clear, for 100% renewable scenaria, the storage size reduction thanks to PV introduction is not very important. This happens because, as the Hydrogen cycle must cover all unsupplied demand, the storage is not allowed to be empty at any time and the relevance of the smaller intermittence of solar resources is not as important as in the peak shaving cases. As seen in Fig. 11, in Case 2, when 100% intermittent limit is accepted but no storage is used, the electricity supplied by wind is about 50% on yearly basis. With storage based in Hydrogen Cycle (Cases 4 and 6) one manage to increase the renewable penetration to loo%, and virtually eliminate the use of Diesel blocks, that are kept as backup.
Care 2 Wind 100%
Case 4 FCWind 100%
Case 6 FCSolarWind 100%
Fig. 1 1. The demand supply per technology, in different cases of 100% intermittent limit, year 2010.
M Lerer et al.
28
It was shown how using hydrogen loop the renewable penetration could be increased to 100% in Porto Santo (cases 4 and 6). Since that would currently be too expensive a possible first step would be to install peak shaving systems (cases 3 and 5). Also, using hydrogen for transport would be straightforward, as seen in case 7. In order to have a reference case, a conservative scenarium for renewable energies penetration, as described in case 1, is based on the proven technology of wind farms, with up to 30% penetration on hourly base. In this case, and keeping the rejected renewable energy lower than lo%, more 5x300 kW wind turbines would have to be installed until 2010. These turbines could provide about 1I% of the electricity used in Porto Santo. Adding a peak shaving FC, fuelled on renewable Hydrogen, even if the annual load is of about 1% only, can increase the renewable penetration to about 16% of the supply, but this scenarium needs an infrastructure with electrolysers and H2 storage. The decision on including the still expensive PV panels together with the wind turbines should be weighted considering the reduction in cost for Hydrogen storage. Although the diversification of energy sources can be an advantage, because the overall intermittence is reduced and security of supply is increased, the decision must consider geographical and economical concerns as well as strategy arguments. In case we allow for up to 100% penetration of intermittent renewable energy on hourly basis, with no energy storage (case 2), about 50% of electricity could be provided by wind energy. The switch to 100% renewable (cases 4 and 6) corresponds to a high investment in generation (PV and wind turbines on one hand, electrolysers and FC on the other) and energy storage. Acknowledgements The authors would like to thank the European Comission for supporting the project in which this paper was based: “Renewable Energy Solutions for Islands RENEWISLANDS” Accompanying Measure of the European Commission’s Framework 5 ‘Energy, Environment and Sustainable Development’ Research Programme. Contract NO. NNE5-2002-00073. References [ l ] European Commission, Communication from the Commission - Energy for the Future: Renewable Sources of Energy, White Paper for a Community Strategy and ActionPlan, COM (97)599 final (2611 1/1997),
http://europa.eu.intfcommlenergy/library/599fi~en.pdf. [2] United Nations, Report of the Global Conference on the Sustainable Development of Small Island Developing States, NCONF. 16719, Bridgetown, Barbados, 1994, http://www.unep.ch/islands/dsidscnf.htm. [3] INSULA - UNESCO - European Commission - Consello Insular de Menorca, First European Conference on Sustainable Island Development, The Minorca Commitments, European Island Agenda, Ciutadella Declaration, Minorca, 1997.
Energy Planningfor the Penetration of RES
29
[4] Lund H, Munster E. (2003) Modelling of energy systems with a high percentage of CHP and wind power. Renewable Energy 28: 2179-2193. [51 HOMER, The optimisation model for distributed power. http://www.nrel.gov/homer/. [6] HYDROGEMS, HYDROGen Energy Models. http://www.hydrogems.no/. [7] Duic N, Carvalho MG. (2003) Increasing the penetration of intermittent renewable energy sources in island energy supply. CD, Proc. of the 2nd Dubrovnik
Conference on Sustainable Development of Energy, Water and Environment Systems, Dubrovnik. [S] Empresa de Electricidade da Madeira. http://www.eem.pt/. [9] Melim Mendes JM, Oliveira F, Freitas D, Henriques C , Olival E, Branco S. (2002) Plano de politica energetica da Regiao Autonoma da Madeira, Funchal. [lo] Lundsager P, Binder H, Clause NE, Frandsen S, Hansen LH, Hansen JC. (2001) Isolated systems with wind power. Main Report, Riso-R- 1256(EN), Riso National Laboratory, Roskilde, http://www.risoe.dklrispubWEA/veapdf/ris-r-1256.pdf. [ll] Milborrow DJ. (2001) Wind and storage - and a look at Regenesys. "Windpower on Islands" Conference, Gotland, 12-14 September. Gotland University. [ 121 Milborrow DJ. (2002) Assimilating wind. IEE Review 48. [ 131 Milborrow DJ. Penalties for intermittent sources of energy. http://www.cabinetoffice.gov.uWinnovationl;!002/energy/report/working%2Opapers/Milborrow .pdf. [ 141 Stavrakakis G. (2000) Conclusions, Workshop: Dissemination of the advanced control technologies and SCADA systems for the isolated power networks with increased use of renewable energies. Ajaccio, Corsica, http://power.inescn.ptljaneca/pdfs/CorsicaConclusions.pdf. [15] Altmann M, Niebauer P, Pschorr-Schoberer E, Zittel W. (2000) WHySE WindHydrogen Supply of Electricity, Markets - Technology - Economics, Wind Power for the 2 1st Century, Kassel, Germany. [ 161 Putnam R. (1996) How difficult is it to integrate wind turbines with utilities? Wind Energy Weekly 680.
This page intentionally left blank
THE SHAPE OF COMPLETE RENEWABLE ENERGY SYSTEMS
IN THE WORLD H. YAMAMOTO (I), K. YAMAJI(*) (I)
Socioeconomic Research Center, Central Research Institute of Electric Power Industry (CRIEPI), Tokyo, Japan. "School of Engineering, The University of Tokyo, Tokyo, Japan
The uses of fossil fuels cause not only the resources exhaustion but also global warming. Before they exhaust all the fossil fuels or fall into catastrophic climate change, they need to develop the new energy systems that are fossil-fuel-free and complete renewable. In the study, the authors developed a global land use and energy model including new technological concepts such as intermittent renewables (photovoltaic and wind) directly connected with electricity storage of small capacity and large capacity. Using the model, the authors conducted simulations and obtained the following results. In the sustainable scenario in 2060 (when they achieve zero fossils and completely renewable energy systems) the dominant resource of primary energy supply will be photovoltaic directly connected with electrolysis at the amount of 455 EJ.yeai' in the world. The second and the third largest resources are biomass and photovoltaic directly connected with large capacity storage at 322 EJ.yeai' and 118 EJ.yeai', respectively. At that time, the dominant energy resource of the power generation connected with the electricity grid will be the photovoltaic directly connected with large capacity storage at 118 EJ.year-l in the world. BGCC (biomass gasifier combined cycle), hydropower, and photovoltaic directly connected with small capacity storage are above 40 EJ.yeai' each in the world. However, the energy system costs in the sustainable energy scenario will be at 18 trillion US$.year-l and be almost three times as expensive as the costs in the base case (without constraints of fossil fuels or C 0 2 emissions). The sustainable energy systems are the expensive energy systems.
1. Introduction
The uses of fossil fuels cause not only the resources exhaustion but also environmental problems such as global warming. Before they exhaust all the fossil fuels or fall into catastrophic climate change, they need to reach the sustainable energy systems that are fossil-fuel-free and completely renewable. However, in the past studies the clear shape of the sustainable energy systems was not found. Even in the cases of high carbon taxes the authors could not find the sustainable energy path in the past study [ 11. In order to find the shape of the sustainable energy systems, the authors developed a global land use and energy model (GLUE) [2-31. The model figures the global energy supply systems in the future considering the cost minimization. The model includes overall energy resources including fossil fuels and renewables and overall energy conversion technologies including a variety of technologies concerning power generation, gasifier, and liquefaction. In the study, the authors developed the new version of the model (GLUE 3.0). It is based on a new calculation method of foresighted dynamic recursive. In addition it includes new technological concepts of intermittent renewables (photovoltaic and wind) directly connected with electricity storage of small capacity and large capacity. The authors assume a sustainable energy scenario that they will set a target in 2030 when they will reach no fossil fuel and complete renewable energy systems by 2060 (in 30 years after 2030). The authors conduct simulations under not only the sustainable energy scenario but also a base scenario (that is without constraints on fossil fuel uses or 31
32
H. Yamamoto & K. Yamaji
C 0 2 emissions). Then the authors discuss the shape and the cost of the sustainable energy systems.
2. The Model of Glue 3.0 2.1. Outline of the Model In this section the authors outline the structure of Global Land Use and Energy model version 3.0 (GLUE3.0) model. The world is divided into 11 regions in the model (Table 1). The model can conduct simulations from 2010 to 2200 with ten-year time step. The authors show the simulation results until 2060 in the study. The model consists of two parts: an energy systems part and a land use part. The energy systems part is based on a global energy systems model named New Earth 21 (NE21) and the land use part is base on a global land use and energy model (GLUE-11) (Fig. 1) [2][3]. The land use part covers a wide range of land uses and biomass flow including food chains, material recycling, and discharge of biomass residues. Those two parts are connected through common variables concerning supply potential of bioenergy and photovoltaic. The model minimizes the summation of the energy system costs. The authors prepare data set for the model using the data of FAO, IPCC, World Bank, DOE, and so on. For example, the final energy demand will increase following IPCC SRES-B2 scenario. Most input data are explained in the reference [2]. 2.2. Renewables in the Model
In the model the authors considers the following renewables: bioenergy (including fourteen kinds of bioenergy), hydropower and geothermal power (hydropower is the representative and includes geothermal power in the study), wind power, photovoltaic, and innovative renewables (such as space power system). Intermittent renewables (such as wind power and photovoltaic) are divided into four kinds such as type 1 (without electricity storage), type 2 (directly connected with electricity storage of small capacity), Table 1. Regions in the model.
CO2 balance in land use part (wood biomass
-.
.
I
CO2 balance in ene
I
Energy part
Land use part
pasture, other land, addtiional arable
each primary energy (resources and wsts; except bioenergy)
wnversion
slash-and-burning, (biomass pmduction per
(mundwocd to timber, cerealto meat, etc.)
(population,
technologies (ind efficiency and
secondary
demands per
TI
lntrcduction
energy. (including bioenergy)
of each sewndary energy
r
'
t Determination of biomass flow
~upp~ycurve of each bioenergy -+
Discharge of chemical product scrap Determination of energy flow
34
H. Yumumoto & K. Yumaji
type 3 (directly connected with electricity storage of large capacity), and type 4 (directly connected with electrolysis and not connected with electricity grids). It is assumed that type 2 improves stability of the outputs of the renewables and eases the upper limits of introduction of the intermittent renewables in the electricity grid compared with type 1. Type 2 is being developed in Japan funded by NED0 (New Energy and Industrial Technology Development Organization) [2]. It is assumed that type 3 is similar to dammed hydropower that can supply electricity to the grid following the demands. Therefore type 3 is a kind of stable energy and is not suffered the constraint of the electricity grid stability. The authors consider that type 3 will be realized when the price of the capacity (in kWh) is reasonable. Type 4 is not connected with electricity grids and is not suffered the constraint of the electricity grid stability. 2.3. Sustainable Energy Scenario and Base Scenario The authors assume a sustainable energy scenario that they will set a target in 2030 when they will use no fossil fuel by 2060 (in 30 years after 2030). In the other words, they can use only renewables (such as hydropower, wind power, photovoltaic, and innovative renewables (that the authors assume at Space Power System) in 2060. Under the assumption of resource estimation of WEC and an energy demand scenario of IPCC SRES B2, they will exhaust reserves of conventional oil conventional natural gas until 2060 (Figs. 2 and 3)[2]. In 2060 there will be plenty of reserves of the other energy resources (such as unconventional oil, unconventional natural gas, and coal). However the depletion of conventional oil and conventional natural gas will have human reconsider the relationship between human and exhaustible energy resources. It is assumed that they will be able to use intermittent renewables with electricity storage or electrolysis (see the previous subsection). It is assumed that they will adopt nuclear (LWR) phase-out where they will introduce LWR following the LWR scenario at IAEA by 2020 and will phase out the capacity after 2020 and abolish it in 2060 [2]. In the study the authors conducts not only the sustainable scenario but also a base scenario. The base scenario is the scenario without constraints about fossil energy uses and C 0 2 emissions. The authors discuss the results such as the structures and the costs of the energy systems concerning the two scenarios. 3. Simulation Results
Using the model of GLUE, the author conducted simulations of two scenarios (that are the sustainable energy scenario and the base scenario). The authors found feasible solutions even in the sustainable energy scenario. The authors explain the simulation results of primary energy supply, electric power supply, and energy trade.
The Shape of Complete Renewable Energy Systems in the World
35
3.1. Primary Energy Supply In the base scenario the dominant resources of primary energy and power generation are both coal in the world.
-+Cumulative consumption 20000 18000 16000 14000 12000 10000 8000 6000 4000
,
,
40
35 30 25 % 0 20 .= 15 10 2 h
-
5 0
2000
I
45
1970 2000 2030 2060 Fig. 2. Proven reserve and RIP ratios of conventional oil.
+Cumulative consumption FUP ratio(left-hand axis)
a,
2 c
-
g %.z 0 0
k
20000 18000 16000 4000 2000 0000 8000 6000 4000
70
I
60 50
2
30
'z
20
$
m a, 40 2 0 L
10
2000 0
0 l J l U
LUUU
LUJU
LUUU
Fig. 3. Proven reserve and RIP ratios of conventional gas.
36
H. Yamamoto & K. Yamaji
In the world in 2060 in the sustainable energy scenario the largest primary energy resource will be photovoltaic type 4 (that is photovoltaic directly connected with electrolysis and supplies hydrogen) at 455 EJ.year-'. The second and the third largest resources are bioenergy at 322 EJ.year-' and photovoltaic type 3 (that is photovoltaic directly connected with electricity storage of large capacity) at 118 EJ.year-', respectively (Fig. 4).
I2Oo
1 w Chemical products scrap w Bioenergy w SPS w PV and electrolysis w PV(with large capacity storage)
1000
i i 5
3
800
rn PV(with small capacity storage)
=-a. -
Y
<
PV(without storage) Wind and electrolysis Wind(with large capacity storage) &I Wind(with small capacity storage) w Wind(without storage) Hydro rn Nuclear Natural gas 00 11 w Unconv Oil w Coal
a
600
P (I
W
$
400
.-E k
200
0
2010
2020
2030
2040
2050
2060
Fig. 4.Primary energy supply (in the world in the sustainable energy scenario).
3.2. Electric Power Supply
In the world in 2060, the largest electric power source is photovoltaic type 4 at 455 EJ.year". However, photovoltaic type 4 that supplies not electricity but hydrogen is categorized not as power supply but as hydrogen supply. As photovoltaic type 4 is excluded, the largest source of electric power supply is photovoltaic type 3 (that is directly connected with electric power storage of large capacity) at 118 EJ.year-' (when lkWh is converted into 3,600 J). Subsequently, BGCC (biomass integrated gasifier combined cycle), hydropower, and photovoltaic type 2 (that is directly connected with electric power storage of small capacity) are at 52, 46, 43 EJ.year-', respectively (Fig. 5). However, each of photovoltaic type 1 (that is conventional photovoltaic) and the wind power types 1 to 4 is below 11 EJ.year-'. The amount of photovoltaic type 1 is restrained for the grid stability. The amount of wind power is restrained by the wind resource that is assumed at 28 EJ.year-' [4].
The Shape of Complete Renewable Energy Systems in the World
37
rn PV (with large capacity storage)
300
PV (with small capacity storage) PV (without storage) Wind and electrolysis Wind (with large capacity storage) rn Wind (with small capacity storage) rn Wind (without storage) rn Hydro power Space power system (SPS) Nuclear Coal IGCC with CO2 recovery rn Methanol power H2 power Natural gas power Oil power Coal power Chemical Drod. ScraD incineration Q
-
250
=j
200
L
m
w
w ._ b
-
.-0 p 150 i 100 50
0 2010
2020
2030
2040
2050
2060
I
Fig. 5. Power supply (in the world in the sustainable energy scenario).
3.3. Energy Trade Energy trade in 2060 in the sustainable energy scenario is explained (Fig. 6). On the overall trend, the low latitude regions (where photovoltaic and bioenergy are in good conditions) export energy to the high latitude regions. In the low latitude regions the conditions of photovoltaic and bioenergy are better than those in the high latitude regions. Concerning bioenergy, the main export regions will be Latin America and Oceania, and the main import regions will be the Middle East & North Africa, Westem Europe, and China (incl. Mongol, Vietnam, North Korea, etc.) Methanol will become the largest amount of energy of exports and imports. The total exports (that are equal to the total imports) will be 119 EJ.year-' in 2060. The main export regions are Middle East & North Africa and Sub-Sahara Africa, and the main import regions are North America, Western Europe, Japan, China, Latin America, Southeast Asia, and South Asia. Hydrogen will become the second largest amount of energy of export and import at 98 EJ.year-'. Hydrogen will be exported from Middle East & North Africa to Western Europe and Soviet Union & East Europe, and will be exported from Southeast Asia to China. Hydrogen will be produced mainly by photovoltaic type 4 (that is directly connected with electrolysis) in Middle East & North Africa and Southeast Asia where the incident light is in good conditions, and be exported to the high latitude regions by pipeline.
H. Yamamoto & K. Yamaji
38
IWestern Europe
North America
0Japan
0 Oceania
IChina etc.
Middle East and N-Africa
ISub-Sahara Africa
Latin America
IFSU and Eastern Europe
ISoutheast Asia
0South Asia Electricity Import Electricity Export H2 Import
H2 Export CH4 Import CH4 Export Methanol Import Methanol Export (Syn)Oil Import (Syn)Oil Export Coal Import Coal Export Bioenergy Import Bioenergy Export
0
I
I
I
I
20
40
60
80
100
120
140
Energy export and import(EJ/year) Fig. 6. Export and import of energy (in 2060 in the sustainable energy scenario).
3.4. Energy Systems Costs The energy system costs in 2060 in the world will be 7 and 18 trillion US$.year-' in the base scenario and the sustainable energy scenario, respectively. The latter cost is almost
The Shape of Complete Renewable Energy Systems in the World
39
three times as expensive as the former cost. We must recognize that the sustainable energy systems are the expensive energy systems. 4. Conclusions
In order to find the shape of the sustainable energy systems that are fossil-fuel-free and completely renewable, the authors developed a global land use and energy model (GLUE) including new technological concepts such as intermittent renewables (photovoltaic and wind) directly connected with storage of small capacity and large capacity. Using the model, the authors conducted simulations and obtained the following results. In the sustainable scenario (where they achieve zero fossils and complete renewables in 2060) the dominant resource of primary energy supply will be photovoltaic directly connected with electrolysis at the amount of 455 EJ.year-' in the world in 2060. The second and the third largest resources are biomass and photovoltaic directly connected with large capacity storage at 322 EJ.year-' and 118 EJ.year-', respectively. At that time, the dominant energy resource of the power generation connected with the electricity grid will be the photovoltaic with large capacity storage at 118 EJ.year-' in the world. BGCC (biomass gasifier combined cycle), hydropower, and photovoltaic directly connected with small capacity storage are above 40 EJ.year-' each in the world. However, the energy system costs in the sustainable energy scenario will be at 18 trillion US$.year-' and be almost three times as expensive as the costs in the base case (without constraints of fossil fuels or C 0 2 emissions) in the world in 2060. The sustainable energy systems are the expensive energy systems.
References [ l ] Yamamoto H, Yamaji K. (2004) Sustainable energy path. In Sustainable Development of Energy, Water, and Environment Systems. In N.H. Afgan et al. (eds), pp.23-27. A.A. Balkema Publishers. [2] Yamamoto H. (2003) A path toward sustainable energy systems evaluated by a global land use and energy optimization model, CRIEPI report Y02022, CRIEPI, Tokyo. [ 3 ] Yamamoto H. (2003) A role for renewables toward sustainable energy systems. In J. Gales & Y. Kaya (eds), Greenhouse Gas Control Technologies, Volume I, pp. 925-930. Pergamon. [4] Nakicenovic N. et al. (1993) Long-term strategies for mitigating global warming, Energy 18 (5): 401-609.
This page intentionally left blank
EXPERIMENTAL INVESTIGATION AND MODELLING OF THERMOELECTRIC GENERATORS FOR USE IN HYDROGEN BASED ENERGY SYSTEMS J. ERIKSEN('), R. GLOCKNER('),V.A. YARTYS('), B.T. HAFSTEINSSON(~),
T.I. SIGFUSSON ('I (I'
Institute for Energy Technologv (IFE), Kjeller, Norway, (2) Varmaraf ehj Reykjavik, Iceland ('IUniversity of Iceland (HI), Reykjavik, Iceland
The thermoelectric generator (TEG) represents a silent and reliable solid-state means of converting heat flow directly into electrical energy. The objective of the project is to investigate the potential for integration of TEGs in hydrogen based energy systems with focus on the concept development and evaluation of an integrated MH/TEG-unit for hydrogen energy storage purposes. In order to identify the potential for such MH/TEG-units, a TEG-model has been developed that is currently being verified based on laboratory experiments. Over the past several years hydrogen energy system experts at IFE have developed and managed a collection of hydrogen energy models (HYDROGEMS) that is increasingly being used to develop renewable-hydrogen (RE/H2) projects in countries all around the world. By interfacing the developed TEG-model with the HYDROGEMS library we will be able to virtually connect the TEG-technology to various other technologies, such as metal hydrides, electrolysers, fuel cells, photovoltaics, and wind energy conversion systems, in order to nodeling the design and control of such RE/TEG/Hz-systems. Based on the preliminary nodeling results and experimental data some key system parameters and design guidelines are appointed.
1. Introduction
The thermoelectric generator (TEG) represents a silent and reliable solid-state means of converting heat flow directly into electrical energy. Also being a powerful current source with rapid load following capabilities, high scalability, and the potential of low auxiliary power consumption, the TEG posses several beneficial properties especially in regard to hydrogen based stand-alone power systems (HSAPS). Traditionally, high-temperature energy sources have been utilised due to the inherent higher efficiency at high temperature differences. However, in recent years it has been realized that in situations where the supply of heat is cheap or free, as in the case of natural heat sources and various forms of waste heat, the TEG efficiency is not an overriding consideration. A number of hydrogen energy systems (HES) are currently being demonstrated around the world. Often, these systems associate low-level energy sources, such as solar heat and various forms of waste heat, which are candidates for increased system efficiency through thermoelectric conversion, Fig. 1. Especially, in the case of autonomous HES increased electrical energy efficiency might be of crucial importance. Furthermore, since the reverse operation of the TEG is also possible (i.e., converting electrical energy into a temperature gradient), the TEG might represent an interesting alternative in the heat balance of metal hydride (MH) storages.
41
J. Eriksen et al.
42
2. Objective The main objective of the project is to investigate the potential for integration of thermoelectric units in hydrogen based energy systems as first pointed out by Sigfusson in 2002 [l]. The work, which includes both experimental investigations and model simulations, focuses on the conversion of heat flow directly into electrical energy by means of TEG and the concept development and evaluation of an integrated MHITEGunit for hydrogenenergy storage purposes.
Fig. 1. Example of a HES with integrated TEGs for increased system efficiency.
The major expected outcome of the project will be an assessment of the potential for integrated MWTEG-units and design guidelines with regard to the heat integration of the two devices. Furthermore, the critical system parameters will be identified and, where possible, quantified. 3. System of Reference
In order to identify the potential for such integrated MH/TEG-units, a TEG-model has been developed that is currently being verified based on laboratory experiments and consequent modelling runs. The TEG of reference to this work, a C16 Thennator, was obtained from Varmaraf ehf [2] and integrated into an existing test-bench facility for R&D in hydrogen energy technology at IFE [3]. In addition to the TEG-unit, the testbench setup includes a MH-storage, photovoltaics, a proton exchange membrane (PEM) fuel cell (FC), secondary batteries, and an electronic load connected such that each component can be operated individually, at subsystem level, or at system level, Fig. 2. A detailed description of the MH/TEG interface (indicated by the dashed line in Fig. 2) is
Thermoelectric Generatorsfor Use in Hydrogen Based Energy Systems
43
given in Fig. 3, while the key figures for the TEG-unit and the MH-storage are summarised in Table 1 and Table 2, respectively.
0 ,
1
/
DPump
@Valve
/-Relay
*Heater
1 GND
Fig. 2. Schematic of the test-bench facility for R&D in hydrogen energy technology.
t
TEG bypass (hot)
-
water source
lot in
water source
:old in Cold out&
& Hand valve
a
MH bypass (hot)
I
Pump
Hot out&
MH*
:
1,
Hot water 1, out
-cyct;k
-
:
*) During MH discharge hot water is passed through the MH-storage, while cold water is passed through the MH-storage during MH charge.
Fig. 3. Schematic of the MH/TEG interface. Table 1. Specification for the TEG-unit.
Parameter Thermoelectric element Nominal power Nominal voltage Nominal current Typical fluid flow rates Weight
Value/type BiTe-based
Unit
100
W,I
12 8.3 6-9 25
Vdc Adc
Nl/min kg
Comment Bismuth-telluride @AT = 80°C @ AT = 80°C @ AT = 80°C
J. Erihen et al,
44
Table 2. Specification for the MH-storage.
Parameter Hydride alloy Nominal H2-capacity Hydride amount Volume, MH canister Volume, water canister Max. pressure, H2 Max. pressure, water Max. operating temperature
Valueltype TiZrMnVFeNi 1 5.5 2.5 0.75 10 6 80
Unit Nm3 kg 1 1 bar, bar, "C
Comment AB2-type hydride Approximate value Approximate value Approximate value Approximate value
4. Configurations and Applications Primarily, three different configurations for the integrated MH/TEG-system exists: 1) the heat source and the heat sink can be regarded as different sections of the MH-unit, 2) either the heat source or the heat sink can be considered as an external system with respect to the MH-storage, or 3) both the heat source and the heat sink are external systems with respect to the MH-storage (as for the configuration given in Fig. 3). While the first configuration probably is best suited for heat management and MHcompression by means of the Peltier effect, the two others have the potential for thermoelectric generation. A typical MWTEG-application based on the second configuration is to utilise the heat generated during charging of the MH-storage for power generation, as this heat should be removed from the storage unit anyway to prevent undesirable H2-pressure build-up. Also, the opposite configuration is possible since discharging of the MH-storage requires heating. Similar reasoning can be used when both the heat source and the heat sink are supplied externally as in the third configuration. Here, it is not the thermal energy due to the chemical reactions between hydrogen and the hydride alloy that is utilised for the power generation, but the thermal energy associated with the external sources. However, the main advantages with this configuration, compared to the configuration where the MH-storage is either the source or the sink, are: 1) for the power generation - a more powerful heat source can be utilised (e.g., natural heat sources or heat wasted by other components such as engines and furnaces), 2) for the heat balance of the MH-storage - a better controlled heat management can be maintained, and 3) for a laboratory setup - a better control with the energy flow in the system is achieved. Thus, the work performed in this project will focus on the configuration where both the heat source and the heat sink are supplied externally as shown in Fig. 3.
5. Modelling Over the past several years hydrogen energy system experts at Institute for Energy Technology (IFE) have developed and managed a collection of hydrogen energy models (HYDROGEMS) that is increasingly being used to develop renewable-hydrogen (RE/H2)
Thermoelectric Generatorsfor Use in Hydrogen Based Energy Systems
45
projects in countries all around the world, including Australia, USA, and Europe, and chosen as the preferred tool for evaluation of hydrogen demo projects in Annex 18 of the IEA Hydrogen programme [4,5]. By interfacing the TEG-model developed in this project with the already existing HYDROGEMS library, TEG technology can be virtually connected to various other technologies, such as MH-storages, PEM electrolysers (PEMELY) and fuel cells (PEMFC), alkaline fuel cells (AFC), photovoltaic (PV) generators, wind energy conversion systems (WECS), and diesel engine generator systems (DEGS), in order to optimize the design and control of such RE/TEG/H*-systems. 6. Theoretical Background
A TEG consists of several n- and p-type semiconductor pellets connected electrically in series and thermally in parallel sandwiched between two ceramic plates as illustrated in Fig. 4.
Fig. 4. Schematic of a thermoelectric module in power generation mode (TEG), with the temperatures thermal conductances (G), heat flow rates ( Q 7, and polarities indicated.
(0,
The bottom plate is bonded to a heat sink and, with the application of a heat source to the top surface of the thermoelectric module, electrical power is generated. Electrical power is derived from the movement of electrical carriers brought on by heat flow through the semiconductor pellets. Positive carriers (or holes), move to the heat sink side of the p type pellet making that junction electrically positive. Similarly, electron flow in the ntype pellets results in a net negative charge at the heat sink side of the n-type pellet. In this way, the current generating potentials of the n- andp-type pellets are series-added. 7. Method
The TEG modelling is based on a 1D model of thermoelectric coupling where the semiconductor elements are of homogeneous structure and insulated, both electrically and thermally, from their surroundings except at the junction contacts. This is a common approach to the problem [e.g., 2,6,7, 81.
46
J Eriksen et al.
Two different generic TEG-models have been developed that will be referred to as Model #1 and Model #2 below. While Model #1 models the TEG performance based on Fourier’s heat flow, the Seebeck effect, and Joule’s heat, Model #2 also accounts for the Thomson effect. The main challenge with including the Thomson effect is that, in order to achieve an accurate result, the TEG-hardware needs to be specified in great detail. Thus, whenever such detailed information is not available, the faster and less demanding Model #1 should be used. In order to develop realistic TEG-models featuring a high degree of usability, great effort has been made to estimate the junction temperatures Thj and Tcj of the thermoelectric elements (Fig. 4) as accurate as possible based on the inlet and outlet temperatures of the hot and cold fluids that are readily available for most systems. Here, the thermal conductances of the heat exchangers are of special importance. Both the TEG-models developed at IFE, which are intended for simulation of integrated MH/TEGunits, uses the lumped thermal conductances GI (hot side) and G2 (cold side) in order to model the irreversibility caused by finite rate heat transfer, i.e., the temperature differences (Th-Thj) and (Tcj-Tc).Assuming that the thermal resistance of the heat exchangers due to convection dominates the thermal resistance due to conduction (assumed constant over the whole contact surface area), the lumped thermal conductance method is generally valid if the temperature of the body can be described or approximated by a single temperature. In this case, this validity condition is met quite well using the mean temperature of the circulating hot and cold fluids. 8. Main Equations
Then, under steady-state conditions, the electrical power output (P)and the efficiency ( q ) for a thermoelectric generator composed of y2 identical thermocouples can be expressed by (Model #l):
where aj is the Seebeck coefficient for the junction at the temperature of interest, R = R, + Rp and K = K, + Kp are the total electrical resistance and thermal conductance of the ntype andp-type elements, respectively, and the junction temperatures ThJand Tcjare given by:
Thermoelectric Generatorsfor Use in Hydrogen Based Energy Systems
47
The equivalent equations for Model #2 are:
where z = Z, + zp is the total Thomson coefficient of the n-type andp-type thermoelectric elements and the superscript star (*) indicate the parameter accounts for the Thomson effect according to:
wn =- z n I Kn Ln
, cop=- zPI
KPLP
Here, the junction temperatures Thjand Tcj are given by:
J. Eriksen et al.
48
9. Laboratory Experiments
In order to investigate the TEG hardware and then, in the next phase, to verify the software models several experiments were performed. The key measurement parameters and control signals associated with these experiments are listed in Table 3 and depicted schematically in Fig. 5. Typically, the fluid flow rates and inlet temperatures would be kept constant within a single experiment, while the TEG output current I ~ E G were operated in a steps of 0.5A drawing a constant current over a 1-3 minute period and measurements performed at 10-second intervals. Table 3. The key measurements and control signals associated with the TEG experiments.
Parameter Hot fluid, inlet temperature
Hot fluid, outlet temperature Cold fluid, inlet temperature Cold fluid, outlet temperature Hot fluid, volume flow rate Cold fluid, volume flow rate TEG terminal voltage TEG output current
Symbol
Unit "C
Measurement
Thot,out
"C
Tco/din
"C
Tco/+,,
"C
X X X X X X X
Thot,in
V'hot
1 min-'
v'co/d
1 mid'
UTEG ITEG
V A
X
Control X
X
X X X
I
Fig. 5 . Schematic of the TEG-unit, with the key measurement parameters and control signals indicated.
49
Thermoelectric Generatorsfor Use in Hydrogen Based Energy Systems
As a result of the averaged measured data, the current-voltage (N) and current-power (IP)characteristics of the TEG-unit at different temperature differences AT, and fluid flow rates were found. In Fig. 6, a comparison between experimental data and Model #1 at two different moderate temperature differences across the thermoelectric elements is given. As seen from the figure, the generic model is in fairly good agreement with the measured data. However, the results also show that that the deviation between the measurements and the model is at a maximum around the maximum power point, which's location might be the main reason for doing the model simulation. Thus, at this time, the model is seen too inaccurate to give the desired effect in an integrated MH/TEG study. However, consequent modelling runs clearly indicate the necessity from accurate temperature and flow rate measurements since both metal hydrides and the TEG-model is very sensitive to these parameters. The current version of Model #2 and the specification of the TEG-unit do not allow for a fair comparison with the measured data, and, thus, is avoided.
I
201
'
I
'
I
'
I
'
I
'
I
'
I
18
S
A 70 -
{ 60
16 14
4 50
12
4 40
10
a
30
6
{ 20
E n
4
110
2 0 0
2
4
6
8
10
12
-30 14
~ T E G[A1
Fig. 6. IU-and IP-characteristic of the TEG-unit at two different temperature differences across the thermoelectric elements (Run #1: AT,, = 29.6 K, Run #2: AT,e = 40.4 K). Comparison between Model #1 (thin lines - voltage, tick lines -power) and measured data (squares -voltage, cross -power).
10. Conclusions
Three different configurations for the integrated MH/TEG-system have been identified, of which two have been recognised as potential thermoelectric generator configurations; a) MWTEG-systems where either the heat source or the heat sink can be considered as an external system with respect to the MH-storage and b) both the heat source and the heat sink are external systems with respect to the MH-storage, where the latter has been implemented and tested at IFE as this configuration was found to suite a laboratory setup and the modelling task the best.
50
J. Eriksen et al.
A set of modelling equations for realistic simulation of thermoelectric generators based on readily available system parameters in order to promote a high degree of usability have been identified. One of the generic models based on these equations (Model #l), which models the TEG performance based on Fourier’s heat flow, the Seebeck effect, and Joule’s heat, is in fairly good agreement with the experimental data. Due to a deviation between the measured data and the model, especially close to the maximum power point, the model is seen too inaccurate to give the desired effect in an integrated MH/TEG study at this time. Nevertheless, the necessity from accurate temperature and flow rate measurements has been addressed since both metal hydrides and the TEG-model has shown to be very sensitive to these parameters. Also, the configuration and modelling of the heat exchangers has shown to be of key importance for the TEG performance. That is, even now critical system parameters and some main design guidelines have been appointed.
Acknowledgments This material is based upon work supported by Nordic Energy Research (NEFP) under the NORSTORE project 46-02 ”Integration of Advanced H Storage Materials and Systems into the Hydrogen Society”.
Nomenclature G I K L n P
Q’ R T U V’
thermal conductance of the heat exchangers, WK-’ electrical current, A thermal conductance of the thermocouples, WK-’ length of the thermoelectric elements, m number of thermocouples, electrical power, W heat flow rate, W electrical resistance, R temperature, K voltage, V volume flow rate of circulating hot and cold fluids, 1s.’
Greek symbols
a 7 z
Seebeck coefficient of the thermoelectric elements, VK-’ thermal efficiency of the thermoelectric generator, Thomson coefficient of the thermoelectric elements, VK-’
Subscripts c, coldcold side h, hot hot side in inlet
Thermoelectric Generatorsfor Use in Hydrogen Based Energy Systems
51
junction load negatively doped (n-type) thermoelectric element n out outlet positively doped (p-type) thermoelectric element p te thermoelectric element TEG thermoelectric generator
j
L
Superscripts
*
indicate the parameter accounts for the Thomson effect
References [ 11 Sigfusson TI. (2002) The physics of thermoelectricity. Proc IOIh Physics in Iceland
[2]
[3]
[4]
[5] [6]
[7] [S]
Conference, Reykjavik, Iceland, 203-2 13; Ed. Ari Olafsson, publication by the Icelandic Physical Society. Hafsteinsson BT. (2003) Performance analysis of a thermoelectric heat exchanger generator system. M.Sc. thesis (in Icelandic), University of Iceland, Reykjavik, Iceland. Eriksen J. (2003) Experimental design of a data acquisition and control system for a small-scale PV-H2 system. Ph.D. thesis, University of Oslo, Norway. Ulleberg 0. (1998) Stand-alone power systems for the future: Optimal design, operation & control of solar-hydrogen energy systems. Ph.D. thesis, Norwegian University of Science and Technology, Trondheim, Norway. The HYDROGEMS homepage is found at www.hydrogems.no. Cobble MH. (1995) Calculation of generator performance. In D.M. Rowe (ed.), CRC Handbook of Thermoelectrics, CRC Press, Boca Raton, Florida. Seifert W, Ueltzen M, Muller E. (2002) One-dimensional modelling of thermoelectric cooling. Phys. Stat. Sol. (a) 194 (1): 277-290. Chen L, Gong J, Sun F, Wu C. (2002) Effect of heat transfer on the performance of thermoelectric generators. Int. J. Therm. Sci. 41: 95-99.
This page intentionally left blank
INTELLIGENT OPERATION MANAGEMENT OF FUEL CELLS AND MICRO-TURBINES USING GENETIC ALGORITHMS AND NEURAL NETWORKS A.M. AZMY , I. ERLICH Institute of Electrical Power System Engineering and Automation, University of Duisburg-Essen, 0-47057, Duisburg, Germany This paper demonstrates a new-two-stage intelligent technique to manage the operation of Distributed Generating “DG’ units for residential utilization. In the first stage of the optimization process, a Genetic Algorithm “GA” is used to define the optimal settings of DG units depending on detailed economic models. For online applications and to avoid the repetitive time-consuming optimization process, the procedure is generalized in the second stage using an Artificial Neural Network “AN”’. The objective is to develop an intelligent management tool, which can be used in the online mode and depends only on the parameters obtained from the structure of the ANN. Variations of load demands and operating tariffs can easily be simulated online as they represent the main inputs of the ANN. The first stage of the management process is applied alternatively to a single fuel cell unit, three fuel cell units operating in parallel and a micro-turbine unit. However, the ANN generalization process is applied only with the single fuel cell unit as it shows the most economic operation regarding the operating costs. The results obtained in this research encourage the use of this technique in order to achieve a simple, fast and effective online management of DG units for residential applications.
1. Introduction With the increasing demand on electrical energy, DG technology can offer important support to the conventional centralized power sources [l]. Therefore, DG is predicted to play a significant role in the electric power system in coming years [2, 31. The DG, in general, can be understood as the integrated or stand-alone utilization of any generation near consumer’s load terminals [4]. DG technology can provide significant benefits for both consumers and electric distribution utility [5, 61. This includes improving availability and reliability of power supply system, voltage support, improved power quality and postponing or avoiding transmission and distribution investments. Power loss reduction, possibility of cogeneration and emission reduction represent also additional advantages of utilizing DG units within distribution network [7]. Fuel cells and micro-turbines are candidates as DG units to be utilized either integrated into distribution systems or in the stand-alone mode [7-91. Also, a hybrid configuration comprising the two units, which provides relatively high efficiency, is possible after solving some related technical difficulties [9]. One of the important applications of DG units, where fuel cells and micro-turbines are particularly suitable, is the utilization of small-modular commercial or residential units for onsite service. In this case, the capacity of the DG unit can be chosen to cover most of the load demand most of the time, where the surplushhortage is exported to or imported from the main grid system. Therefore, the operation of the DG unit has to be properly managed, considering both electrical and thermal power, to reduce the operating cost to the minimum level. This reduction in the operating cost can significantly contribute to decrease the total energy price and hence, improving the economic feasibility of these units. 53
54
A.M. A m y & I . Erlich
In this paper, the operation of Proton Exchange Membrane “PEM’ fuel cells and micro-turbines with a residential load are managed using a new two-stage intelligent approach. This requires the development of suitable economic models to describe the daily operating cost of the selected units. Moreover, a robust optimization tool, which can deal with the nature of the problem, has to be applied. In the first stage, GAS are applied to define the optimal daily performance of the DG units under different operating conditions. This process is applied for three alternative scenarios: utilizing a single fuel cell, using three fuel cell units in parallel and utilizing a micro-turbine unit. In spite of the significant reduction in operating costs, the results from this stage show a strong impact of the fuel and electricity tariffs on the optimal settings of the units. In addition, the optimal settings depend on the load demand, which necessitates carrying out new optimizations after each change in the operating tariffs and load demands. To avoid repeating the optimization process and to enable online updating of the operating parameters, a second stage is applied to the management process based on ANN generalization capability. In this stage, the ANN is trained and tested using database extracted from the first stage. This is carried out only with one fuel cell unit connected to the residential load as it shows the most economic operation among the three investigated cases regarding the operating cost. The ANN, which is trained and tested offline, succeeded to recognize and re-simulate the optimal behaviour of the fuel cell. The well-trained ANN can then be used onsite in the online mode. To simulate the variations in the operating conditions, fuel costs and load demands, as inputs to the ANN, are modified online. The results obtained in this research encourage the implementation of this approach with different DG units to achieve both fast adaptation and optimal operation with the commercial and residential applications.
2. Economic Models of the Selected DG Units Fig. 1 shows the structure of the domestic system including both the electrical and the thermal energy paths. The electrical and thermal demands of the load are supplied mainly by the DG unit(s). However, the shortage in electricity and the surplus electrical power can be covered from or sold back to the main grid system at different tariffs. Two energy meters can separately measure the purchased and the sold electricity f r o d t o the grid system depending on their tariffs. The thermal energy produced in the DG source(s) is utilized for water and space heating of the residential building. The load is provided also by natural gas to compensate any possible deficiencies in the produced thermal energy. Most suppliers offer several natural-gas tariffs depending on the field of application (e.g. residential, industrial, electricity generation...etc). The consumptions of natural gas are measured independently for DG units and residential load to calculate the cost of each part depending on its tariff. With the abovementioned structure, the daily operating cost “DOC ($)”, which has to be minimized, can be developed in terms of payments (for natural gas and purchased electricity) and incomes (for sold electricity) in the following form:
Intelligent Operation Management of Fuel Cells and Micro-Turbines
DOC = DFC + DCPF + DCPE - DISE + 0 & M + STC
55
(1)
M,: gas flow meter for the residential load Fig. 1. Structure of the residential system supplied by DG unit(s)
The daily fuel cost “DFC $” to supply DG units(s), daily cost of purchased fuel “DCPF $” for residential load, daily cost of purchased electricity “DCPE $”, and daily income for sold electricity “DISE $” are described by the following equations:
where: C,, Cz : Fuel price to supply DG and residential load respectively ($/kWh) C3,C4 : Tariffs of purchased and sold electricity respectively ($/kWh) T : Time duration between two successive settings of the DG units (h) PJ :.Net electrical power produced at interval J (kW) Pa :.Power required for auxiliary devices (kW) qr :.DG efficiency at interval J Lth,j :.Thermal demand at interval J (kW) Pth,J :.Thermal power produced at interval J (kW Lel,j : Electrical demand at interval J (kW)
56
A.M. Azmy & I . Erlich
The operating and maintenance cost “O&M’ is assumed as a constant value per kwh, while the start-up cost “STC $”depends on the temperature of the unit and hence on the time terminated before start up and is given as:
where: a :Hot start up cost and a + j3 represent the cold start up cost
toff
:The time duration, where the unit is off (h) :The DG unit cooling time constant (h) This objective function is applicable for a single fuel cell and a single microturbine by choosing adequate parameters. Considering the case of three fuel cells operating in parallel, the daily fuel cost is calculated depending on the fuel consumption of each unit individually, taking into account its efficiency depending on its operating point. On the other hand, the cost of purchased electricity, income for sold electricity and cost of purchased gas for residential applications are calculated depending on the accumulated electrical and thermal power from the three units together. Typical efficiency curves for fuel cells and micro-turbine as well as typical relations between electrical and thermal powers produced in the units are used in the economic models. Nonlinear functions are developed to identify these relations depending on the supplied electrical power. Fig. 2 shows the efficiency curves and the relation between the thermal and electrical powers in the fuel cell unit as used in the economic model. z
Electrical output power (kW)
Electrical output power (kW)
Fig. 2. Efficiency curves and the thermal power in the fuel cell unit.
Intelligent Operation Management of Fuel Cells and Micro-Turbines
57
In the given objective function, there are four tariffs affecting the setting points i.e. C1, C2, C3 and C4. These four tariffs represent the four decision variables, which affect the optimal settings of the energy source(s). The minimization of the objective hnction (1) is restricted by many operational and technical constraints. This includes the unit capacity constraints, unit ramp rate constraints, minimum up/down time limits (continuous running/stop time constraint) and the maximum number of starts and stops per day. The mathematical description of these constraints is given in detail in a previous paper [ 101.
3. GA-Based Optimization Process Since the presented economic models of the DG units with domestic loads are discontinuous and nonlinear in nature, the GA will be a convenient choice as it represents a powerful probabilistic search algorithm taking into account its capability of searching in a population of points in parallel [ 1 1- 131. The multi-population structure is chosen, where individuals migrate periodically between subpopulations to transfer information between them. To handle the constraints in the economic models, the penalty-function method is used, whereby the constrained problem is converted to an unconstrained one by augmenting the main cost function with additional cost terms. The additional terms assign nonlinear costs for solutions that violate any of the constraints depending on their locations relative to the feasibility boundary. The evolution process begins with initiating the population by formulating a number of individuals, which represent the possible output electrical power from the DG unit(s) over one day. The maximum values of these individuals are limited to 4 in the case of single fuel cell or micro-turbine and 1.3 in the case of three fuel cells operating in parallel. This is equivalent to a maximum electrical power of 4kW and 1.3kW respectively. For one day, 96 setting values have to be calculated as the setting of the unit is assumed to be updated every 15 minutes. The individuals are evaluated depending on the operating cost in addition to the penalty terms. Then, the individuals are ranked and suitable fitness values are assigned to them. Strings with higher fitness values are selected using the roulette wheel technique and then the recombination process is performed. Using the two well-known recombination processes, i.e. crossover and mutation, a new generation is produced. Some of the fittest members of each generation are saved and copied into the next one to ensure that best solutions are not lost when moving from one generation to the next. Some modifications are introduced to the GA-optimization program to carry out the management of the three fuel cell units simultaneously. Now, each individual comprises 288 unknowns, with 96 unknowns belonging to each unit. Two times during the evolution process, the individuals are divided into three sub-individuals and the calculations are carried out considering each sub-individual separately. The first one is when the daily operating cost of each individual fuel cell is calculated. In this case, the computation has to be carried out for each unit according to its efficiency and the additional penalty costs, which depend on the operating points. The total cost is then
A.M. Azmy & I . Erlich
58
calculated by adding the cost of the three units together, since the total power is required for further calculations rather than the power from individual units. The second time, where individuals are divided into three sub-individuals, is when the new offspring is create as the crossover and the mutation have to be applied to each unit separately.
4. Results of the Optimization Process The GA-based optimization process is carried out to manage the performance of a PEM fuel cell at different electricity and fuel prices as well as at various daily load demands. More tan 540 cases are considered including ten typical load curves corresponding to different seasons and realistic tariffs. Fig. 3 shows the optimal settings of the fuel cell for a certain load curve with three different tariffs of sold electricity (C4). This involves the case where no electricity is sold back to the utility. The other 3 tariffs, i.e. C1, C2 and C3, are held constant in the three cases. The strong variation of the optimal settings with the change of this tariff is obvious. The other three tariffs have also similar strong impact on the optimal performance of the fuel cell. This necessitates repeating the optimization if any of the operating tariffs is changed, which represents a real challenge and requires advanced knowledge and experience from the operator. Electrical power (kW)
Thermal power (kW)
~
Load demand
a
6 0
Time (h) *0.0 $ k W h - -
- -0.07 $ k W h
- -. -0.1 $ k W h '
Fig. 3. Effect of varying the sold electricity tariff on the optimal settings of the fuel cell.
It is noticeable from the results that the fuel cell generates low power for prolonged periods within the day, which is also the case in most of the investigated cases. It supplies the rated or near rated power only for short time. Hence, it is necessary to answer the question whether the utilization of smaller identical units with equivalent total capacity can be more economic regarding the operating costs. In this case, one of the identical units is used as a base source and the other units are added as required.
59
Intelligent Operation Management of Fuel Cells and Micro-Turbines
The optimization process is carried out again using the same load curves and tariffs as with one fuel cell unit to manage the performance of three fuel cell units simultaneously. Figs. 4 and 5 show the optimum output electrical power from the three fuel cells as well as the total electrical and thermal power for two different cases. For comparison purposes, the optimum electrical and thermal powers from one fuel cell under the same conditions are also illustrated in the figures. Table 1 summarizes the used tariffs in the two optimization processes as well as the total operating costs when the load is supplied by one and three fuel cell units. The total electrical and thermal output powers from the three units together are similar to that obtained from a single fuel cell. The contributions from the three units vary depending on load curves and operating tariffs. In some cases, one or two units are not Table 1. Operating tariffs and daily costs when optimizing the operation of a single fuel cell and three units operating in parallel.
case (1) case (2)
0.03 0.03
0.07 0.09
0.16 0.16
0.1 0.0
1.6032 3.6551
1.915 4.5912
Electrical uower of unit 1 (kW)
1 T .
'I
0.5
\
,
I
I
I
Electrical power of unit 2 (kW)
1.5
I
_.I
Electrical power of unit 3 (kW)
I
1.5
0.5 I
I I
I
Total electrical power (kW)
2
I
10
0
I
I
I
Total thermal power (kW)
-3 units I
0
4
I unit]
I
8
12
Time (h)
16
20
24
Fig. 4. Optimal settings of one and three fuel cells to supply a residential load: case (1).
used for the whole day. In other cases, one or two units operate only for a short period during the day as shown in Fig. 5 . Generally, the total operating cost using three units is
A.M. Azmy & I . Erlich
60
more expensive than utilizing only one unit. The operation of one or more units for a short time increases the operating cost as a result of the start-up cost. In addition, fuel cells exhibit lower efficiency at higher operating power as shown in Fig. 2. Since the three units are operating at relatively higher percent power, the resultant efficiency of each unit is lower than that of a single fuel cell. As an alternative DG source, the management process is carried out for a microturbine to optimize its performance with the same strategy, In spite of the similarity of the economic models of the fuel cell and the micro-turbine, the dissimilarity between the parameters of the two units as well as the type of efficiency and thermal power curves result in significant changes in the optimal operation of the two units. Fig. 6 shows the efficiency curve and the relation between thermal and electrical powers in the microturbine as used in the economic model. Compared to the curves of the fuel cell shown in Fig. 2, a considerable difference can be observed regarding both the nature and magnitudes of the efficiency and the thermal power.
2
Electrical power of unit 1
1
1-
0
Electrical power of unit 2
I
Electrical power of unit 3
21
I
1I
I
I
Total electrical power
5-
- 3 units
0
ma
0 .
I
I
r-0.0. I
To evaluate the performance of the micro-turbine compared to that of the fuel cell, two cases from the optimization process are given, where the operating tariffs and the total operating cost are given in Table 2. Figs. 7 and 8 illustrate the electrical and the thermal output power from the micro-turbine for the two cases. The load demand and the optimal output power from a single fuel cell unit are also illustrated in the same figures.
Intelligent Operation Management of Fuel Cells and Micro-Turbines
61
Fig. 6. The efficiency curve and the thermal power of the micro-turbine unit. Table 2. Operating tariffs and daily costs usinga fuel cell unit and a micro-turbine unit.
case (1) case (2)
Gel+
CIl1
C"2
C4 -p
($kWh) 0.03 0.04
($kwh) 0.07 0.05
($kwh) 0.14 0.12
($km)
12
16
0.1 0.07
Total operating cost ($/day) Fuel cell Micro-turbine 1.2412 2.8997 3.6942 4.3213
Electrical power (kW)
0
4
Load demand
8
Time (h)
- - - - micro-turbine output
s
20
fuel cell output
Fig. 7. Optimal settings of a micro-turbine unit and a fuel cell unit: case (1).
24
A.M. Azmy & I . Erlich
62 Electrical power (kW)
Thermal power (kW) 12
A
10
Time (h)
Load demand
- - - - micro-turbineoutput * * * * * fuel cell output
Fig. 8. Optimal settings of a micro-turbine unit and a fuel cell unit: case (2).
In most investigated cases, the settings of the micro-turbine are mainly affected by the thermal demand. In Fig. 7, the fuel cell produces a high value of electrical power due to the high tariff of sold-electricity and the low tariff of purchased fuel, while the outputs from the micro-turbine are close to the thermal load demand. This is due to the high thermal power generated in the micro-turbine compared to the electrical power as seen in Fig. 6. Covering the electrical load demand results in excess thermal energy, which would be wasted. The higher electrical efficiency of the fuel cell causes lower operating cost in most investigated cases. This is due to the high price of electrical energy compared to that of the thermal energy. For loads with high thermal and low electrical demands, microturbine units may be favourable in terms of operating costs. To evaluate the potential reduction in the total daily cost when this approach is applied, the results of optimizing the fuel cell, as the most economic choice among the three cases, are compared with three conventional settings. The first one is to operate the unit at its rated power. The second and third cases are to track the electrical and thermal load demand respectively. Table 3 gives the average cost as well as the average difference of the three conventional settings with respect to the GA-based optimal case for one load curve under 8 1 different operating conditions. Table 3. Cost savings by optimization the operation of the fuel cell.
Optimal settings (from GA) Settings=rating Settings=electricaldemand Settings=thermaldemand
Average cost ($/day) 3.122 8.582 4.855 4.631
Average difference with respect to the optimal case Average difference ($/day) 1 Average percentage difference 0.0 0.0 4.860 130.575 % 1.133 30.441 Yo 24.584 % 0.915
Intelligent Operation Management of Fuel Cells and Micro-Turbines
63
5. Management Generalization Using ANN The results from the first stage of the optimization process showed the possibility of reducing the operating cost considerably by optimizing the power generated in the DG units. However, the need for new optimization after each change in the load demand or the operating tariffs restricts the online application of this approach. In the second stage of the management process, an ANN is used to generalize the results obtained in the first stage. The ANN has high capability of generalizing such nonlinear complicated problems [13-151. This is carried out only for the single fuel cell as it is the most economic choice as explained earlier. After training and testing the ANN, it can be used onsite for the online application. The ANN comprises three hidden layers, with 40, 30, and 20 neurons. The tansigmoid transfer function is chosen for all neurons in the hidden layers, while the logsigmoid transfer function is used for the output neuron. 54 inputs including the four operating tariffs and present electrical and thermal demands are used. In addition, historical and forecast powers for three hours are introduced at the input layer. As it is assumed that the setting point will be updated 4 times each hour, 12 previous values and 12 prognoses of both the electrical and thermal load demands are introduced at the input layer. The single output represents the desired optimal electrical-power of the fuel cell in the next time step. The ANN is trained offline using more than 56000 patterns and is then tested using new load curves as well as new operating tariffs. Figs. 9 and 10 compare between the GA-based optimal targets and the actual output from the ANN in two cases. The operating tariffs and the corresponding daily costs depending on both GA-based optimized settings and the ANN output are given in Table 4. Table 4. Comparison between daily costs using GA-based optimal settings and ANN outputs
case (1) case (2)
0.03 0.04
0.05 0.09
0.12 0.16
3.4734 3.0473
0.0 0.1
3.6371 3.1723
Thermal power 4. 3'
0
4
8
12
Time (h)
16
20
240
4
8
12
Time
16
20
Fig. 9. A comparison between GA-based optimal target and ANN output: case (1).
24
A.M Azmy & L Erlich
64
5
"0
4
8
12
Time (h) -
16
20
ANN outputs
4
24"O ea
0
e
*
8
12
Time
16
20
24
GA-based optinlal target
Fig. 10. A comparisonbetween GA-based optimal target and ANN OUQUS: case (2).
The c o m p ~ s o n sgive an in~cationabout the agreement between the outputs from the ANN and the optimal settings. Hence, it is expected that defming the settings d e ~ e n d i n ~ on ANN decision will not lead to a significant increase in the operating cost compared to the optimal case. To emphasize this fact, a comparison between the daily operating cost using GA-based o p ~ z a ~ and o n ANN outputs is carried out for 78 different cases at various operating tariffs and the results are illustrated in Fig. Z 1. The average daily cost using the GA-based optimal settings is about 3.689 $/day for the 78 investigated cases. Using the quasi-optimal settings, which are defined by the ANN, increases the daily operating cost to 3.869$/&y. Cornpared to the reduction achieved using the proposed technique, this difference represents a minor increase in the daily operating cost. These results reflect the success of the ANN to capture the optimal behaviour of the unit.
Investigated case Fig. 11. Daily operating cost using optimal settings and applying ANN.
Intelligent Operation Management of Fuel Cells and Micro-Turbines
65
6. Conclusions This paper deals with the optimal management of electrical and thermal power in DG units when used to supply residential loads. The investigation involves the management of a single fuel cell, three fuel cells operating in parallel and a single micro-turbine unit. Analysis of the obtained results revel a significant reduction achieved in the daily operating costs using the management process, which contributes in improving the economic feasibility of DG units. Supplying the residential load using a single fuel cell unit resulted in lower daily operating cost compared to the other two cases. The high electrical efficiency of the fuel cell results in lower operating cost compared to the use of a micro-turbine. Also, the nature of the fuel cell efficiency curve, which decreases with the increase of the supplied power, shifts the optimum allocation in favour of using a single unit rather than using smaller identical units operating in parallel. The paper proposes also formulating the management process in a general frame using ANN to avoid the need for repetitive optimization after changes in operating conditions take place. The ANN is trained and tested using data-base extracted from the GA-based optimization for different load curves and operating tariffs. The inputs of the ANN, which represent the load demand and the operating tariffs, can be updated online to simulate the variations in the operating conditions. The effectiveness of the suggested approach is confirmed by the agreement between the optimized settings and the outputs from the ANN. The online adjustment of the he1 cell settings in a fast and simple manner demonstrates the viability of this approach for optimum deployment of different DG units for residential applications.
References [I] Hoff TE, Wenger HJ, Farmer BK. (1996) Distributed generation: an alternative to electric utility investments in system capacity. Energy Policy 24 (2): 137-147. [2] Barker PP, De Mello RW. (2000) Determining the impact of distributed generation on power systems: Part I. Radial distribution systems. Power Engineering Society Summer Meeting 2000 IEEE, 16-20 July, 3: 1645-1656. [3] Barker PP, Johnson BK. (2002) Power system modeling requirements for rotating machine interfaced distributed resources. Power Engineering Society Summer Meeting, 2002 IEEE, 21-25 July, 1: 161-166. [4] Daly PA, Morrison J. (200 1) Understanding the potential benefits of distributed generation on power delivery systems. Rural Electric Power Conference 200 1, 29 April-1 May, pp A2-1 - A2-13. [5] Delfmo B. (2002) Modelling of the integration of distributed generation into the electrical system. Power Engineering Society Summer Meeting, 2002 IEEE, 21-25 July, 1:170-175. [6] McDermott TE, Dugan RC. (2002) Distributed generation impact on reliability and power quality indices. Rural Electric Power Conference 2002 IEEE, 5-7 May, pp: D3-D3-7.
66
A.M. Azmy & I . Erlich
[7] Azmy AM, Erlich I. (2003) Dynamic simulation of fuel cells and micro-turbines integrated with a multi-machine network. The 2003 IEEE Bologna Power Tech Conference, June 23-26, Bologna, Italy. [8] Ordubadi F. (2001) PEM fuel cells and future opportunities. Power Engineering Society Summer Meeting, 2001 IEEE, 15-19 July, 1: 710-716. [9] Azmy AM, Erlich I. (2003) Dynamic simulation of hybrid fuel cell/micro-turbine units integrated into large power systems. The Ninth International Middle-East Power Systems Conference (MEPCON), Egypt, 16-18 December. [lo] Azmy AM, Erlich I. (2003) Management of PEM fuel cells for residential applications using a real coded genetic algorithm. The Ninth International MiddleEast Power Systems Conference (MEPCON), Egypt, 16-18 December. [ 113 Damousis IG, Bakirtzis AG, Dokopoulos PS. (2003) Network-constrained economic dispatch using real-coded genetic algorithm. IEEE Transactions on Power Systems, February, 18 (1). [12] Ling SH, Lam HK, Leung FHF, Lee YS. (2003) A genetic algorithm based fuzzytuned neural network. Fuzzy Systems, 2003, FUZZ ‘03, the 12th IEEE International Conference, 25-28 May, 1 : 220-225. [13] Chow TT, Lin Z, Song CL, Zhang GQ. (2001) Applying neural network and genetic algorithm in Chiller system optimization. Seventh International IBPSA Conference, 13-15 August, Rio de Janeiro, Brazil. [I41 Chen LH, Chiang CH. (2002) An intelligent control system based on multiobjective genetic algorithms and fuzzy-neural network. Systems, Man and Cybernetics, 2002 IEEE International Conference, 6-9 October, 3: 6. [ 151 Charytoniuk W, Chen MS. (2000) Very short-term load forecasting using artificial neural network. IEEE transactions on Power Systems, February, 15 (1).
MUNICIPAL SOLID WASTE VALORISATION AS ENERGY FOR ISOLATED COMMUNITIES G. TAVARES (I), (I)
z. ZSIGRAIOVA(*),v. SEMIAO"), M.G. CARVALHO(')
IST - Instituto Superior Tecnico, Lisbon, Portugal
Technical University of Kosice, Faculty of Metallurgy, Kosice, Slovakia, presently at IDMEC,
IST, Lisbon The amount of generated municipal solid waste (MSW), which is a result of human activities, is continuously increasing and following trends of economy and population changes. Sustainable waste management requires a good integrated waste management system and government policies that encourage waste prevention, reuse and both materials and thermal recycling. Eventually, landfills will only be used for stabilised materials. Suitable legislation is necessary with incentives and raise public awareness, to maximise the valorisation of waste materials. Municipal solid waste incineration represents a considerable potential for reduction in reactivity of any residual landfilled material on one side and potential energy source on another. Therefore, incineration as a way of municipal solid waste valorisation has a high importance in resources recovery within the integrated waste management. Power generation from waste could be a significant contribution to the reduction in overall emissions and an alternative to displace energy produced from fossil fuels. Therefore, a demand for increasing amounts of MSW to incinerate is expectable. Heat recovery, efficient flue gas cleaning systems and significant greenhouse gas abatement capacity makes municipal solid waste incineration an attractive environmentally sound option. In the present paper the authors focus on the potential of MSW incineration and its implementation to energy systems for isolated communities such as islands. Islands have several particular characteristics that considerably influence their energy and sustainable waste management options, which will be addressed. Based on a selected case, it will he demonstrated what are the possible ways of implementation and perspectives for municipal solid waste treatment and exploitation.
1. Introduction
Production of MSW is an inevitable consequence of human activities. The generated amount is continuously increasing following trends in development of economies and population. Sustainable waste management requires a good integrated waste management system and government policies that encourage waste prevention, reuse and both materials and thermal recycling. Eventually, landfills will only be used for stabilised materials. MSW incineration represents a considerable potential for reduction in reactivity of any residual landfilled material on one side and potential energy source on another. Power generated from waste could be a significant contribution to the reduction in overall emissions and an alternative to displace energy produced from fossil fuels. In the present paper the authors focus on the potential of MSW incineration and its implementation to energy system for isolated communities such as islands. Island communities have several particular characteristics that considerably influence their energy and sustainable waste management options. 2. Integrated Management of MSW
The direction of the industrial development is towards integration of available waste treatment technologies aiming to obtain a waste management environmentally and economically sustainable. 67
68
G. Tavares et al.
However, looking at the situation in European Union (EU), waste strategy goals have not been reached: waste prevention measures have not stabilised waste production, and landfilling is still the most common treatment method for more than 70% of the total collected waste, despite significant progress in recovery and recycling [ 11, [2]. An Integrated System of MSW Management should comprise: Prevention (to avoid, eliminate or reduce a waste at its source); Reuse; Recycling; Thermal treatment (with or without energy recovery); 0 Disposal to landfill. However, this hierarchy, organised according to the importance given by EU, is too simple to be generally applicable to more varied and complicated cases, as it does not always correspond to a more adequate option of the waste treatment. Therefore, other decision supportive tools are demanded. From the environmental point of view, considering the long-term generation of the emissions, it might be a wrong decision to adopt a cheaper model, as for example the waste landfill. The negative impacts of the emitted methane and leachate can proceed several years to be fully visible.
3. Problems of Isolated Communities The situation of waste treatment in isolated communities is even more visible since they have some particular features that influence significantly their choices of the acceptable management options [3], such as: 0 Islands are by nature isolated and in the most cases without sufficient energy resources and primary energy transportation prices may be prohibitive; 0 The availability of land is very limited (in terms of amount and price), which in practice, reduces landfill disposal of waste in large extent; Recycling and recovery of materials is usually problematic especially due to absence of a market for recycled materials; 0 High price of small scale fossil fuel based technologies (usually diesel); 0 High dependence on fossil fuels (diesel) in power and water production; Tourism represents seasonal fluctuation in the population number, and thus it influences waste generation as well as energy planning. To address the mentioned problems the MSW incineration process combined with energy recovery brings reasonable solutions for islands or other outmost regions [4]. 4. Thermal Treatment of MSW with Energy Recovery
In the past incineration systems were often seen as a disposal route only. However, modern incinerators designed and engineered today have energy recovery as an essential component of the system, to derive economic benefits from the disposal of the waste. The incineration utilises the energy content of the waste to produce heat, electricity or both.
Municipal Solid Waste Valorisation
69
The most developed and widely deployed form of energy recovery from MSW is mass bum incineration. Mass burning represents a large scale incineration of unsorted mixed MSW in a single-stage chamber unit with a moving grate to support, mix and transport the burning waste during the combustion process [4]. During the last decades the industry and science have been acting adequately to respond to the challenge either in forms of improvements of the incinerators design or in terms of investments in the process of control and modifications allowing the emission prevention [ 5 ] . As result overall emission levels from modem municipal waste incinerators, employing sufficient flue gas cleaning system, are well comparable with power plants running on natural gas, which are considered one of the cleanest technologies for energy production from fossil fuels. Waste incineration is a significant option in the waste treatment and disposal in many countries. It has a number of advantages over landfill when applied to MSW while combined with energy recovery: Incineration plants can be situated near the point of collection; It reduces the waste to a biologically sterile ash product to about 10% of its original volume and to 30% of its original weight; It destroys the organic component of biodegradable waste which, when deposited to landfill, would produce a landfill gas (methane), which is a greenhouse gas (GHG) and therefore, a significant contributor to global warming; It generates lower COz emissions than coal, oil or gas when used for the production of electricity; It is the best practicable environmental option for many hazardous wastes such as highly flammable, volatile, toxic and infectious waste, or makes them more suitable for disposal in normal landfills; It recovers useful energy from waste at comparably low costs to produce steam for electric power production, industrial process heating, or hot water for district heating, thus reducing the need to import non-renewable fossil fuels and cutting down emissions to the environment; The bottom ash can be used for materials recovery or as aggregates in construction industry; Furthermore, it may have beneficial effects on local economic development and employment, with conservation of large areas (otherwise used as landfill sites) which can be exploited for other purposes such as tourism or agriculture. There is a wide range of other thermal processes used to treat a wide variety of wastes, i.e. fluidised bed, rotary kilns, cement kilns [7]. They are more appropriate for smaller scale waste throughputs from 1 to 2 tonnes per hour for treatment of sewage sludge, clinical and hazardous waste. However, some of them have not been yet fully commercially proven for MSW treatment.
G. Tavares et al.
70
5. Study Case - Cape Verde Islands
The MSW management means a great challenge for Cape Verde with the direct impact on the level of the environment and health of the population. The prime objective of the present study is the project of an incineration plant treating MSW that can be integrated in the national strategy of the Integrated Waste Management as the part of the concept of sustainable development for the country. Cape Verde is a small state consisting of 10 islands located in the Atlantic Ocean, off the west coast of Africa, in front of Senegal. The population of less than a half million is spread over 9 islands. The largest and the most populated island is Santiago, where the capital Praia is also situated. From the economic point of view it is a developing country constantly growing. It has not its own fuel resources and there is a shortage of drinking water. Tourism as a new sector shows increasing importance. Projections and calculations covering the period of the years till 2020 based on the data from the National Institute of Statistics of Cape Verde (INE) have been performed herein. Presently about 55% of the total population of Cape Verde lives on the island of Santiago, and almost 25% of all inhabitants lives in the capital Praia. It is expected that in 2020 the population of Santiago will rise to 58% and to 30% in Praia. The growth is related to larger work opportunities after independence of the country in 1975. With regard to the anticipated energy consumption there will be a significant increase in comparison with the consumption in 2000 reaching 800 GWWyear in 2020 (Fig. 1).
,
1 000
700 000
600 000 800 500 000
700 600
400000 B
z
g
-
.-0
5
500 0
300000
400
21
P
300
200 000
200 100 000 100
0
0 2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
Year
Fig. 1. Projections of population growth and energy consumption
The average physical composition of produced MSW in Cape Verde is given in Fig. 2. The part with higher calorific value, such as paper, plastics and organic materials,
Municipal Solid Waste Valorisation
71
represents about 75% of MSW, which demonstrates its good quality and suitability for incineration. To cover a part of the demand on the production of electricity and drinking water using energy recovery from MSW, it is proposed to build an integrated power plant. The plant in consideration should be located close to the urban sites as well as the coastal area to facilitate transport of MSW, to allow use of sea water to supply the plant and for eventual discharges. The plant system will comprise of different kinds of technologies including: 0 Incineration unit; 0 Steam turbogenerator; Thermal water desalination unit; Reverse Osmosis water desalination unit; 0 Turbogenerators with Pelton hydraulic turbines. Metals
Oraanic materials
Fig. 2. Physical composition of MSW.
Taking into account the whole arrangement of the presented system the overall efficiency can increase significantly in comparison to a single unit (Fig. 3).
Fig. 3. The proposed scheme of the integrated plant.
G.Tavares et al.
72
The incineration plant consists of a furnace with moving grates for mass burning of MSW, a steam boiler to supply a turbogenerator and a sophisticated filtering system providing treatment of flue gases. The turbogenerator will generate power for the plant internal consumption and the excess of electricity can be supplied directly to the grid. The generated steam expanded at the exit of the turbine can be further exploited in the thermal unit for sea water desalination. A part of the energy produced will supply the Reverse Osmosis unit, where the pre-treated sea water from the thermal desalination unit will be finally processed. This combination of the desalination units will allow a significant increase of lifetime of the membranes, as well as less frequent maintenance. The Reverse Osmosis is a process working with very high water pressure. Therefore, such a high pressure of the discharged water at the Reverse Osmosis unit exit encourages the combination of it with hydraulic Pelton turbines, coupled with generators of small dimensions to exploit the water energy potential, and thus to increase the overall efficiency of the complex system.
6. Alternative Scenarios To make the investments in the integrated plant as much profitable as possible, some alternative strategies of the waste management in the form of two applicable scenarios are presented in this section. The first scenario considers only the collection and the processing of MSW from the island of Santiago. The projections of the plant throughput and power generation for this scenario are displayed in Fig. 4. 15,O
,
500
450 +Energy -+-Potential
Consumption (GWh) Energy (GWh)
400
350
e~
gg 300
~2 a m
250 2
s.
200 150 100
50
0
Fig. 4.Projections of the plant capacity and potential energy production-Scenario 1.
x l
FF
t s
73
Municipal Solid Waste Valorisation
The second proposed alternative is to collect MSW also from the remaining eight inhabited islands of Cape Verde and to transfer it to the plant on the island of Santiago. The projections of the plant capacity and potential produced energy for this case are shown in Fig. 5. Besides the two previous scenarios, the possibility of processing simultaneously used oils and worn out tyres with the aim to eliminate the soil contamination and its disposal or burning in the open fire has also been analysed. In terms of resources they represent an enormous energetic potential since they are derivates of fossil fuels.
7. Discussion of Results In the present section the obtained results for the proposed scenarios will be summarised and discussed. 500
30,O Capacity (Vh)
25,O
+Energy -+-Potential
-
20,o
450
Consumption(GWh) Energy (GWh)
400
350 . p
c
300
s 15,O
250
s
E;E g ;F
mn
C
e
s w
8
10,o
100 50
Fig. 5. Projections of the plant capacity and potential energy production-Scenario 2.
7.1. Scenario 1 Analysing the projection of the MSW production at Santiago up to 2020 (see Table l), it can be concluded that the available waste amount 424 t/d in 2020, considering the efficiency of the collection, will be enough to feed an incineration plant of 9 t/h of capacity with an electricity output 5 MW. In 2015 the quantity of the collected waste can supply the plant of the capacity 12 th,and in 2020 14 t/h generating 6 and 7 MW, respectively. This gradual increase may facilitate both the division of the investments demand and the extension of the incineration plant capacity to a few delayed phases, which allows the erection of it more smoothly.
G. Tavares et al.
74
Table 1. Results of the projections for the different scenarios.
Unit
2010
2015
2020
t/d t/h MW
424 9 5
49 1
570 14 7
Total MSW Production
t/d
743
856
Plant Canacitv
till
Potential Power
MW
16 8
20 10
Seen Total MSW Production
Plant Capacity Potential Power
~~~~
12 6
988 25 12
7.2. Scenario 2
Considering the collection of the waste according to the Scenario 2 we have obtained the following results (Table 1): in 2010 the incineration plant can process around 16 tonnes of waste per day with 8 M W of power output. In 2015 the capacity of the plant can rise to 20 t/d and to 10 MW of electricity. The data for 2020 show that it will be possible to extend the daily throughput of the plant to 25 t/h generating 12 M W of electricity. It will correspond to 22% of the total energy consumption anticipated to Santiago. Furthermore, it will increase the availability of the required investments and at the same time it will decrease costs of energy and water production. Additionally, an outstanding advantage of this scenario is non-necessity to build landfills in the other islands. The municipalities can carry out the collection, preparation and transfer of MSW to Santiago. Thus the possible investments in landfill can be transferred to improve the conditions of waste collection, to subsidy the transfer costs and to increase of the plant capacity. The power production on Cape Verde is presently based on the diesel technology. Even taking into account the future use of wind energy, which shows promising potential for some of the islands, it is not possible to fully replace fossil diesel production. In 2020 it is projected that wind can cover 30% of the energy consumption [6]. Comparing the present projections of the waste-to-energy from Scenario 1 and Scenario 2, it is expected that in 2020 MSW incineration may substitute up to 13% (Fig. 6) and 22% (Fig. 7), respectively, of the energy otherwise generated from diesel. Diesel 57%
Waste-to-energy 13%
Wind 30%
Fig. 6. Potential energy generation by resources in 2020 - Scenario 1.
Municipal Solid Waste Valorisation Diesel 48%
Waste-to-energy 22%
Wind
30%
Fig. 7. Potential energy generation by resources in 2020 - Scenario 2.
Nowadays Cape Verde imports more than 600 tonnes of tyres per year, which represents more than 1,6 tld, and this trend appears to increase yearly. Additionally, the quantity of imported lubricants is about 1200 tonnes per year, i.e. more than 3 tld. Compared to other resources, these materials have a significant energy potential equivalent to approximately 7% of the total waste amount, in terms of thermal input to the incinerator. However, this part requires more deep investigation and further quantifications.
8. Conclusions The establishment of a properly working, sustainable integrated waste management system is necessary for any economy. It requires government policies that encourage waste prevention, reuse and both materials and thermal recycling. Suitable legislation is necessary with incentives and raise of public awareness, to maximise the valorisation of waste materials. The whole structure of the waste management can vary with the conditions of the community, in order to hll y satisfy all its requirements. Such particularities have to be taken into account. This is especially true for isolated regions, which have some specific features and usually face problems of energy and water supply. The presented study of two scenarios for Cape Verde islands, one considering only waste collection from the island of Santiago and the other considering collection from the 9 inhabited islands, shows that a waste-to-energy recovery up to 22% of the total predicted energy consumption is possible to obtain for Santiago in 2020. An outstanding benefit of such scenario is the non-necessity of landfilling in the other islands. Furthermore, it represents advantages of environmental character by decreasing demand for fossil fuels, as well as social and economical, since it reduces the national dependence on significant fuel import. It will open new job opportunities and save land reservations that would certainly be used for landfill disposals of waste. The MSW incineration has a considerable potential for reduction in reactivity of any residual landfilled material. This, associated with heat recovery, efficient flue gas
16
G. Tavares et al.
cleaning systems and significant greenhouse gas abatement capacity makes the MSW incineration an attractive environmentally sound option. Another benefit is the transfer of high technologies and a possibility to gain financial support from developed countries, according to the framework of the Kyoto Protocol under its flexible mechanisms, to reduce emissions of greenhouse gases. Acknowledgements
This work has been carried out under the National Energetic Plan of Cape Verde (PEN) and the ECOTHERM project funded by the European Commission in the 5'h Framework Programme. The authors would like to acknowledge the Portuguese Ministerio da Ciencia e Tecnologia for financial support for the scholarship of Eng. Gilbert0 Tavares, and EC Marie Curie Fellowship Programme for the scholarship of Dr. Zdena Zsigraiova. References
[I] OECD Environmental data compendium, Chapter 7, Waste. (1999) OECD, Paris. [2] ISWA (2002) Energy from Waste, State-of-the-Art-Report, 4thEd., ISWA, Denmark. [3] Project WTE-ISLE. Waste management in Island Communities: Strategy to integrate waste-to-energy policies, 5thFP of the EC, DG-TREN. [31 Waste as resource for sustainable development, URL: www.scientecmatrix.com. [4] Niessen WR. (2002) Combustion and Incineration Processes, 31d Ed., Marcel Dekker, Inc. New York. [5] Goerner K. (2003) Waste incineration European state of the art and new developments. IFRF CombustionJournal, Article No. 2003. [5] Williams PT. (1999) Waste Treatment andDisposa1. John Wiley & Sons Ltd. [6] Duic N, Alves LA, Chen F, Carvalho MG. (2003) Potential of kyoto protocol clean development mechanism in transfer of clean energy technologies to small island developing states: case study of Cape Verde. Renewable & Sustainable Energy Reviews 7: 83-98.
PRODUCTION OF SUSTAINABLE HYDROGEN USING THERMOCHEMICAL GASIFICATION OF BIOMASS J. ANDRIES, W. DE JONG, H. SPLIETHOFF
Section Energy Technology, Delft University of Technology, Deyt, The Netherlands Hydrogen production systems based on thermochemical gasification of biomass typically consist of the following basic steps: pre-treatment, gasification, gas cleaning, upgrading and separation. The development of the technology depends on the successful integration of these steps and technical and economic analysis is required to match the optimum technology with available feedstock to produce a product of the necessary quality for the target application. The section Energy Technology is participating in a new EU-funded project which is aimed at developing and optimizing an energy-efficient and cost-efficient method to produce hydrogen-rich gases from biomass, including residues, using circulating fluidised bed technology. This gas can then be upgraded to commercial quality hydrogen or to synthesis gas for further upgrading to liquid fuels such as DME and methanol or Fischer-Tropsch diesel. An existing 100 kWth circulating fluidised bed gasifier will be modified and used to assess the influence of operating parameters such as oxidant type (oxygen concentration, steam) on the gasifcation process experimentally. The influence operating conditions relevant for hydrogen production on the formation of the main components and contaminants will be determined under. An experimental and theoretical analysis will be made of the fate of alkali metals and particulates throughout the system including the high temperature gas filter. An analysis of the bottlenecks caused by these components will be given and a detailed description of the Delft tasks and their role in the multidisciplinary R&D project will be presented.
1. Introduction
The growing consumption of energy, which is now mainly produced from fossil fuels, leads to a further depletion of the limited fossil reserves, increased dependency on fossil fuel imports from often politically unstable countries and furthermore gives rise to an increasing concentration of greenhouse gases (mainly COz) in the atmosphere. There are more and more indications that this contributes to global climatic changes. One of the ways to reduce the anthropogenic contribution to this problem is the application of a C02-neutral primary energy carrier instead of the net C02-producing fossil fuels, which are currently used. Hydrogen potentially is such a carrier that can also reduce dependency on fossil fuels [ 11. Sources for hydrogen are depicted in Fig. 1. Today the least expensive and most widely applied technology to produce hydrogen is steam reforming of natural gas. The global hydrogen production currently exceeds 1 billion m3/d of which 48% is produced from natural gas, 30% from oil, 18% from coal, and the remaining (4%) via water electrolysis [ 2 ] . The production of hydrogen from fossil fuels, however, is accompanied by the production of COz, CH4, and other polluting gases that contribute to the enhancement of the natural greenhouse effect [3]. Hydrogen on the other hand can be produced sustainably in a number of ways, that are nowadays certainly not applied at large scale: the most promising ones are water splitting by electrolysis using sustainably produced electricity (e.g. wind, hydro and sun) and processes based on sustainable production and utilization of biomass. 77
78
J. Andries, K de Jong & H. Spliethof
Biomass-based processes can be either thermochemical or biological and can produce hydrogen directly or via an intermediate, storable product. A recent overview of the status of the different technologies is given in [4].
Fig. 1. Hydrogen production and utilisation alternatives [ 5 ] .
To enable the large-scale introduction of sustainably produced hydrogen, a number of technical and non-technical issues have to be addressed, such as the production technology chain, storage options, smooth integration into the existing energy system and social acceptance. The transition from the present fossil fuel-based energy system into a sustainable, hydrogen-based energy system will be fragmented and involve a diverse mixture of fossil and renewable energy sources and carriers. The use of hydrogen produced from fossil fuels will probably be an initial step toward the realization of a sustainable energy system based on hydrogen. This paper will focus on large-scale application by means of thermochemical biomass gasification. This route has two attractive features [6]: It is omnivoric, i.e., a very broad range of biomass feedstock can be completely converted; Syngas is a universal intermediate to a broad range of commodity chemicals, transportation fuels and specialty chemicals.
Hydrogen using Thermochemical Gasifcation of Biomass
19
2. Hydrogen from Thermochemical Gasification of Biomass
Thermochemical gasification of biomass generates so-called biosyngas, rich in CO and HZ. Fig. 2 shows an example of a biosyngas production scheme for tri-generation of methanol, hydrogen and electricity. This process scheme typically consists of the following basic steps: biomass pre-treatment, gasification, gas cleaning, reforming of higher hydrocarbons, shift to obtain suitable H2:C0 ratios and gas separation for hydrogen production and purification. Optional are a gas turbine or boiler to utilise the unconverted gas, and a steam turbine to produce electricity.
2.1. Pretreatment Pretreatment of biomass (drying, chipping, sizing, densification, and optional torrefaction) is of great importance for a reliable feeding system and furthermore has a big influence on storage, handling and transportation characteristics. Integration of the drying process within the system (steam drying) and development of reliable feeding systems for pressurised gasifiers are areas needing substantial R&D.
pretreatment
4
optimal
Fig. 2. Biosyngas plant scheme for tri-generation of methanol, Hz and power [7]
2.2. Gasification Fluidised bed (FB) gasification together with entrained flow (EF) gasification of solid biomass feedstock are technologies which can basically realise the required throughput for a hydrogen production facility with comparatively high thermal efficiencies. The two technologies have their own specific advantages and disadvantages. The greatest risks reported for EF gasification are slagging phenomena in the gasifier and fouling in the syngas cooler, whereas for circulating fluidised bed (CFB) gasification risks are mainly concentrated in the fouling of the tar cracker and syngas cooler [6]. The CFB process and also the (pressurised) bubbling FB process, contrary to EF gasification, is much more fuel flexible: EF gasification needs very small particles for good fuel conversion. This requires an extra process step of biomass milling+sizing or torrefaction+milling+sizing.
80
J. Andries, W. de Jong & H. Spliethoff
Biosyngas production costs for both technologies were reported to be quite similar: 6-13 euro/GJ, depending on the biomass price, which were assumed to be 0.4-4 euro/GJ [ 6 ] . Furtheron, we will concentrate on fluidised bed technology as primary gasification step. The extensive and complicated downstream equipment (gas cleaning, reforming, shift, purification and separation) requires a product gas with low nitrogen dilution to reduce equipment size. This can be produced by an oxygen blown or indirectly heated process. The Institute for Gas Technology (IGT) and the Batelle Columbus Laboratory (BCL) processes are potentially attractive processes for the production of hydrogen from biomass. The IGT gasification process is using a pressurised, oxygen-blown bubbling fluidised bed gasifier. It produces a C02-rich gas and the hydrogen content can be increased by using more steam. The H2:CO ratio must be shifted using the shift reaction and the CH4 in the product gas must be reformed or separated and used in a gas turbine. Pressurised gasification allows a large throughput per reactor volume and diminishes the need for pressurisation downstream. So less overall power is needed in this process. The BCL gasification process uses an atmospheric, indirectly heated, CFB gasifier. It produces a gas with low C 0 2 content. The hydrocarbons in the product gas must be reformed to COz and H2.The large CO fraction must be shifted to H2. 2.3. Gas Cleaning
Particulates, bound sulphur species (HzS, COS, CS2), halogens (mainly HCl), bound nitrogen species (HCN, NH3), alkali (Na, K) and other trace species and tars have a potentially harmful1 effect on downstream equipment, especially catalysts. The maximum acceptable levels are very stringent and are reported to be 10 - 20 ppb [8]. Further experimental in-depth investigation, though, is necessary towards achievable and tolerable levels of these minor gas constituents for the specific system components applied. As gas cleaning options, a choice can be made for either low temperature “wet” or high temperature “dry” gas cleaning. The latter type of gas cleaning is advantageous for the overall energy balance but many process elements are not proven technology. The ultimate selection depends on the properties of the produced gas and the end use. The sequence and temperature levels of the different steps require a careful analysis of input properties and output requirements. 2.4. Hydrogen Yield Maximization
The syngas can contain a considerable amount of methane and other light hydrocarbons, representing a significant part of the heating value of the gas. Steam reforming converts these compounds to CO and H2 driven by steam addition over a nickel catalyst. Autothermal reforming combines partial oxidation in the first part of the reactor with steam reforming in the second part, thereby optimally integrating the heat flows. It is not clear from the literature whether steam reforming can deal with the high CO and COz
Hydrogen using Thermochemical Gasijcation of Biomass
81
content of the biomass syngas and whether carbon deposition can be avoided in both types of reformers. The syngas produced by the BCL and IGT gasifiers has a low HZ- CO ratio. The water gas shift reaction is a common process operation to shift the energy value of the CO to HZ, which can than is separated using pressure swing adsorption. The equilibrium constant for the water gas shift reaction increases as temperature decreases. Hence, to increase the production to H2 from CO, it is desirable to conduct the reaction at lower temperatures, which is also preferred in view of steam economy. However, to achieve necessary rates higher temperatures are required. New processes are being developed as part of the US hydrogen development programme. An example is the selective partial oxidation or Praxair-Air Products process with membrane separation and catalytic reformation. 2.5. Purification and Separation
After reforming and shifting of raw syngas, hydrogen can be separated and compressed. Today’s hydrogen plants use different process concepts. New hydrogen plants use pressure swing adsorption (PSA) €or purification. The hydrogen quality is a major issue for its eventual automotive application. Specifically, CO is a strong poison to polymer electrolyte membrane (PEM) fuel cells. Studies indicate that levels as low as 1-2 ppb will deactivate the platinum anode material of such fuel cells. Membrane technology is promising for gas separation, due to simple design and the ability of combining shift and separation in one reactor. Much R&D effort is put in decreasing the pore size, so that membranes can be molecular sieves. Membranes are expected to perform better at high temperature, implying an important energy advantage in combination with hot gas cleaning, as no temperature drop would have to occur between gasifier and gas turbine. The membrane surface also catalyses the water gas shift reaction. This reaction is driven to hydrogen as it is removed by the selectively permeable membrane. Ceramic membranes have the advantage of a broad temperature and operating pressure range. Construction of membrane separation units is potentially very simple and cheap when compared to other technologies like PSA. Moreover, membranes do not suffer the efficiency losses and high capital costs for heat exchangers, associated with the need to cool the syngas. 2.6. Research Requirements
Several analyses have been made of the hydrogen production cost using biomass [6-9111. A recent overview is given in [7]. The main conclusion is that the production costs still have to be decreased and improved performance can be obtained by: Application of new and improved technologies; 0 Combining fuel and power production; Using economy of scale.
82
J. Andvies, W. de Jong & H. Spliethoff
The technology development depends on the successful integration of the different process steps and technical and economic analysis is required to match the optimum technology with available feedstock to produce a product of the necessary quality for the target application. Although a lot is known about each of the different subsystems separately, this knowledge is often obtained during research aimed at very different applications. Because of this, experimental data on the main subsystems has to be obtained under conditions which are applicable to hydrogen production using thermochemical gasification of biomass. Research towards ceramic hot gas filtration has shown that there are still quite some research questions to be solved, especially regarding the fate of trace species and the use of additives [12]. Special attention has to be paid to scale effects in order to enable optimal sizing and a reliable assessment of the economies of scale. A great similarity exists between biomass-based processes producing hydrogen, methanol, and clean syngas for liquid fuel production, causing substantial synergy between the projects in these areas. Currently, a number of large, government funded R&D programmes are on-going in the USA [ 131, Japan, the EU and in The Netherlands.
3. Projects The section Energy Technology of the Delft University of Technology is currently involved in two related projects regarding production of hydrogen rich syngas, of which one, “SENECU”, has started in January 2004 and the other, EU FP6 integrated project “CHRISGAS” [ 141, is being evaluated by the EC. 3.1. Sustainable HydrogenSENECU
A multidisciplinary R&D project, SENECU, has started as part of a TU Delft sponsored programme aimed at studying the introduction of sustainable hydrogen in the Netherlands. The overall project aim is to identify and solve the bottlenecks in the hydrogen production technology chain including fuel cells by design, modelling, and optimisation of an integrated hydrogen production system based on thermochemical gasification of biomass. The information on the different sub processes will be integrated into a flowsheeting model, describing a complete installation producing hydrogen from biomass. This model, which will be based on validated experimental data as much as available, is targeted to predict the process efficiency, the production cost of the hydrogen and the environmental effects of the installation. The dynamic behaviour of the different sub-processes and the integrated system under transient conditions will be determined as well. The results will be used to develop, optimised and test a robust and reliable control system. Special attention will be paid to potentially cost-reducing technologies. An economic analysis will be made to determine the feasibility of hydrogen production using thermochemical biomass gasification. Scaling laws for the different processes will be assessed and used to determine the optimum production size.
Hydrogen using Thermochemical Gasijication of Biomass
83
3.2. Production of Clean Hydrogen-RIch Synthesis GAS-CHRISGAS
The transport sector represents an increasing share of the world’s total fossil fuel use. In order to fulfil the Kyoto Protocol obligations the transport sector’s dependence on oil must be reduced. Therefore, it is aimed at increasing the use of vehicle fuels produced from renewables. This project will develop and optimise an energy- and cost-efficient method to produce hydrogen-rich gases from biomass, using CFB technology [ 141. This gas can be upgraded to commercial quality hydrogen or synthesis gas for further upgrading to liquid fuels like DME and methanol or Fischer-Tropsch diesel. New equipment will be developed, tested, and implemented in a demonstration facility to produce clean biosyngas for vehicle fuel production. This facility, the 18 MWth Varnamo plant, is to be reconstructed so as to be able to produce ca. 7000 m:/h biosyngas. Also included in the project are studies related to the large-scale use of such plants and their impact on the environment and society. The project consists of tasks for which the objectives are: Conversion of solid biofuels into syngas by PCFB gasification using a steam and oxygen mixture; Improved in-situ desulphurisation by means of dedicated additives; Cleaning of the generated gas from particulates in a high temperature filter; Purification of the generated gas by catalytic Autothermal steam reforming of a broad spectrum of hydrocarbons, to generate a raw synthesis gas consisting mainly of carbon monoxide and hydrogen as energy carriers; Conditioning of the raw syngas, meeting the quality stipulated for manufacture of DME (dimethylether) or other potential products. This is attained by catalytic water gas shift to adjust the CO/H2 ratio, catalytic hydrogenation of minor contaminants followed by removal of carbon dioxide and other acidic gases in a conventional wash system or in novel, selective processes for deeper S-removal; Studies of fuel production from various biofuels, at a scale and cost representative of biomass fuel chains in various regions in Europe, thereby rendering realistic data for future production potential on a European scale; Large scale feeding system testing of a compact, high-capacity piston feeder with the advantage of much lower total energy consumption than using lock hoppers. At the laboratory of Energy Technology, an existing 100 kwth airblown CFB gasifier (see Fig. 3) will be modified and used to assess the influence of operating parameters such as fuel species, oxidant type (oxygen, steam), stoichiometry and additives on the gasificationlgas cleaning process. The influence on the formation of the main components and contaminants will be determined under operating conditions relevant for large-scale hydrogen production. High temperature (HT) gas cleaning by means of advanced Pall-Schumacher ceramic candle filters will be further developed and optimised for conditions relevant for hydrogen production. Filtration behaviour will be tested behind the Delft gasifier. An experimental and theoretical analysis will be made of the fate of tarshoot, alkali-, other trace metals and solids in the system including the HT gas filter.
J. Andries, W. de Jong & H. Spliethoff
84
REMARK:TEMPERATURE AND PRESSURE SAMPLING NOT INCLUDED ON THE DRAWING 'TICAL PORT - ELlF
BACK-PULSING NITROGEN VESEL
0-16 kglh
N2, 14 bar. 0-18 kglh
d
\ir max. 45 kglh Jitrogen
rT
rT
0 4 5 kglh DN 25
DN25
4
4 bar
AirINitrogen preheater - 400'C
040 kglh
Fig. 3. Schematic of the existing 100 kWthCFBG test rig at TU Delft.
4. Conclusions and Recommendations The two projects described in this paper are both aimed at the development and societal implementation of biomass-based production of hydrogen. The integration of the different process steps and the optimisation of the whole process chain should lead to lower production cost of hydrogen. There is a large common area between systems producing hydrogen, methanol, and hydrogen-rich syngas from renewable biomass for the production of liquid transportation fuel. Exchange of information and integration of projects working on these processes is very important to make optimal use of available funding. References [ 11 Turner JA. (1999) A realizable renewable energy future. Science 285: 687-689. [2] US DOE Energy Efficiency and Renewable Energy. Hydrogen, fuel cells & infrastructure technologies programme/production & delivery, 2003, http://www.eere.energy.gov/hydrogenandfuelcells/hydroge~production. html. [3] Collot AG. (2003) Prospects for hydrogen from coal. Report CCC/78 IEA Clean Coal Centre, London. ISBN-92-9029-393-4.
Hydrogen using Thermochemical GasiJicationof Biomass
85
[4] Milne TA, Elam CC, Evans, RJ. (2001) Hydrogen from biomass, state of the art and research challenges. IEA report IEA/H2/TR-02/00 1. [5] EC Hydrogen energy and fuel cells - A vision of our future. Report, Brussels, Belgium, European Commission, 2003. High level group on hydrogen and fuel cells. Available at: http ://europa.eu.int/comm/research/energy/pdf/hydrogen-report-en.pd~ [6] Calk HPA, Haan JP, Boerrigter H, van der Drift A, Peppink G, van den Broek R, Faaij APC, Venderbosch RA. (2002) Proceedings of the expert meeting on pyrolysis and gasfication of biomass and waste, 30 September-1 October, Strasbourg, France, pp. 403-417. [7] Hamelinck CN, Faaij AF’C. (2002) Future prospects for production of methanol and hydrogen from biomass. Journal ofpower Sources 111 (I): 1-22. [8] Tijmensen MJA, Faaij APC, Hamelinck CN, van Hardeveld MRM. (2002) Exploration of the possibilities for production of Fischer Tropsch liquids and power via biomass gasification. Biomass and Bioenergy 23 (2): 129-152. [9] DeLuchi MA, Larson ED, Williams RH. (1991) Hydrogen and methanol: Production from biomass and use in fuel cell and internal combustion engine vehicles - A preliminary assessment. PU/CEES Report 263, Center for Energy and Environmental Studies, The Engineering Quadrangle, Princeton University, Princeton, New Jersey, USA. [lo] Larson ED, Katofsky RE. (1992) Production of methanol and hydrogen from biomass. PU/CEES Report 27 1, Center for Energy and Environmental Studies, The Engineering Quadrangle, Princeton University, Princeton, New Jersey, USA. [11] Komiyama H, Mitsumori T, Yamaji K, Yamada K. (2001) Assessment of energy systems by using biomass plantation. Fuel 80 (5): 707-715. [12] Unal 0, Bode A, de Jong W, Spliethoff H. (2002) Pressurised fluidised bed gasification of biomass and browncoal: Hot gas cleaning with ceramic channel flow filters. Proceedings of the conference ‘Neue Entwicklungen bei der Heissgasreinigung von Brenn- und Rauchgasen ’ (CD-ROM), 14 November, Leverkusen, Germany. [13] U.S. DOE A multiyear plan for the hydrogen R&D program. Rationale, structure, and technology roadmaps, 1999. [14] Thornqvist T. (2003) Clean hydrogen-rich synthesis gas. CHRISGAS, Vaxjo University, Sweden. Available at: ftp ://ftp.cordis.lu/pub/sustdev/docs/energy/bioenergy~bO4 .pdf.
This page intentionally left blank
CHARACTERIZATION OF KENAF POTENTIAL IN PORTUGAL AS AN lNDUSTRIAL AND ENERGY FEEDSTOCK A.L. FERNANDO, M.P. DUARTE, J. MORAIS, A. CATROGA, G. SERRAS, B. MENDES, J.F.S. OLIVEIRA
Grupo de Disciplinas de Ecologia da Hidrosfera / Unidade de Biotecnologia Ambiental, Faculdade de CiEncias e Tecnologia da Universidade Nova de Lisboa, Quinta da Torre, 2829-51 6 Caparica, Portugal; e-mail: alawct. unl.pt orjyso@fct. unl.pt The purpose of this work was to investigate the influence of crop management on the kenaf biomass quality and productivity, in Portugal, in order to access its potential as an industrial and energy feedstock. To do so, the effects of different varieties, sowing dates, plant populations, different irrigation and nitrogen fertilization levels in the biomass quality and productivity, were studied. Productivity was affected by the level of irrigation but not by the level of nitrogen fertilization. Higher productivities were obtained in the fields with 301 and 400 mm of water added. Biomass quality was not affected either by the level of irrigation of the fields either by the level of N-fertilization. Productivity and biomass quality were affected by the sowing date but not by the plant population. Plants sowed earlier presented better productivities and better quality for fuel and pulp purposes than plants sowed latter. Tainung 2 and Everglades 4 1 presented similar productivities and mineral composition. Everglades 41 showed better quality for fuel purposes and Tainung 2 better quality for pulp purposes. Possibilities of industrial utilization under different field conditions were assessed and discussed, in the perspective of sustainable growth and development.
1. Introduction
Kenaf (Hibiscus cannabinus L.) is a short day, annual, herbaceous plant processing high quality cellulose. It is a member of the Malvaceae family along with cotton and okra, and is endemic to Africa [l]. The entire plant can be used to produce pulp for the paper industry. Lower quality paper can be made from the short wood fibres of the inner core, while high quality paper can be made from the long fibres of the bark [ 2 ] . Kenaf, as a high yielding plant, is also a potential energy crop when used as a whole crop. The residues from different industrial processes can, as well, be utilized as energy sources [ 11. In the scope of the project Biomass Production Chain and Growth Simulation Model for Kenaf (Biokenaf), supported by the European Union, the purpose of this work was to investigate the influence of crop management on the kenaf biomass quality and productivity, in Portugal. As this crop is being introduced in Portugal, as a new crop, possibilities of industrial utilization under different field conditions are assessed and discussed, in the perspective of sustainable growth and development.
2. Materials and Methods The experimental fields are situated in the Peninsula of Setubal, in the south border of the river Tejo, near the estuary and the Atlantic coast (latitude 38"40'N, longitude 9" W, altitude of 50 m) where the climate is warm temperate. During the experimental period, July 2003-January 2004, the average minimum temperature was 15.6"C and the average maximum temperature was 23.0°C, with a total of 443 mm rainfall. The experimental plots were established in a clayey and alkaline soil. For the nitrogen and irrigation levels study, one kenaf variety was studied, Tainung 2. The fields were sowed at 4'h July using a row spacing of 0.50 m and a distance within row of 0.10 m (20 seeds per m'). P-fertilizer (60 kg P205.ha-'), K-fertilizer (120 kg K20.ha-') and !h Nfertilizer were applied at the time of sowing. The other % N-fertilizer was applied when 87
A.L. Fernando et al.
88
the plants reached approximately 20 cm height (about 1 month after sowing). Three different levels of N-fertilizer were applied: 0, 75 and 150 kg N.ha-'. At early stages of growth, all the fields were fully irrigated in order to compensate the water deficit of the soil, and to prevent water stress. 41 days after sowing, irrigation was differentiated, and four different levels were applied: 0%, 25%, 50% and 100% PET. During the growing season a total of 204, 253, 301 and 400 mm of water were added to the fields. A factorial 3 (three levels of N-fertilizer) x4 (four irrigation levels) split-plot design in 3 blocks was used. Standard basic plots had a surface area of 5x9 m2. For the sowing dates and plant population study, two kenaf varieties were studied, Tainung 2 and Everglades 41. The fields were sowed at 26" June and at 1lthJuly. A row spacing of 0.50 m was used and two different distances within row were studied: 0.10 m (20 seeds per m') and 0.05 m (40 seeds per m2). P-fertilizer (60 kg P205.ha-'),K-fertilizer (120 kg K20.ha-') and % N-fertilizer (37.5 kg N.ha-') were applied at the time of sowing. The other !h N-fertilizer was applied when the plants reached approximately 20 cm height (about one month after sowing). All the fields were fully irrigated in order to compensate the water deficit of the soil, and to prevent water stress. A randomized block design with three replications was used. Standard basic plots had a surface area of 5x8 m'. At the end of the growing season the vegetable material was harvested and the total aerial dry weight, the nitrogen content, the phosphorus content, the fibre content and the gross heat of combustion were determined in order to evaluate the productivity and the quality of the biomass. The chemical analyses were performed according to the following procedures: a) nitrogen content: by the Kjeldahl method; b) phosphorus content: by the ascorbic acid method, after digestion of the sample; c) fibre: by the Weende method; d) gross heat of combustion: using an adiabatic calorimeter.
3. Results and Discussion 3.1. Biomass Productivity Fig. 1 shows the differences, in terms of the dry matter productivities among fields with different nitrogen application and among fields with different levels of irrigation. Fig. 2 shows the differences, in terms of the dry matter productivities between fields sowed at two different dates, between fields with different varieties and between fields sowed with different plant densities.
-
.->
3 0
2 n
1
1 1
0 Kg 75 Kg 150 Wha Nlha Kg Wha
H20 H20 H20 H20
Nlrcgen fertilization
Irrigation levels
~
204 rnrn
253
301
rnm
m
400 ~
rnrn
Fig. 1. Productivities obtained in fields with different nitrogen application levels and in fields with different irrigation levels.
Kenaf Potential in Portugal as Energy Feedstock
89
According to these results there were only significant differences among productivities obtained in fields with different levels of irrigation. Higher productivities were obtained for the 50% and 100% PET fields, respectively with 301 and 400 mm of water added. There were no significant differences among productivities obtained in fields with different levels of nitrogen, probably because the soil was rich in nitrogen. E 8.00
%
1
1
7.00 E 6.00
5.00
2
4.00
3.00
5 2.00
5 .-
1.00
0.00
c
z U
n e
1
Sowing date
1
Variety
1 Rant population
I
Fig. 2. Productivities obtained in fields sowed in two different dates, in fields with different varieties and in fields sowed with different plant densities.
According to these results there were only statistical significant differences between fields sowed at two different dates. Plants sowed earlier (26th June) presented significantly higher productivities than plants sowed 15 days later (1 lthJuly). Everglades 41 presented higher productivities than Tainung 2, although this difference was not significant. This result was already expected since both varieties belong to the same group of the late maturity varieties. Productivities obtained in the fields sowed with the higher plant density (40 plants/m2) were higher than those obtained in the fields sowed with the lower density (20 plants/m2). However, this difference was not significant, chiefly because Kenaf reduces its population during the growing season, being this effect more pronounced in the fields sowed with a higher density [3]. The productivities obtained were significantly lower than those obtained by other Mediterranean partners of the Biokenaf project [4]. These lower productivities were mainly due to the difficulties experienced during the first year of the project, namely, due to the late sowing date and the heat wave experienced during the 2003 Summer [ 5 ] .
3.2. Biomass Quality In order to meet the processing needs, as for the energy sector as for the pulp production sector, by the time of harvest, the crop should have low mineral and water contents. Kenaf biomass is usually harvested after leaf fall, allowing the return of nutrients from the fallen leaves, namely, nitrogen, phosphorus and other minerals, back to the soil. Table 1 shows the results obtained concerning the moisture content, the nitrogen content and the phosphorus content of the kenaf biomass obtained in the experimental fields. In all the fields, the only variable that showed some influence on the biomass quality was the sowing date. Effectively, in terms of nitrogen content, plants sowed fust, presented significant lower nitrogen content than plants sowed fifteen days later. Nevertheless no influence was observed as for the moisture content as for the phosphorus
90
A.L. Fernando et al.
content. No statistical significant differences were observed between the two varieties (Tainung 2 and Everglades 41), between plants obtained in fields sowed with two different plant densities, among the plants obtained in fields with different levels of nitrogen and in fields with different levels of irrigation. Bark presented less moisture than core material and, although not significant, presented higher nitrogen and phosphorus content than core material. Table 1. Moisture content, nitrogen content and phosphorus content of core and bark of Kenaf biomass obtained in the experimental fields. Moisture (“h) Nitrogen (% dry matter) Phosphorus (“hdry matter)
Core
Bark
53 f 18 0.4 40.3 0.13 f 0.06
36 f 20 0.9 f 0.4 0.25 f 0.08
3.3. Potential Energy Production
The potential energy produced during combustion of the harvested biomass was evaluated as the productivity (t/ha) x gross heat of combustion (GJ/t) (Tables 2 and 3). Table 2. Potential energy production (GJiha) of core and bark of Kenaf biomass obtained in fields with different nitrogen application levels and in fields with different irrigation levels. Irrigation Level
204 mm H20 253 mm H 2 0
301 m m H 2 0
400 mm H 2 0
Nitrogen Fertilization 0 kg Niha
75 kg Nlha 150 kg Niha 0 kg N k a 75 kg N/ha 150 kg N/ha 0 kg N k a 75 k g N k a 150 kg N k a 0 kg N h a 75 kg N k a 150 kg Niha
Core
Bark
Total
14 18 21 28 21
7 9 9 12 12
21 27 30 40
23
8
42 39
16 17
31 58 56 38 53 52 79
25
13
37 35
16 17 26
53
33
In terms of gross heat of combustion, there were no statistical significant differences among the plants obtained in fields with different levels of nitrogen and with different levels of irrigation. Then, in terms of the potential energy production, differences among the fields were observed, namely among the different levels of irrigation, only due to the differences in the productivities. Highest values were obtained in the most irrigated fields, with 301 mm and 400 mm H20. No statistical significant differences were observed among levels of nitrogen fertilization. The potential energy production was considerably higher in the fields sowed first than in the fields sowed fifteen days later, mainly due to the highest productivities obtained in those fields but also due to the highest gross heat of combustion of the plants harvested from those earlier fields. Although not statistically significant, higher potential energy production was also observed in the fields sowed with a higher plant population. This resulted from the higher productivities obtained in those fields because, in terms of the gross heat of combustion, no significant differences were observed between plants
Kenaf Potential in Portugal as Energy Feedstock
91
obtained in fields sowed with two different plant densities. The bark of Everglades 41 presented a higher gross heat of combustion than the bark of Tainung 2. No differences were obtained between the inner core of both varieties. But? in terms of potential energy production no significant differences were observed. In all the experimental fields, the bark presented lower quality and quantity for energy purposes than core, due to its lower gross heat of combustion and to its lower productivity. Table 3. Potential energy production (GJ/ha)of core and bark of Kenaf biomass obtained in fields sowed in two different dates, in fields with different varieties and in fields sowed with different plant densities.
Sowing date
Variety Tainung 2
26’h June Everglades 41 Tainung 2 1lthJuly Everglades 41
Plant population
Core
Bark
Total
20 seeds/m2 40 seeds/m2 20 seeds/m2 40 seeds/m2 20 seeds/m2 40 seeds/m2
60 88 80 80 32 50 34 46
26 26 20 34 14 22 19 24
86 114 101 114 46 72 54 69
20 seeds/m2 40 seeds/m2
~
3.4. Potential Pulp Production The potential pulp produced after the harvest of the biomass was evaluated as the productivity (t/ha) x fibre content (“XIdry matter) (Tables 4 and 5). Table 4. Potential pulp production (tiha) of core and bark of Kenaf biomass obtained in fields with different
nitrogen application levels and in fields with different irrigation levels. Irrigation Level
204 mm H 2 0
253 mm H20
301 mm H 2 0
H2O 400 ~~l~ll
Nitrogen Fertilization 0 kg N/ha 75 kg N/ha 150 kg N/ha 0 kg N/ha 75 kg N/ha 150 kg N/ha 0 kg N/ha 75 kg N/ha 150 kg N/ha 0 kg N/ha 75 kg N/ha 150 kg N/ha
Core
Bark
Total
0.36 0.42 0.53 0.68 0.58 0.56 1.15 1.04 0.63 0.90 0.88 1.37
0.22 0.28 0.30 0.36 0.38 0.18 0.51 0.56 0.45 0.43 0.41 0.85
0.58 0.70 0.83 1.04 0.96 0.74 1.66 1.60 1.08 1.33 1.29 2.22
In terms of fibre content, there were no statistical significant differences among the plants obtained in fields with different levels of nitrogen and with different levels of irrigation. Then, in terms of the potential pulp production, differences among the fields were observed, namely among the different levels of irrigation, only due to the differences in the productivities. Highest values were obtained in the most irrigated fields, with 301 mm and 400 mm H20. No statistical significant differences were observed among levels of nitrogen fertilization.
92
A.L. Fernando et al.
Table 5. Potential pulp production (&a) of core and bark of Kenaf biomass obtained in fields sowed in two different dates, in fields with different varieties and in fields sowed with different plant densities.
Sowing
Variety
date
Tainung 2 26'h June
-
Everelades 41
Tainung 2 th
Everglades 41
Plant population 20 seeds/m2 40 seeds/m2 20 seedslm2 40 seeds/m2 20 seeds/m2 40 seedslm2 20 seeds/m2 40 seeds/m2
Core
Bark
Total
1.09 2.04 1.83 1.49 0.86 1.46 0.80 1.21
0.75 0.79 0.47 0.73 0.42 0.70 0.45 0.56
1.84 2.83 2.30 2.22 1.28 2.16 1.25 1.77
The potential pulp production was considerably higher in the fields sowed first than in the fields sowed fifteen days later, mainly due to the highest productivities obtained in those fields because in terms of the fibre content, there were no significant differences between plants obtained in fields sowed at two different dates. Although not statistically significant, higher potential pulp production was also obtained in the fields sowed with a higher plant population. This resulted from the higher productivities obtained in those fields because, in terms of the fibre content, no significant differences were observed between plants obtained in fields sowed with two different plant densities. The bark of Tainung 2 presented higher fibre content than the bark of Everglades 41. No differences were obtained between the inner core of both varieties. But, in terms of potential pulp production no significant differences were observed. In all the experimental fields, the bark presented lower quantity for pulp purposes than core, due to its lower productivity, but a better quality for pulp purposes due to a lower content in lignin. 4. Conclusions
Productivity was affected by the level of irrigation of the fields but not by the level of Nfertilization. Better productivities were obtained with the highest irrigation levels, namely 301 mm and 400 mm H20. Biomass quality was not affected either by the level of irrigation either by the level of N-fertilization. Productivity and biomass quality were affected by the sowing date but not by the plant population. Plants sowed earlier (26" June) presented better roductivities and better quality for fuel and pulp purposes than plants sowed latter (1 1ti? July). Tainung 2 and Everglades 41 presented similar productivities and in terms of mineral composition there were no differences between the two varieties. The inner core material of the two varieties, presented also, the same quality for energy and pulp purposes. However, the bark of Everglades 41 presented a higher calorific value than the bark of Tainung 2 and the bark of Tainung 2 a higher fibre value than the bark of Everglades 41. In the perspective of a sustainable growth and development, for the production of Kenaf in the Portuguese climatic conditions, fields should be irrigated with an amount equivalent to 50 % PET. Irrigation with 100% PET, although giving better productivities, can contribute to a depletion of the water resources. Because of the ability of kenaf to mobilize soil nitrogen, nitrogen fertilization does not need to exceed 75 kgha and should be adjusted with regards to the soil contribution possibilities.
Kenaf Potential in Portugal as Energy Feedstock
93
Still in the perspective of a sustainable growth and development, there is no point on choosing the highest plant population density (40 plants/m2). In fact, no significant differences were observed in terms of the productivities, the potential energy production and the potential pulp production, in relation to the fields with the lower plant density (20 plants/m2). Overall, higher plant density fields are more costly, economically and environmentally, due to the high market price of the seeds and due to the higher content of fossil fuel necessary for its production. Considering the potential energy production and the potential pulp production, any of the varieties studied can be used, since no significant differences were observed in terms of this production parameters. Considering the results obtained, fields should be sowed, at latest, at 26'h June. A later sowing can compromise the productivities and, therefore, the net economical and environmental gain of this crop. Sowing at middle of May can be considered as a good hypothesis to be studied and applied. Acknowledgments
This work was supported by the European Union (Project QLK5-CT-2002-01729). References
[I] El Bassam N. (1998) Energy Plant Species. James & James (Science Publishers) Ltd. [2] Manzanares M. et al. (1993) Yield and development of kenaf (Hibiscus cannabinus L.) crop in relation to water supply and intercepted radiation. Biomass and Bioenergy 5 (5): 337-345. [3] Scott AW. (1990) Kenaf research, development and commercialization. Proc. from the Association for the Advancement of Industrial Crops: 27-33. [4] Alexopoulous E. et al. (2003) Minutes of the 2ndBiokenaf Meeting, Athens, October. [5] Fernando A. et al. (2004) 1st Individual progress report of FCT/UNL to the Biokenaf project.
This page intentionally left blank
IMPLEMENTATION OF ANAEROBIC DIGESTION PLANTS IN BREWERIES - DIFFICULTIES AND BENEFITS G. PESTA, R. MEYER-PITTROFF Chairfor Energy and Environmental Technologies ofthe Food Industvy, Center of Life Science Weihenstephan, Technische Universitaet Muenchen, Germany The anaerobic microbial conversion of organic matter into biogas is state of the art. Organically high loaded wastewater is treated preferably by anaerobic digestion (AD), e.g. wastes and wastewater of the food industry. Breweries generate not only above-average amounts of wastewater and organic wastes but also a very inhomogeneous wastewater. Treating a brewery’s wastewater presumes knowledge of its origin and composition to optimise the digestion process. Principles of anaerobic biological wastewater treatment are illustrated as well as typical factors for operating a biogas plant in breweries. These factors can be influenced only in a certain range. In general the implementation of AD plants in breweries shall be accompanied by restructuring the energy and wastewater management. Especially the utilisation of biogas respectively surplus heat asks for an intensive consideration of the production process. In Germany changing legislative restrictions and acts on granting priority to renewable energy sources offer thought-provoking impulses.
1. Introduction During the last years legislative restrictions concerning waste treatment and environmental protection were aggravated. Among others companies of the food industry were forced to think their waste-management over, including wastewater management. It became not only unpopular to produce organically high loaded wastewater but also very expensive. The organic carbon freight is numerical characterised by the Chemical Oxygen Demand (COD; unit: mg 02/1). In general the COD is basis for calculating wastewater fees. Especially the beverage industry, e. g. breweries, represents an extreme water consumptive branch of industry. Besides its exceeding use of fresh-water it produces aboveaverage amounts of wastewater. 2. Principles of Biological Wastewater Treatment
2.1. Aerobic Treatment with Anaerobic Stabilisation of Sludge Biological wastewater treatment in municipal sewage plants is usually an aerobic process. To stabilise the sludge surplus anaerobic digestion (AD) processes are widely used. Symbiotic living bacteria convert under anaerobic conditions organic carbon predominantly into methane and carbon dioxide, so called biogas. Fig. 1 shows a carbon mass balance of aerobic wastewater treatment with anaerobic treatment of sewage sludge. According to the carbon-flow-scheme about 50 % are lost by producing worthless C 0 2 and about 50 % by producing biomass and sewage sludge respectively. After stabilising the sludge surplus by AD or drying, the sludge residual needs to be treated adequate-causing additional costs.
95
G. Pesta & R. Meyev-Pittroff
96
The utilisation of the AD byproduct biogas in combined heat and power units (CHP) is technical and economical feasible: in general electric power as well as the heat is directly used in the sewage plant, latter to temper e. g. the AD fermentation tank. waste water influent 100 % organic C
28 - 36 Yo 50 % gaseous phase 1
\' return water
1
influent
1
1
sludge residual
mmRoBIc
1
Fig. 1. Carbon mass balance of aerobic wastewater treatment
2.2. Anaerobic Treatment and Biogas Production
The organic carbon freight of wastewater is rather a valuable energy source than a worthless wastewater component. Due to this fact it is obvious to obtain a benefit by cleaning wastewater: to provide an energy source. Instead of spending additional energy to reduce carbon freight by aeration it is reasonable to produce biogas, a renewable energy source. Fig. 2 shows a carbon mass balance of anaerobic wastewater treatment. Treating wastewater by AD converts more than 90 % of its carbon freight into biogas. Operating a biogas plant is feasible for concentrated wastewater with an COD higher than 3 500 mg Od1. To reduce the COD of the fermentation residues finally to a runoff ditch level, an advanced effluent treatment by aeration is still necessary. The degradation by anaerobic bacteria takes place within two main steps: hydrolysis (decomposing of organic material) and methane-genesis (production of biogas). The degree of degradation as well as biogas quality can be affected in various ways by technological and procedural measures. The advantage of a two-step fermentation system is the possibility to exert influence on single degradation processes, hydrolysis and methane-genesis. Each step takes place in a digester on its own.
Implementation of Anaerobic Digestion Plants in Breweries
97
In general two-step fermentation systems are preferable used for inhomogeneous substrates differing in their biodegradability. Microorganisms always degrade substrates first, of which assimilation or dissimilation takes less energy. At a predicted dwell time waste water influent ^ ^
-.
. -
CH,, COz BIOGAS
COZ
A
A
I
1-3 -%
I
effluent 1 % organicc
v sludge residual
influent
ANEROBIC
AEROBIC
Fig. 2. Carbon mass balance of anaerobic wastewater treatment.
and a continuous feeding of a mixture of easy and hard degradable substrates the microorganisms utilise the easier degradable one first. The hard degradable substrate will be utilised only fragmentary or even not at all. Two-step fermentation systems reach up to a COD-reduction of 98 %, even with inhomogeneous substrates. AD is not only used to stabilise sewage sludge. AD offers an alternative waste management for a wide range of substrates. Besides the channeling of waste treatment the production and usage of biogas is a primarily aspect. Basically any liquid or solid organic wastes from e. g. food and agricultural industry or municipal waste collections, like whey, slaughterhouse wastes, flotation fat, manure and energy crops-just to name a fewcan be utilised for AD. Approximately 1 700 biogas plants in Germany (2003) with a total capacity of 140 MW demonstrate the technical and economical feasibility of AD of organic wastes and wastewater.
3. Characteristics and Use of Biogas Biogas is a mixture of mainly methane (CH4, 50-85%-vol.), carbon dioxide (C02, 1550%-vol.) and trace gases, e. g. water (H20), hydrogen-sulfide (H2S) or hydrogen (H2). Dewatering and the removal of H2S are necessary before biogas is utilisable.
98
G. Pesta & R. Meyer-Pittroff
The calorific value (c. v.) of biogas depends on its CH4 content and varies between 4 and 7.5 kWhimN3.Biogas with 65 %-vol. CH4 has an average c. v. of 6.0 kWh/mN3, which is equivalent to 0.62 1 fuel oil. Biogas is used in CHP-Systems to generate electric power and thermal energy. It substitutes natural gas in gaseous fuel steam vessels to produce steam. Besides reducing the consumption of fossil energy carriers the use of biogas made out of biomass reduces also C02-emissions. In addition to the C02-trading guidelines biogas has an emissionfactor of 0 tons C02perTJ-compared to fuel oil with 78 t and natural gas with 56 t C 0 2per TJ. In Germany the Act on Granting Priority to Renewable Energy Sources (Renewable Energy Sources Act, EEG) is one of the key elements to enlarge the use of renewable energy sources, following the national targets in the EU directive [4]. The EEG regulates the feeding into the grid and compensation of electricity from renewable energy sources. The EEG obliges grid operators to purchase electricity generated from renewable energy sources and to pay minimum compensation to the generators of this electricity. The amount of compensation is based on the costs that arise from the generation of electricity to facilitate the economic viability of installations where electricity is generated from renewable energies. Depending on the installed electrical capacity of the CHP between 0.089 and 0.1 15 Euro per kWh are paid in 2004. Additional grants are paid for using CHP or innovative energy techniques. 4. Anaerobic Digestion of Brewery’s Wastewater
4.1. Characteristics of Brewery’s Wastewater At an average brewery 240-750 1 wastewater emerge form the production of 100 1 beer. Table 1 shows some characteristics of brewery’s wastewater and also points up the wide range of these figures, depending on the state of the art of the brewery. It is obvious that a consistent water management reduces fresh water consumption as well as wastewater production. The additional benefit is an increased specific organic load caused by an effect of concentration. In general wastewater from breweries is very inhomogeneous. Depending on the step within the brewing-process various parameters are changing, e. g. quantity, pH, temperature, organic load, solids, cleaning and disinfecting supplies’ content. The knowledge of wastewater’s origin is necessary for implementing an absolutely reliable wastewater management. E. g. separating most of the wastewater lines and flows leads to different organic freight concentrations in each fraction. This supports a constant quality of wastewater for the digestion process by blending the single fractions.
Implementation of Anaerobic Digestion Plants in Breweries
99
Table 1. Characteristics of brewery’s wastewater (SB = Sales beer) [l-31
Special waste treatment, e. g. dewatering of brewer grains, increases not only the organic freight but also the production of biogas. Although wastewaters of breweries do have a high organic load they can be characterised as easy degradable. This fact is shown in the ratio of COD and BOD, which ranges between 1.5 and 1.&similar to municipal wastewater. Because of its valuable and easily available carbon-freight this kind of wastewater is very welcome in municipal sewage plants. Despite these facts the amount of wastewater fees depends on the COD. Without serious negotiations wastewater fees will be extremely high. Nevertheless the anaerobic digestion of wastewater offers its benefits to reduce these costs as well as energy costs by reducing the consumption of fossil fuels. 4.2. Digestion Systems for Brewery’s Wastewater
A two-step fermentation system with an upstream blending- and buffering-tank proved to be an appropriate system for a comprehensive treatment of all wastewater flows in a brewery. Fig. 3 shows an example of an anaerobic digestion plant for treating brewery’s wastewater. f
biofilters wastewater influent
& J.
buffer- and mixing-tank
J.
hydrolysis
biogas utilisation
sludge concentration
advanced effluent treatment aeration
sludee drvine
Fig. 3. Anaerobic sewage plant for treating brewery’s wastewater [l].
100
G. Pesta & R. Meyev-Pittroff
After separating solid and extreme grainy contraries (e.g. pieces of glass, crown caps) the wastewater flows into a blending- and buffering-tank. This provides for a constant wastewater composition, which is necessary for a steady-going feeding of the fermentation tanks. In the buffering tank a lively microbial activity is already apparent - more facultative anaerobic than anaerobic. For example visualises a sulphur crust above the fluid surface this activity. Sulphur oxidising bacteria are responsible for this crust, literally called “sulphur-flower” (Schwefelblume). Additional indicators are the increasing odour emission and the beginning of acidification - a degradation step within the hydrolysis. Bacteria of the hydrolysis digester degrade organic matter to short chained fatty acids, ammonia, hydrogen sulphide, phenolic carboxylic acids and hydric fatty acids. Adjusting the pH between 5.6 and 6.5 encourages the activity of acid forming bacteria and represses activity of methane forming bacteria which also exist in there. Biogas produced in the hydrolysis digester consists of 85-90 % Carbon dioxide (COz), 5-10 % Methane (CH4) and various olfactory active substances. Biogas out of hydrolysis is preferred cleaned by biological filters or thermal processes. Afterwards the acidificated wastewater flows through the methane generating digester. Biogas is produced by methane forming bacteria. Besides current UASB-Reactor-Systems (Upflow-Anaerobic-Sludge-Bed) the two-step-fermentation-system“Spradau-Schraufstetter”provides long time reliability and availability (Fig. 4) [2]. The Spradau-Schraufstetter fermentation system consists of a hydrolysis digester enclosed by the methane generating digester with solid-state fermentation. The influent flows through the reactors by a hydraulically connected piping system. Substrate’s degradation up to 90 YOcan be achieved. hydrolysis gas (biofilter) ~~~
influent
effluent
Fig. 4.Scheme of the two-step fermentation system “Spradau-Schraufstetter” for treating organically high loaded wastewaters [2].
Implementation of Anaerobic Digestion Plants in Breweries
101
This fermentation system allows a very compact construction and requires very low specific space. Especially organic high loaded wastewater with a COD of 15 000 to 70 000 mg 02/1 can be handled easily, e. g. distiller’s grain - if a distillery is attached to the brewery. Independent of the fermentation system an advanced effluent treatment by an aerobic process ensures the COD reduction below the limit value for special wastewater fees or the runoff ditch.
5. Integration of Thermal and Electric Energy In breweries there are two possibilities for the utilisation of biogas: the production of steam (steam vessel) or the production of thermal and electric energy (CHP). Fig. 5 shows exemplary the flow-scheme of wastewater and energy in a 2-mio-hlbrewery. Electric and thermal energy are totally used within the brewery’s production and waste management plants. Biogas in steam vessels substitutes natural gas. Combined burners for biogas and natural gas are state of the art. The investment costs are low, in comparison to a CHP concept. The alternative utilisation of biogas with a CHP provides not only thermal energy by using waste heat, e. g. from exhaust gas. Electric energy is generated, covering the companies basic load and cutting the peak demand or feeding the grid. In Germany the EEG regulates the compensation fees for feeding the grid with electric energy. In certain applications feeding the grid provides additional grants for the brewery. advanced effluent treatment/ aeration
. . l . . _ . .
I
lectric e n e r g y 215 kW surplus beat 375 k W 7 5 - 9 0 OC
Bottling plant canslkegslbottles 7 9 5 m’/d
__
._. -,“.-
I
anaerobic treatment
biogas
Brewhouse I steam 600 kW Brewhouse II 4 2 5 m’ld
CHP or steam vessel
W o r t cooling 2 5 0 m’ld
(thermal orland tlectric e n e r g y :
F e r m e n t a t i o n cellar 5 0 m’ld
Bright beer tanks
I
1
S t o r a g e cellar 1 5 0 m’ld Filtration cellar 400 m3/d
Fig. 5. Scheme of the wastewater and energy management of a 2-mio-hl-brewery [l].
102
G. Pestn & R. Meyer-Pittvoff
Concerning the C. v. of biogas using a CHP reclaims about 30 % in electric energy form and 55 % in thermal energy form. An amount of 30 % of thermal energy is used for tempering the digesters of the biogas plant. Surplus heat is integrated into the production process or can be sold e. g. by feeding a local heat grid. Absorption chillers are an additional alternative for surplus heat utilisation. Using CHP-systems with special cooling systems provide surplus heat with temperatures higher than 100 "C and opens further possibilities €or heat utilisation in the brewery.
6. Specific Problems of Anaerobic Digestion in Breweries An AD-plant in a large brewery with an annual production of one Mio. hl sales beer depicts an investment of 3-8 Mio. Euro. Therefore it is necessary to straighten out the demands on a wastewater management as well as to set the stage for a reliable process. On target are a constant quality of biogas, no odour emissions, highest degradation degrees and following the lower limits for discharging. Failing or deficiency leads to higher costs and a negative company's image. Running an AD plant is an extension of the brewing process and demands engagement and knowledge of the process. To ensure a constant quality, besides technologically and procedurally excellent equipment, trained staff and accompanying documentation are necessary. 6.1. Definition of Detergents and Disinfectants Clearness should exist on the composition of detergents and disinfectants used in the brewery. Absolute acids or bases are microbial inoffensive. Exceeding use of phosphoric acid may result in a phosphor-elimination-step before discharging the effluent. Sulphur acid increases the production of sulphuric substances and H2S under anaerobic conditions. Tenside solutions may affect a stable process negative. Nevertheless, AD plants have to be able to handle and tolerate detergents and disinfectants in a certain range. Therefore a dialog between the biogas plant's engineer, the detergent producers and the brewery is necessary prior to the engineering of the biogas plant. 6.2. Supervision of the Siogas Plant
The times of placing the biogas plant in operation and starting the process up are to short for transferring specific knowledge on the process to the operator of a biogas plant. Especially knowledge on digestion problems including accurately timed identification and trouble shooting is essential. Long term support by the plant constructor and engineer should be given to the operator.
Implementation of Anaerobic Digestion Plants in Breweries
103
6.3. Adjusting the pH- Value of the Hydrolysis Digester A pH-value of the hydrolysis digester between 5.5 and 6.5 is proofed to be favourably for an optimal degradation effect. The organic load of brewery’s wastewater acidifies very easy and fast, resulting in a pH-value of 4 and even lower. Concentrated pure bases are added to level the pH-value. The dosage of bases may effect the salt freight of the effluent. An increased concentration of sodium may additionally cause metabolic disruptions of the bacteria. A well-balanced equilibrium of alkali- and earthy base-metals prevents these problems. If it is necessary to ad base for pH-adjustment, a specified mixture of sodium (NaOH), potassium (KOH), calcium (added as CaC03) and magnesium (added as MgCO3) should be taken. The dosage of alkali may be supplemented by using partial recirculated effluent. This process water is saturated with COZ and its soluble salts, particularly bicarbonates. This measure increases the systems buffer capacity and helps to economise the application of base. In certain cases magnesium may increase complexation and precipitation. Experiments will verify the extent. 6.4. Biological Cleaning of Extracted Air with Biological Filters During the wastewater cleaning process several flows of extracted air, loaded with vapour and various organic substances are generated from buffering- and mixing-tanks, hydrolysis digester and advanced effluent treatment. Some examples of organic substances are alcohols, aldehydes, esters, ketones and dimethyl sulphide (DMS). Among others the cleaning with biological filters is possible. If the biological filter works proper, the organic load of the extracted air will be degraded under aerobic conditions into C 0 2 and water. Sulphuric substances will be oxidised until elementary sulphur - which has to be removed regularly. Ammonium and phosphor should be added because these nutrients are not available in the extracted air. A lack of oxygen has to be corrected by adding fresh air or discharged air form production rooms. To keep the biological filter moist sometimes extra water might be necessary.
6.5. Thermical Utilisation of Extracted Air Besides cleaning with biological filters olfactory loaded extracted air might be thermically utilised. The extracted air can only substitute a certain amount of combustion air because of the danger of coating within the steam vessel. 6.6. Danger of Infection of the Brewery’s Production Process by Anaerobic Bacteria In general the danger of infections of the brewery’s production process by Pectinatus and
104
G. Pesta & R. Meyer-Pittroff
Megasphera are well known. Both species are detectable in biogas plants. The danger of infection of the product lines by bacteria out of the biogas plant is not higher than in a brewery without wastewater treatment. It is presumed that the biogas plant is operated appropriately and accurately. Anyway, product lines are generally closed lines and are separated from the AD-facilities. Additionally the AD facilities are operating with a constant slight lower pressure, because of exhausting air. Bacteria can hardly attain outward.
7. Conclusions Anaerobic digestion is state of the art in a wide range of applications. AD of breweries wastewater combines two effects: On one hand it depicts an effective wastewater treatment to reduce wastewater fees. On the other hand AD produces a renewable energy source, biogas, which is more than a simple byproduct. Biogas fired in steam vessels or combined heat and power units, or even in combination of absorption chillers, helps to substitute fossil fuel sources. Proofing economical feasibility is not only done by calculating an biogas plant. It is also necessary to elaborate a total energy consumption and energy distribution scheme. Each application needs to be treated as a unique project with a unique general framework. This ensures a customised biogas plant design and absolutely reliable processing. Acknowledgements
The authors would like to thank the Stiftung Nagelschneider, Muenchen (Germany), for the financial support of this project. References
[I] Behmel U. (1993) Mehrstufige Methanisierung von Brauereireststoffen. Technische Universitaet Muenchen, Lehrstuhl fiir Energie - und Umwelttechnik der Lebensmittelindustrie,Dissertation. [2] Pesta G. (2003) Anaerobe Behandlung organischer Reststoffe der Lebensmittelindustrie - Vergaerung von Treberpresssaft. In einer Biogasanlage. 36. Technologisches Seminar Weihenstephan. [3] Scharf P. Energieversorgung eines Brauereibetriebes mit Brauereisreststoffen. Lehrstuhl f i r Energie-und Umwelttechnik der Lebensmittelindustrie, Dissertation. Technische Universitaet Muenchen. [4] Act on Granting Priority to Renewable Energy Sources (Renewable Energy Sources Act, EEG). 01.04.2000, Citation date: 01.01.2004.
APPLICATION OF BIOFUELS TO COMPRESSION IGNITION
ENGINES A. KOWALEWICZ Technical University of Radom, Radom, Poland A review on application of alcohols and biofuels, especially crude vegetable oils and their esters has been
carried out. Dual fuel engine, fuelled with diesel fuel (DF) and alternatively with rapeseed oil methyl ester ( M E ) as a main he1 and bioethanol injected during inlet stroke into the inlet port as supporting burning of RME was developed and investigated. The concept of ethanol injection consists in this, that after ignition by burning D F / M E droplets, ethanol-air mixture bums very quickly and promotes combustion of main fuel droplets, resulting in shorter combustion period. Experiment was carried out with the use of one-cylinder direct injection diesel engine adapted for ethanol injection. Investigations were performed at two constant loads, for which the ratio of ethanol and base fuel were changed: high and low, at each load engine was run at three speeds for three injection timings. Results of investigation showed considerable decrease of C 0 2 and smoke level for all operating conditions of the engine load at high load also CO and HC emission, while at low load NO, emission. Optimum ratio of ethanol energy to both fuels energy on account of brake fuel conversion efficiency and benefit of emission decrease was found to be about 20 for DF and 25 percent for M E . ~
~
~
1. Introduction For the last two decades a worldwide trend of biofuels application to internal combustion engines has been observed. These fuels are mainly vegetable oils and alcohols. However, vegetable oils have very high viscosity in comparison with that of diesel fuel (DF) and therefore they did not commonly enter the fuel market. They are base materials for fabrication of fatty acid methyl esters (FAME), which have much lower viscosity in comparison with crude vegetable oils, comparable to this of DF and/or added to diesel fuel as fractions. Rapeseed oil methyl ester ( M E ) and alcohols are these renewable biofuels which have been used for some time in Western Europe, South Asia and South and North America. Presently they will be used in Poland as components to diesel fuel and gasoline, respectively.
1.1. Alcohols 1.1.1. Methanol A lot of work was carried out on fuelling SI and CI engines with methanol, also in Technical University of Radom. As far as CI engines are concerned, Kowalewicz and Luft [ 11 developed, in Technical University of Radom, the dual-fuel engine fuelled with methanol vapour (methanol evaporated in the engine cooling system and was fumigated to inlet air) as a main fuel and DF as a pilot ignition fuel. For high load brake fuel conversion efficiency of this engine was higher than base diesel engine. However, methanol still remained as a typical fuel for SI engines. Although methanol may be fabricated from biomass, its production is expensive (more cheaper is fabrication of methanol from coal). In the last years, the interest of alcohol as a fuel focussed on ethanol (bioethanol especially). It is worthy to remind, that Henry Ford’s cars operated on ethanol fuel (1 880).
105
A . Kowalewicz
106
1.1.2. Ethanol Ethanol is a good fuel to spark ignition engines, but its application to compression ignition (CI) engines meets with many troubles. The European Commission expects that bioethanol will be introduced into European markets as blends with gasoline and diesel fuel, as E85G (85% of ethanol in gasoline) and E15D (15% of ethanol in diesel fuel) [2]. Scania S.A. launched a couple of city-busses which are fuelled with the blend of 5% (by vol.) of ethanol and 95% of diesel fuel [3]. 1.2. Vegetable Fuels
Vegetable fuels, on account on their physico-chemical properties, mainly high cetane number and low temperature of selfignition, may be applied to CI engines.
1.2.1. Raw Vegetable Fuels and their Blends with DF Raw vegetable oils, which can be considered as fuels are: rapeseed oil, olive oil, soybean oil, palm oil, sunflower oil and maize oil. Their properties as potential fuels to CI engines are in high degree similar (see Table 1). Table 1. Physical and chemical properties of some vegetable oils.
Carbon content,
Oxygen content,
At first, in early seventies of the last century, crude vegetable oils were tested as fuels to CI engines. On the base of numerous experiments, it was recognized, that these oils have -in comparison with DF fuelling-longer injection time, shorter ignition delay and longer combustion time [4-61. Their short-term combustion in CI engines is feasible. However, long term test showed faster than for diesel fuel engine degradation: wear of injection pump, carbon deposits on valve seats, injector cocking and valve steams, deposits built up in combustion chamber and piston land [7, 81. Engine performance was considerably deteriorated: engine power decreased and brake specific fuel consumption increased [7].
Application of Biofuels to Compression Ignition Engines
107
Similar results were obtained when waste cooking oil was applied to DI CI engine: injector cocking and piston rings sticking, fuel system clogging and carbon deposits in the combustion chamber [7]. In general, in the case of fuelling CI engine with raw vegetable oils in comparison with DF fuelling, power and torque decrease; 0 engine efficiency decreases. In order to improve performance, the beginning of injection should be earlier. Decrease of performance is a result of lower heating value, worst atomisation of the fuel and worst combustion [8]. Raw vegetable oils have also very high temperature of cold start [9], so heating of the fuel for start is necessary. According to the investigations carried out in [lo], heavy duty engines can be fuelled with the mixture of 30% RO and 70% DF (only more often injectors should be cleaned). In conclusion, raw vegetable oils can be used as fuels under the following conditions: they are components of DF (up to 30%) andlor they are transesterified (FAME). Vegetable fuel which undergone most of tests is rape oil methyl ester ( M E ) . 1.2.2. Rape Oil Methyl Ester (RME)
A lot of work has been done in Europe (France, Austria, Germany, Poland) and America since early nineties on application of RME to CI engines, e.g. [ 10-131. Summing up the results of numerous experiments carried out on fuelling with neat RME andor its blends with DF it can be stated that: RME as individual fuel (neat) may be applied to CI engines but inspection of fuelling system, especially injectors, should be carried out more often [ l 11; brake fuel conversion efficiency of the engine fuelled with RME is a little better than for fuelling with DF, mainly for middle speed; 0 ignition delay of RME fuel and its blends with DF is shorter than neat DF; 0 all emissions (except of NO,) of the engine fuelled with RME are lower than for fuelling with DF; wear of the engine is higher for fuelling with RME than for DF but, after of some plastic and rubber elements, is comparable to that for DF; exchange 0 as far as mixtures of RME and DF are concerned, 30% fraction of RME demands no exchange of engine elements, does not influence engine performance and emissions, even improves them. 2. A Novel Concept of an Ecological Engine
This work concerns a novel concept of diesel engine, which was invented, developed and investigated in Technical University of Radom. The concept consists in dual fuelling: a base fuel, which-in the first approach-is diesel fuel (DF) and-in the second approachrape oil methyl ester (RME) is injected by the standard fuel system and additionally ethanol is injected into inlet port during suction stroke, in order to enhance burning of DF/RME fuel droplets. Before self-ignition of DF/RME ethanol forms with air homogeneous mixture. After ignition by burning DF or RME droplets (which have self-
A . Kowalewicz
108
ignited earlier), ethanol vapour-air premixed combustion accelerates burning of DF/RME droplets, and-as a result-combustion period is shorter. In such a way the total time from the beginning of DF/RME injection to the end of combustion may be shortened, resulting in smokeless combustion. At this moment it is worthy to mention, that while fuelling an engine with RME, which is biofuel, net production of greenhouse gas (COz) is equal to zero. Moreover, due to that ethanol has higher mass ratio of hydrogen to carbon (H/C ratio) in the molecule than DF, lower emission of CO2 is expected. 3. Objectives of the Work The main objectives of this paper are a follows:
to investigate whether ethanol has any effect on combustion in CI engine fuelled with DF and RME as a base fuel; to measure emissions and efficiency in function of the ratio of ethanol energy to both fuels energy for different engine operating conditions; to determine the optimum ratio of ethanol to DF and RME from the point of view of emission and engine efficiency; to determinate the optimum angle of beginning of the base fuel (DF and RME) injection from the point of view of emissions and efficiency for different engine operating conditions.
4. Engine Test Stand and Course of Investigation Engine data are shown in Table 2. Engine test stand is shown in Fig. 1. Engine torque was measured by means of the eddy-current dynamometer Vibrometer 3WB15. RME fuel consumption was measured with the use of the automatic dosemeter PG-80. Ethanol dose per cycle was measured indirectly by measuring the time of consumption of its definite mass. Air flow was measured with the use of a flowmeter installed on the inlet air surge tank which reduced pressure pulsation.
Sensor Fig. 1. Test stand.
Application of Biofuels to Compression Ignition Engines
109
Table 2. Engine data.
Type of the engine No. of cylinders Swept volume Compression ratio BorelStroke Max power Max torque Injection pump* Injector nozzle Orifice diameter Injection pressure
1HC102 (Polish production) 1 980 cm3 17 102/120 mm 11 kW at 2200 rpm 55 Nm at 1500 rpm plunger type pintile type 0,95 mm 13,2-14,2 MPa
*Standard fuel injection pomp was replaced by another one, giving higher fuel delivery
Investigation was carried out at two loads: 20 Nm and 40 Nm; at each load three speeds: 1200, 1800 and 2200 rpm and three injection timing of base fuel: 25, 30 and 35 deg BTDC. Measurement points were chosen in such a way, that the comparison of engine parameters and emission could be obtained for the same load but for different proportions of ethanol to base fuel (DF or M E ) . 5. Results and Discussion
5.1. Emissions Emissions were measured in the function of the ratio of ethanol energy to the energy of both fuels (base: DF or RME and ethanol) fiE for engine operating conditions mentioned earlier. The best results with regards to emission were obtained for greenhouse gas and smoke. For example, emission of carbon dioxide for two base fuels vs. fiE is shown in Figs. 2-5. Table 3. Physico-chemical properties of diesel fuel, RME and ethanol. Chemical formula Molecular weight, g/mol Density @ 20"C, kg/m3 Calorific value, MJ/kg Calorific value of stochiometric mixture, MJkg Heat of evaporation, kJ/kg Temperature of selfignition, K Stochiometric aidfuel ratio, kg air/kg fuel Lower flammability h, Higher flammability h h Kinematic viscosity @ 4 0 T , mm2/s Octane number: motor(M0N) research (RON) Cetane number Flame temperature, K Molecular composition (by mass) C H 0
-
-170 838 41,03 -
270 -500 14,s 0,98 0,19 2,97 ~
3,85
-400 13,6
665
2,06 0,30 4,58 -
~
I
89
58 -
0,870 0,130 -
0,775 0,121 0,104
0,522 0,130 0,348
A . Kowalewicz
110
Base fuel - DF, T = 20 Nm, n = 1800 rpm
+injection timing = 25 deg CA BTDC ai
& injection timing = 30 deg CA BTDC ru injection timing = 35 deg CA BTDC
10
X
0 C 0
E6 4
0
0.1
0.2
0.3
0.4
0.5
0.E
The ratio of ethanol energy to both fuels energy, Ql Fig. 2. Greenhouse gas emission vs. C ~ at E low load for engine fuelling with diesel fuel and ethanol. 14
s ai
0
12
10
x
.-0 0
8
[I:
0
E6 4
2 0
0.1
0.2
0.3
0.4
0.5
The ratio of ethanol energy to both fuels energy,
0.E
a,
Fig. 3. Greenhouse gas emission vs. C ~ at E high load for engine fuelling with diesel fuel and ethanol.
Base fuel - RME, T = 20 Nm, n = 1800 rpm
+injection timing = 25 deg CA BTDC -+- injection timing = 30 deg CA BTDC
s w-
0
injection timing = 35 deg CA BTDC
10
X
0
0.1
0.2
0.3
0.4
0.5
The ratio of ethanol energy to both fuels energy,
0.f
fi
Fig. 4.Greenhouse gas emission vs. C ~ at E low load for engine fuelling with RME and ethanol.
Application of Biofuels to Compression Ignition Engines
111
14
s d
10
$
8
0
+Injection timing = 25 deg CA BTDC
12
-t-
Injection timing = 30 deg CA BTDC
C 0
e s
0
4 2 0
01
02
03
04
05
01
The ratio of ethanol energy to both fuels energy, Q,
Fig. 5. Greenhouse gas emission vs. !& at high load for engine fuelling with RME and ethanol.
From these and other measurements it may be stated that, C02 emission: in comparison with the engine fuelled with DF, is lower for fuelling with RME; is higher at high load than at low load for any QE for both DF and RME; 0 decreases with increase of OEfor any injection timing and speed for both DF and ME.
The explanations of these results are as follows. RME has less carbon in the molecules than DF, hence less COZ emission for RME fuel. Decrease of C02 is a result of higher QE. Products of ethanol combustion contain less CO2 and more H20. Higher C 0 2 emission is a result of more fuel burnt at higher load. Smoke emission is shown in Figs. 6-9. Smoke emission: 0 in comparison with the engine fuelled with DF is lower for fuelling with RME at any load; at high load is higher than at low load for any QE for both DF and RME; decreases with increase of QE for any injection timing and at any load for both DF and RME; is the highest at the highest engine speed and for fuelling with neat DF or RME. As far as other emissions are concerned, the results are as follows. 20 Base fuel - DF, T = 20 Nm, n = 1800 rpm
18
+Injection liming = 25 deg CA BTDC
---+-Injection timing = 30 deg CA BTDC 14 12 10 8
6 4
2
1
0
0
0.1
0.2
0.3
0.4
0.5
0.E
The ratio of ethanol energy to both fuels energy, 0,
Fig. 6 . Smoke emission vs. 0, at low load for engine fuelling with diesel fuel and ethanol.
A . Kowalewicz
112
80
k
'
70
$ - 60
--
Base fuel - DF, T = 40 Nm, n = 1800 rprn
+Injectiontiming = 25 deg CA BTDC
,
Injection timing = 30 deg CA BTDC
50
E
40 30
E
v) 20
10
0
I
I
I
I
I
1
0.1
0.2
0.3
0.4
0.5
0.6
The ratio of ethanol energy to both fuels energy, 0, Fig. 7. Smoke emission vs. QE at high load for engine fuelling with diesel fuel and ethanol. 20 18
8
16
a
8
=- 14 0 '9 12 .$ 10
Y
E 6 m 4 2 0 0
0.1
0.2
0.3
0.4
0.5
0.6
The ratio of ethanol energy to both fuels energy, Rr Fig. 8. Smoke emission vs. C ~ at E low load for engine fuelling with RME fuel and ethanol. 90
80 70
60 50 40 30
20 10 0
0
0.1
0.2
0.3
0.4
0.5
The ratio of ethanol energy to both fuels energy,
0.1
a
Fig. 9. Smoke emission vs. !& at high load for engine fuelling with RME fuel and ethanol.
NO, emission is lower for engine fuelling with RME than with DF. It depends strongly on load, injection timing and the ratio of ethanol energy to both fuels energy RE, for two base fuels (i.e. DF and M E ) : 0 the higher the load, the higher NO, emission;
Application of Biofuels to Compression Ignition Engines
113
the later injection of base fuel, the lower the NO, emission; at the low load NO, emission decreases with increase of the ratio of ethanol (due to lower temperature level being the result of higher energy of ethanol evaporation) and at high load-vice versa (more heat is evolved due to higher load and the influence of ethanol evaporation on temperature is relatively lower). CO emission is generally higher for engine fuelling with RME than with DF and increases with ethanol addition. Only at part load for neat base fuel CO emission is comparable for both base fuels. The same tendency shows hydrocarbon emission.
5.2. Brake Fuel Conversion Efjciency Brake fuel conversion efficiency (BFCE) for each point of experiment was computed with the use of measurement of both fuel consumption, engine speed, load and calorific values of both fuels. BFCE depends strongly on load and speed; other parameters have rather slight influence on it (Figs. 10-13). From these and other (not shown in the paper) results is may be stated that: BFCE is higher for fuelling with DF than with RME for QZE < 0,4; 0 in general, the higher the load, the higher the BFCE for both DF and RME; 0 BFCE gets maximum at middle speed for both DF and M E ; the influence of ethanol addition on BFCE is not very strong, with exception of high speed and high load, when it increases with increase of ethanol addition for both DF and M E ; the influence of injection timing of DF/RME on BFCE is slight, with exception for late injection, when it increases with ethanol addition. 0.4 0.35 0.3
Ym
0.25
0.2 0.15
I
0
0.1
0.2
0.3
0.4
0.5
The ratio of ethanol energy to both fuels energy,
0.6
0,
Fig. 10. Brake fuel conversion efficiency vs. nEat low load for engine helling with diesel fuel and ethanol.
The first two results are clear: higher efficiency results in higher temperature, which assists the higher load and higher calorific value. The third one follows from the fact, that not all the fuel is burnt at high speed and heat losses are high at low speed. Positive influence of ethanol is visible only in the range of low BFCE at high speed-it promotes burning of DF/RME fuel in the case, when its combustion is not efficient due to too short time to be completed. The last result is also clear: BFCE is higher for earlier injection of the main fuel.
114
A . Kowalewicz 04 0.35
w
Y m Base fuel - DF, T = 40 Nm, n = 1800 rpm
+Injection timing = 25 deg CA BTDC
0.2
& Injection timing = 30 deg CA BTDC Injection timing = 35 deg CA BTDC
0 15 0
01
02
03
0.4
0.5
0.f
The ratio of ethanol energy to both fuels energy, Rl Fig. 11. Brake fuel conversion efficiency vs. C ~ at E high load for engine fuelling with diesel fuel and ethanol.
+Injection timing = 25 deg CA BTDC --W--
Injection timing = 30 deg CA BTDC
- - - ~ - r i - - - -lniection -~ timina = 35 dea CA BTDC
iJ m
03
0 25
0
0.1
0.2
0.3
0.4
0.5
The ratio of ethanol energy to both fuels energy,
0.f
Rl
Fig. 12. Brake fuel conversion efficiency vs. CIE at low load for engine fuelling with RME and ethanol.
0 35
6 '
k
0.3
0.25 Base fuel - RME,T = 40 Nm, n = 1800 rpm
+Injection timing = 25 deg CA BTDC --+-
Injection timing = 30 deg CA BTDC Injection timing = 35 deg CA BTDC
0.15
I
0
0.1
0.2
0.3
0.4
0.5
The ratio of ethanol enerqv to both fuels enerqv,
0.1
ni
Fig. 13. Brake fuel conversion efficiency vs. at high load C ~ for E engine fuelling with M E and ethanol.
Application of Biofuels to Compression Ignition Engines
115
6. Conclusions From the carried out experiment the following conclusions may by drawn: Injection of ethanol into inlet port decreases COZ, smoke and-in the case of high speed and low load-also NO, and at high load also CO emissions for fuelling as well diesel fuel (DF), as rape oil methyl ester (RME), at a good brake fuel conversion efficiency; Ethanol fraction in both fuels (i.e. base fuel and ethanol itself) at low load may reach 50%, less at high load and is designated by diesel-knock; Optimum injection timing of DF and RME on account of minimum NO, emission seems to be late injection-25 deg BTDC and for maximum brake fuel conversion efficiency-30 deg BTDC for the two base fuels; Optimum ratio of ethanol energy to both fuels energy for DF is 20% and for Rh4E-25%; More advantageous is fuelling the engine with rape oil methyl ester than diesel fuel, on account on lower emission of COz, smoke, and in some operating conditions, also NO,; Making use of measured indicator diagrams and computed diagrams of heat release rate and fraction of fuel burnt will focus more light on combustion processes in the engine, resulting in better understanding of the influence of ethanol addition on engine efficiency and emission; Application of rape oil methyl ester as a base fuel will contribute to decrease the share of fossil hydrocarbon fuels in fuel market.
Acknowledgements This work was supported by the grant of Polish Committee for Scientific Research No 5T12D02922. The author wish to express his thanks to the PPHU D.R. Rosiak & Rosiak Co., Poland for gratuitously providing the RME fuel used in experiments.
Abbreviations BFCE BTDC CA CI DF DI CI
brake fuel conversion efficiency before top dead centre crankangle compression ignition diesel fuel direct injection compression ignition
Nomenclature n T RE
engine speed, rpm engine torque, Nm the ratio of ethanol energy to energy of both base fuel (DF or RME) and ethanol
116
A . Kowalewicz
References [ 11 Kowalewicz A, Luft S. (1993) Design and performance of diesel engine fuelled with methanol vapour. 26th ISATA Proceedings, Aachen. [2] Brefing note No.07/2001 of European Parliament: Bioethanol added to fuel, 2001. [3] Boman G. (2001) Ethanol busses strengthen image. Scania World 2. [4] Lotko W. (1997) Zasilanie silnikdw wysokopreznych paliwami weglowodorowymi i roilinnymi. Wydawniciwa Naukowo-Techniczne Warszawa. [ 5 ] Szlachta Z. (2002) Zasilanie silnikdw wysokoprqznych paliwami rzepakowymi. Wydawnictwo Komunikacji i LqcznoSci, Warszawa. [6] Kowalewicz A, Lotko W. (1995) Performances and some combustion characteristics of C.I. engines with rape oil and its ester. CIMAC Congress, Iterlaken. [7] Bari S, Yu CW, Lim TH. (2002) The effects of fuel injection timing with waste cooking oil as fuel in D.I. diesel engine. Journal of Automobile Engineering 216 (D9). [8] Bari S, Yu CW, Lim TH. (2002) Performance deterioration and durability issues while running a diesel engine with crude palm oil. Journal of Automobile
Engineering 216 (Dl 1). [9] Szlachta Z. (1998) Cold startability of diesel engine fuelled with rape fuels. Polish Academy of Sciences - Cracow Branch. Commission of Motorisation, 13, (l), Krakow. [lo] Staat F, Gateau P. (1995) The effects of rapeseed oil methyl ester on diesel engine performance, exhaust emissions and long-term behaviour. A summary of three years of experimentation. SAE Paper 950053. [ 111 Strus M. (2001) Feeding of multicapacity diesel engines by plant origin fuels. I1 International Scientically-Technical Conference Explo-Diesel & Gas Turbine, Gdalisk-MiGdzyzdroje, Kopenhaven. [ 121 Sums T. (1996) Use of biofuels under real world engine operation. Proceedings of 2nd European Motor Biofuels Forum, Graz. [ 131 Zablocki M. (1 994) Zastosowanie paliwa rzepakowego do silnikow wysokoprGinych. Research project No. BP 903 1319101 of Committee for Scientific Research.
KINETIC STUDY OF FROTH FLOTATION FOR PET-PVC
SEPARATION E. AGANTE, T. CARVALHO, F. DURAO, A. PINTO, T. MARIAN0 CVRM- Centro de Geosistemas, Instituto Superior Tecnico, Portugal Froth flotation is a separation process widely used in mineral processing and it has been tested in the separation of plastics from post-consumer waste, namely PET and PVC. The complexity of this process and the scarce comprehension of the physical and chemical mechanisms involved in the plastics flotation make it an interesting subject of research. It is widely accepted that, although time consuming, a kinetic study is the most correct way to evaluate the flotation process. In the work presented here, a four-factor / two levels experimental design was carried out to study the effect of the variables depressor concentration, frother concentration, pH and proportion of plastic in the mixture, on the flotation rate parameters, amounting to z4 flotation tests plus replicated tests. The main and crossed effects of the above referred operating variables on the rate constants were quantified.
1. Introduction
Recycling is nowadays a common procedure in countries committed with sustainable development. Prior to recycling, convenient separation of materials that feed the different recycling industries is needed. In what concerns municipal solid wastes (MSW), some materials are relatively easy to separate from others by mechanical means. This is the case of metals which are separated by magnetic and electromagnetic processes. Other materials are more difficult to separate but their value is high enough to justify a considerable financial effort in population education for source separation. This is what happens with glass that, in several countries, for decades, is collected separately from other solid residues. Plastics are not included in none of these cases. Generally, from post consumer waste, only packaging plastics are recycled being commonly separated at the source and collected in a bulk of packages, composed by different types of plastics and, depending on the local policy, by other packages materials, as well. The recovery of each type of plastic from the mixture is not easy, being commonly hand sorted. As well, their value is very low competing with virgin polymers that still have very low prices. All the non-packages plastics and the packages plastics that are not conveniently presented to the worker performing the hand separation (because they are too small or dirty, for example), are rejected and mixed with the commingled MSW, to be disposed in landfill. Plastics like Polyethylene Terephthalate (PET) and Polyvinyl Chloride (PVC) that appear in considerable proportion in MSW are difficult to separate by mechanical means because their physical properties, like density, are very similar. However, because they have different chemical composition they are potential candidates to be separated by froth flotation. Froth flotation is a versatile and relatively cheap process widely used in mineral separation, where plants have high capacities and are automatic. It is based on the different behaviour of particles in the presence of an inter-phase air-water, which depends on the surfaces properties of the materials. Industrially, it is performed in a vessel containing the water medium. The mixture of particles and water (pulp) is conditioned with specific chemical reagents (in a suitable quantity) to promote selective coupling between particles and air bubbles. This medium is continuously injected with a flow of air to transport the particles. After collision with air bubbles, which move upwards to the top of the vessel, hydrophobic particles adhere to 117
E. Agante et a1
118
them and will be recovered as the floated product. Hydrophilic particles stay in the vessel, becoming the underflow product. The stability of the air bubbles is performed by addition of a controlled quantity of a determined frother into the vessel. PET and PVC, like most plastics, are naturally hydrophobic. As reported in literature (e.g. [l]), the selective flotation of these two plastics can be achieved by different methods. One of the methods is the selective adsorption on the plastic surface of a wetting agent, also called depressor, responsible to rend it hydrophilic ([2] [ 3 ] [4] [5]). However, it should be pointed out that the depression occurs only in a specific physical and chemical environment, so, a reagent can be a depressor of a certain type of plastic depending on the flotation environment. The flotation environment can be affected by some variables such as depressor chemical composition, concentration and conditioning time, frother chemical composition and concentration, pulp pH, solids concentration, impeller speed, air flow rate, particles grain size, type of water medium among others. The most correct way to analyze batch flotation data and to evaluate the effect of each flotation variable, such as chemical environment and operational conditions is through the build up of kinetic models ([6]). To study the influence of variables, the more economic and faster process is the implementation of a factorial plan of experiences ([7]). Flotation is a rate process as the particles are recovered in the floated product proportionally to their concentration in the pulp. It is commonly accepted that the recovery in the jloated product, RJt), of each solid component, as a function of the Jotation time, t, can be approximated by the solution of a first order rate equation, applied to the floatable fraction, (1)
The kinetic parameters R and k are, respectively, the ultimate recovery (floatable haction) and the first order rate constant. It is easy to show that the recovery in the nonjloatedproduct, R(t), is given by R ( t ) =(l-R,)+R,
exp(-kt)
(2)
Replacing (I-R,) by Ro (non floatable fraction) and by R 1 (floatable fraction), the more familiar function is obtained R(t)= R,
+ R, exp(-k
t)
(31
In the work presented here, a kinetic study was performed in order to determine the effects of some important operating conditions: depressor concentration, frother concentration, pH and feed composition on the flotation kinetic parameters of both plastics (PVC and PET).
Kinetic &u& of Froth Flotntionfor PET-P VC Separation
119
2. Experimental ~~~~~~~~~~~
2.1. samples The plastics samples used were composed by shredded virgin packages of PET and of PVC %he study was carried out using only plastics from packages, to delimit the study and because, for the moment, in Portugal, only packages are collected from the MSW for t, of using recycling. Because the separation processes are particle shape ~ e ~ e n d e ninstead plastic pellets, it was used s ~ e d d e dpackages. The packages used were virgin, i.e., non used, to prevent undesired and controlled influence of residues and dirtiness c o n t a ~ ~ a t ~ in o nthe s froth flotation study. The plastic flakcs of the irregular shapes were classified by conventio~~l screening and separated in samples with i-7BOpm -4mm grain size, The e x p e ~ e n t a ltests were performed rasing artificial samples constituted by mixtures of both plastics. The relative percentage of occurrence of PVC and PET in MSW from drop-off points, calculated in a sampling study (@I), was 14% and 8696, respectively. In previous studies it was found that the flotation behaviow: of' one plastic is affected by the presence of another plastic. Therefore it was determined in this study the effect of the c o ~ p o s ~ ~oi fothe n sample. Samples with a proportion of 5%PVC+95%PET (low level) and 2 5 ~ ~ P ~ ~ i (Kigh - 7 ~level) ~ ~were E Tcomposed.
2.2. Equipment The &oth flotation tests were carried out in a Lee& laboratory flotation cell, with 3 litpes volume (Fig. 1). On this equipment the impeller is driven from the bottom of the cell, leaving the top surface free &om turbulence. This cell has airflow rate and impeller's s p e d ~~gl&goB~. The ~ o ~ ~speed t ~ofo650 ~ qpm ~ lwas i ~ ~byothe scell~h ~ y ~ in order to prevent t u ~ ~ u ~ e ~ c e .
Fig. 1. Flotation device.
o
E. Agante et al.
120
2.3. Experiments A total of 16 flotation tests were carried out combining the low and high levels of the four operating variables (see Table 1) as well as 3 replication tests. All the flotation tests were carried out with the following variables kept constant: depressor chemical composition, conditioning time, frother chemical composition, impeller speed, air flow rate, particles grain size, percentage solids weight and type of water medium. The mixtures were conditioned with Na-lignin Sulfonate (NaLS), as depressor, for 5 minutes, and Methyl Isobutyl Carbinol (MIBC) as frother, for 1 minute before flotation. The air flow rate was 120 l/h. This is the value that combined with the rotational impeller speed of 650rpm guarantee a non turbulent regime inside the flotation cell. The % solids weight was 0.8. All these variables were tuned in previous studies. Five incremental concentrates were collected during increasing time intervals: 0.5, 1, 2, 4 and 8 minutes. The flotation products after filtration, drying and weighting were analyzed. The particles of PVC and PET in the concentrates and reject were separated by hand-sorting (allowed by the difference in colour of the plastics) and the recovery was calculated from the resulting weights. Table 1. Low and high levels of operating variables
Low High
NaLS (mgn) 20 40
MIBC (mgn) 10 20
pH
YOPVC in mixture
7.7 11
5 25
3. Results and Discussion
In all flotation tests, the mass fractions R ( t ) of PET and PVC remaining in the cell (nonfloated product), after t minutes of flotation, were well described by the equation (3). The rate equation parameters were estimated by solving a non linear least- squares problem. The objective function is a weighted sum of squared deviations between the observed and the calculated, by equation (2), mass fraction, remaining in the cell at t flotation time, subject to the linear equality constraint RO+R1 = 100 and to the linear inequalities constraints Ro 20, R, 2 0 and k 2 0 . The number of total parameters estimated was 3. Fig. 2 depicts an example of the fit obtained in one test, showing the satisfactory adjustment. The effects of processing operating variables on rate equations parameters were studied by doing a linear regression analysis. Incomplete second order polynomials, including only main and interaction terms were fitted to the experimental results describing the variations of rate equation parameters with the operating variables. The independent variables were normalized in order to permit the comparison between the coefficients of the polynomials. It is worth to remind that R,, and R1are linearly dependent. Therefore, it is only necessary to retain one of these variables. In this case, the polynomial equation for & is presented which has the higher multiple correlation coefficient. The polynomial equations obtained for PVC are the following:
Kinetic Study of Froth Flotation for PET-PVC Separation
& = 0.60+0.12D-0.04F-O.16pH+0.06C-0.06FC+O.O7FpH (Multiple correlation coefficient Rz=0.96) k = 1.78+0.28D+0.05F-0.48pH-0.25C-0.36FC+0.23FpH (Multiple correlation coefficient R2=0.96)
The polynomial equations obtained for PET are the following: &=0.68+0.0 1D-0.03F+0.04pH+O.02C+O.07DpH-0.04CpH (Multiple correlation coefficient R2=0.92) k = 1.01+0.01D-0.13F-0.16pH-0.13C-0.09FC-0.09FpH (Multiple correlation coefficient R2=0.93)
$
0
1
2
3
4
5
6
7
8
8
B Fig. 2. Fitting between actual (symbol) and predicted (line) values for test performed with 40mg/l of NaLS, 20mg/l of MIBC, pH equal to 7.7 and feed content of 25%PVC+75%PET.
In the equations above the factors / variables retained are represented by D (depressor concentration factor), F (frother concentration factor), pH (PH factor) and C (composition of feed factor, in terms of PVC content). The main effects are the terms involving only one factor. The effects of the interaction between variables are labelled with the combination of the designations of both factors; e. g. the effect of the interaction between depressor concentration and pH is DpH. The fit between actual and predicted values are quite high (mainly for PVC), as can be observed in Figs. 3-8.
E. Agante et al.
122
The analysis of the polynomials (4) to (7) permits the following interpretations:
Effect of Depressor Concentration Depressor concentration effect is considerable in the flotation of PVC in terms of recovery as well as in terms of rate constant. In the case of PET, this is the variable that less influences both parameters. This is not surprising in the case of rate constant but it is contrary to what would be expected for recovery. One explanation for the difference of influence of this variable in the kinetic parameters of PVC and PET is that, within the range of the variables studied, the depressor concentration would be enough to promote the depression of PET while the increase of depressor concentration would have also the effect of depression of PVC. It should be noted that the interaction of this variable with pH represents the major effect on the recovery in PET flotation. PVC
Parameter R, adjust to kinetic
Fig. 3. Fitting between actual and predicted values of & to PVC.
Kinetic Study of Froth Flotation for PET-PVC Separation
I
PVC
Fig. 4. Fitting between actual and predicted values of R, to PVC
C
I
PVC
Parameter k adjust to kinetic Fig. 5 . Fitting between actual and predicted values of k to PVC.
123
E. Agante et al
124
PET
20.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Parameter R, adjust to kinetic
Fig. 6 . Fitting between actual and predicted values of Ra to PET
Effect of Frother Concentration For PVC, fiother concentration is the variable that less influences the flotation kinetics parameters. Its interaction effects with pH and feed composition are considerable. In the flotation of PET, frother concentration is a variable that influences significantly both parameters.
I
PET
Fig. 7. Fitting between actual and predicted values of R,to PET
Kinetic Study of Froth Flotation for PET-PVC Separation
125
PET
Parameter k adjust to kinetic Fig. 8. Fitting between actual and predicted values of k to PET.
Effect of pH pH is the key variable to the flotation of both PET and PVC. Its effect is the highest for both recovery and rate constant on both plastics. Besides, the interaction effects with other variables are high as well.
Effect of Feed Composition As expected, the feed composition influences the parameters of PVC and PET flotation kinetics. Its interaction with frother effect is quite high in the flotation of PVC and in the rate parameter of PET. In the PET flotation the interaction of feed composition with pH effect is significant for recovery while for the rate constant the interaction of feed content with frother is considerable. 4. Conclusions
In the study presented here the effects of some important operational variables on flotation kinetic parameters, recovery and rate constant, were quantified. It was concluded that the variable that most influenced the flotation kinetics was pH. The concentration of the depressor didn’t influence significantly the kinetic parameters of the flotation of PET but it has a considerable effect on the parameter kinetics of PVC flotation. The frother concentration and the feed content effects were also quantified.
126
E. Agante et al.
Within the range of operational variables experimented, it was not possible to attain complete selectivity. In average, the rate constants don’t differ considerably and the floated and non floated percentages are approximately equal in both plastics. One explanation for this fact is that the depressor is not the most appropriate. Therefore, the study will continue in the search of a more convenient reagent, taking into account the conclusions of the present study. Acknowledgments
The authors acknowledge FCT (Fundagao para a Ciencia e Tecnologia) for the financial support of the study (project number POCTI/ECM/40799/2001) and Sulpet and Vitor Guedes companies for the samples supply. References [ 11 Fraunholcz N. (2004) Separation of waste plastics by froth flotation - a review, part
I. Minerals Engineering 17: 261-268. [2] Carvalho MT, Durao F, Felicio A. (2003) Separation of plastic from municipal solid wastes. In MG Carvalho & NH Afgan (eds.), New and Renewable Technologies for Sustainable Development: 459-468. Swets & Zeitlinger Publishers, Netherlands. [3] Shibata J, Yamamoto K, Pradip E. (1996) Flotation separation of plastics using selective depressants. Int. Journal of Mineral Processing 48: 127-134. [4] Fraunholcz N. (1997) Plastics Flotation. PhD thesis. Delft University of Technology. [5] Drelich J, Payne T, Kim J, Miller D. (1998) Selective froth flotation of PVC from PVCPET mixtures for the plastics recycling industry. Polymer Engineering Science 38: 1378-1386. [6] Agar GE, Barrett JJ. (1983) The use of flotation rate data to evaluate reagents. CIA4 852: 157-162. [7] Montgomery D. (1997) Design andAnalysis of Experiments. Wiley, 5th Ed. [8] Agante E. (2003) Caracterizacao dos residuos solidos provenientes dos eco-pontos das embalagens. Internal Report. CVRM, Instituto Superior Tecnico.
USE OF BIOMASS IN SMALL DIRECT FIRED SYSTEMS C. SYRED, W. FICK, N. SYRED, A.J. GRIFFITHS
Cardiff School of Engineering, Cardff Universiv, Queens Buildings, Cardiff CF 3AA, Wales, UK This paper discusses the context for the use of biomass for electricity generation in the UK and similar markets and evaluates the possibility of using cyclonic gasification coupled to small gas turbine systems. In the UK the Government has strongly pushed for a significant increase in the use of renewable energy for electricity generation with only very modest success, nearly 3% coming from this source at present, predominantly hydro and wind. Subsidy for the early systems came from an elevated price for generated electricity, but since then attempts at price convergence with that pertaining with conventional fossil fuel generation systems has not occurred and the number of small biomass systems being constructed and their net generating capacity has not increased in line with other technologies. This is in contrast to co-firing of biomass with coal in large utility boilers which has increased substantially over the last few years. Although small scale utilisation technologies exist, (< 50 MWe) and are well proven technologically in Scandinavia, when translated to markets such as the UK, give generating costs which are not competitive with other forms of renewable energy. Problems have arisen with many systems, being predominantly due to foulinghlagging, the different nature of the fuels, and elevated moisture content. In this context this paper describes an EU sponsored programme of work to develop a simple cyclone gasifier and combustor which can produce a medium calorific fuel gas for materials such as sawdust, retain up to about 80% of the total ashiresidues in the system, and fire simple, low cost gas turbines for power generation. The system is shown to have a very wide operating range and can handle sawdust with significant quantities of material up to 4 mm in size, whilst tolerating significant variation in moisture content and capturing very significant quantities of the asWparticulate matter as well as volatile species. The economics of the system are discussed and related to other technologies that are available and markets identified where potential exists.
1. Introduction The context of this paper is the need to develop efficient, low capital cost and small scale power generation systems that can be directly fired on a range of biomass materials, many of which will be grown as energy crops in Europe in this century. The driver for this is the very clear draft EU directive of May 2000. The key features of this proposal are as follows: Member States will be required to adopt national targets for renewables that are consistent with reaching the Commission’s overall target of 12% of energy (and 22.1% of electricity) from renewables by 2010; for the UK this translates to 10% of electricity by that date. There are some ambiguities on what is included (i.e. energy from landfill gas and municipal waste (MSW)). The UK government has set its own targets, from a baseline of 1990, of: Reducing Greenhouse Gas Emissions (GHGE) by 20% by 2010; Increasing renewable electricity to 10% of the total by 2010, in line with proposed EU directives, provided that generation is competitive. Think Tanks in the UK have also discussed 60% reductions in GHGE by the year 2050, in order to help slow down climate change. Reductions of 10% in GHGE have already been achieved on the 1990 figure by the so-called ‘dash for gas’ and the closureireduction to part load for coal fired power stations and now by co-firing of biomass with coal in large Utility boilers. A moratorium 127
128
C. Syred et al.
on the construction of new combined cycle gas fired power stations has recently been lifted and this is could lead to reduced generation of electricity from coal in the UK by 2010 [l]. The industrial use of coal in the UK was also discouraged by a climate change levy, introduced in April 2001, which is 0.25EurocentkWh for natural gas, 1.95EurocentkWh for coal, 1.6 EurocentIkWh for LPG and 0.71 EurocenVkWh of electricity used. However the ability of large coal fired utility boilers to readily co-fire biomass up to levels of 5% thermal or more, with the associated award of profitable ROCs (Renewable Obligation Certificates - a ‘green’ subsidy) has titled the economic balance back towards these units in the short term, whilst longer term the farming of energy crops is being actively encouraged. Independent analysis indicates that in the UK the largest contribution to reduced GHGE up to 2010 will result from reduced COz emissions from power stations due to improved new station efficiencies and the continuing ‘dash for gas’, coupled with a steady increase in the contribution from renewables [ 11, hopefully expected to increase from just under 3% in 2003 to 10% of total generation by 2010. These trends are likely to be reflected in much of Europe, and place question marks over the future role of coal. Long term UK view is that no new coal fired power stations can be built after 2020 without COz sequestration. Turning to the use of biomass and wastes there are many problems to be dealt with. Technically there are many solutions as shown in Austria, Sweden, Finland and Denmark. However when these technologies are translated to other markets, capital costs are very high, and subsidies of some form or other are needed. An example is provided by the straw fired power station in Ely in the UK [2], based on Danish technology. Capital costs for a 36MW (e) station are quoted as Euro 3.1Million per MW, with a contract price for electricity of nearly 10 EurocentdkWh until the year 2013. The UK government has developed it’s so called NFFO (Non fossil fuel Obligation), NI-NFFO, SRO and now ROC schemes to encourage the use of renewables for electricity generation over many years; the price guaranteed for electricity was very favourable in the early schemes. However since the price paid for electricity under these schemes has been made to converge to that from other sources, the number of new schemes that have actually come on line to generate power from biomass has not increased. Analysis of data from the UK Government (News Review-DTI, 2000), Table 1, gives the following data: Table 1. Relative contributions from renewables in UK.
I
Declared net renewable capacity (DNC) as at November 2000 from 343 projects: Contributions from various sectors Landfillgas Municipal and industrial waste Wind Biomass-chicken litterhtraw etc. Hydro Sewage gas Other
I 897.46 MW I 386.0MW(43.01%) I 166.1MW(18.5 1%) 171.34MW(19.09%) 54.8MW (6.1 1%) 40.36MW (4.5%) 25.1MW (2.3%) 58.2MW (6.48%)
Use of Biomass in Small Direct Fired Systems
129
Landfill gas, Municipal Waste, Wind, hydro schemes are all approaching the situation where there are difficulties in evolving significant numbers of new schemes which will actually be constructed. Turning to biomass power generation if electricity is to be generated in a manner at a close to market price, significant advances must be made in reducing capital costs whilst maintaining system efficiency. This is clear from the above table where Biomass has only made a small impact in the total renewables market, about 6%. The capital cost of a specialist boiler, grate, and steam turbine system is very high [2]. Long term capital cost reductions of at least a factor of 2 to say 1.5 Million Euros per MW are needed and for smaller systems, say less than 5 to lOMW direct fired gas turbines are attractive, and when a regenerator is used it appears that cycle efficiency for electricity generation approaching 30% may be possible, the same figure as reported in [2] for much more expensive systems. The European Union has recognised these factors: discussion on the perceived EU requirements are given in Pilavachi [3]. Direct solid fuelled fired gas turbines are of course not new, with many programmes, starting in the 1950si60s with very simple gas turbines. They were discontinued in the late 1960’s due to the low price of fossil fuels, but were restarted in the late 1970’s/1980’s with the various fuel price shocks that were encountered. Two programmes in this area are of note; the pressurised fluidised bed combustion combined cycle system for coal, developed at CRE in the UK, followed by large-scale tests at Grimethorpe. This technically was very successful and led to a number of combined cycle plants being constructed, primarily in Scandinavia. Efficiencies of power generation up to and just past the 40% level were achieved. However, the high capital costs, the ‘dash for gas’ and the slow, but steady improvement in conventional pulverised coal power stations lead to little replication. Few problems are reported from these systems, providing appropriate turbine entry conditions are adhered to (-900°C) as well as maintenance schedules. Typical gas turbine requirements for particulate removal have been quoted as removing all material greater than 5 microns and this was achieved by the use of two large Stairmand high efficiency cyclones connected in series [4]. Similarly in the 1980’s Allison, then part of General Motors, had a well-publicised programme on coal fired gas turbines for trucks, trains and similar vehicles. These gas turbines ran on micronised coal. Little technical information is available. This paper thus analyses the performance of an inverted cyclone gasifier for biomass, designed to produce relatively clean fuel gas to fire a suitably modified gas turbine combustor, with the aim of incorporating this into small gas turbine systems. Cyclonic gasification is well established and can be carried out at high velocity, thus giving very compact pressure vessels and low capital costs. It does require a fairly fine fuel such as sawdust or similar (95% <5 mm), but the costs of producing such fuels are low compared to the extra running and capital costs from pressurised fluidised bed
130
C. Syred et a1
systems. Fluid bed systems are anyway not particularly well suited to very high volatile yield fuels such as sawdust. The concept is thus to gasify the fuel at low temperature (-900°C) to minimise volatilisation of potassium and sodium salts, whilst the provision of collector pockets at the appropriate place on the system can take advantage of the cyclonic action to separate out up to 90% or more of the ash generated, without the extra pressure drop of a parasitic separate cyclone dust separator. The paper describes a family of tests on an atmospheric version of this device and its potential translation to a small gas turbine system. Also described is an associated cyclonic combustor, with integrated ash removal, capable of accepting a wide range of different low calorific value gases, as well as start up fuel. 2. Hot Gas Clean-Up
This is extremely important consideration in biomass fired power plants. Removal of particles, alkali and salts is essential to minimise damage to turbine components with in some cases gas cooling being required. 2.1. Corrosion, Erosion and Deposition in Gas Turbines
Corrosion, erosion and deposition from fuel gases derived from biomass can cause damage to gas turbine components. Particulates can damage turbine blades and gases can cause deposition. Some information however on hot gas clean up requirements is available for IGCC plants that use biomass [ 5 ] , Table 2 below. Table 2. Gas clean-up requirements. NOx
Less than 50 mg NOJMJ
sox
Less than 25 mg SiMJ
Dust
Less than 10 mg/Nm3
Alkali Metals
Less than 0.1 ppm (w)
Dust
Less than 2.5 ppm (w)
Most large scale plants use ceramic filters for gas cleanup. Ceramic filters have performed well especially at low temperatures and reference 5 describes the use of ceramic candle filters in IGCC cycles. However the use of ceramic filters in the present programme was to be avoided as they would increase pressure drops to the system, decrease cycle efficiency and cause additional maintenance costs. Sulphur deposition has appeared to be a problem for some natural gas driven gas turbines even when the inlet gas had low sulphur concentrations [6]. Sulphur deposits were found to form around the throat of the choked control valve and continued downstream. Investigation however showed this deposit could be prevented by moderately increasing the gas temperature upstream of the deposit. This could also remove an existing deposit. This sulphur deposition however should not create a problem
Use of Biomass in Small Direct Fired Systems
131
for this work as the turbine inlet temperatures are in the region of 800°C, and biomass gas generally has low sulphur content. An important area of consideration is thermal barrier coatings. These may allow higher temperatures and hence efficiencies and so play an important role in thermal insulation of the components [7]. Although the system described in this paper uses relatively low turbine inlet temperatures at the moment, future development will demand the use of more efficient turbines and higher inlet temperatures.
2.2. Corrosion and Erosion Damage Simms et al. [8] developed a methodology for the measurement of corrosion and erosion damage on gas turbine material. The amount of acceptable corrosion and erosion determines the life of the component. Assessment of the materials performance is made in five stages, and covers pre-exposure contact and post-exposure opticallimage analysis. This method also allows determination of metal loss, depths of internal corrosion pit depths, as well as taking into consideration the requirements of the material. Tests taken allowed statistical analysis of the material’s degradation to be made and enabled models of material performance to be developed. Neilson [9] gave specifications for acceptable contaminant concentrations in the fuel for General Electrics LM2500 gas turbine. Table 3 shows that the turbine is not very tolerant of alkali metals, as tests showed premature corrosion of turbine blades running in high alkali environments. Table 3. Acceptable contaminant levels in fuel. Solids
See a) below
See b) below
d
600
3000
1 0 < d < 13 pm
6
30
13
0.6
3
Lead
20
100
Vanadium
10
50
Na + K + Li
4
20
Calcium
40
200
12
60
Sulphur Alkali metal sulphates
a) Maximum concentrations in combusted mass flow to the turbine ppb
b) Calculated maximum allowable concentrations in a typical bio fuel ppb
2.3. Alkali and Heavy Metal Removal
Alkali and heavy metal removal from gas flow is of particular importance to reduce corrosion andor erosion of turbine components. Biomass gas contains alkalis formed primarily from Potassium and Sodium in the feedstock. Tolerable concentration for gas turbine applications is very low and is not well documented. The gasifier exit temperature
132
C. Syved et al.
has a large influence on the alkali concentration. At relatively low gasifier exit temperatures the alkalis appear to condense on particulate matter. Thus the alkalis can be controlled by controlling particulates [ 101. Fixed bed gasifiers keep exit temperatures between 500°C and 600°C and hence avoid problems of vaporised alkalis and condensed tars. Fluidised bed gasifiers however have problems of gas quality control due to the substantially higher exit temperature. Alkalis are in the vapour phase and could require fuel gas cooling to condense the alkali. Salo and Mojtahedi [l 11 measured vapour-phase alkalis after the gas cooler in an (air-blown) Integrated Gasification Combined Cycle system. A pressurised fluidised bed gasifier was used with varying biomass fuels. The alkali content of the product gas was reduced to 0.05-0.1 ppm (wt) when the gas is cooled and filtered at 400-5OO0C.,below the specification for their industrial gas turbine. The exit gasifier temperature should be high enough so tars are vaporised. Tars are then burnt in the secondary combustor, and boost the gas heating value. Biomass power projects using fluidised bed gasifiers often require expensive gas cleaning techniques. Lulea University studied biomass powder gasified in a combine cyclone separator and gasifier operated at low temperatures as an attempt to avoid excess alkali compounds in the producer gas. Conditions are such that the volatilisation of corrosive elements such as sodium and potassium is suppressed remaining solid in residual char. Fredriksson [ 121 indicated that alkali separation in cyclone gasification of wood powder was 40-60%. Gabra et al. [13] made a comparison of alkali retention and separation in fluidised bed and cyclone gasifiers. Bagasse biomass was used for the gasification experiments. For cyclone gasifiers stable wall temperatures were found to be in the range of 650°C to 950°C with a fuel equivalence ratio of 0.25 to 0.21. Gasification experiments were conducted 850°C. Alkali separation, or effectively retention by the separated partciculates, of approximately 70% of the input alkali was found with an equivalence ratio of 0.25,,Fig. 1. The fluidised bed showed on average 12% retention after 1.4 hours of operation which dropped to 4% for an operation time of 12 hours. This gives major operational advantages to the cyclone gasifier relative to fluidised bed gasifiers.
CG I
CG 2
FBG 1
FBG 2
Fig. 1. Comparison of alkali retention / separated with char samples from bagasse gasification in FBG/CG
133
Use of Biomass in Small Direct Fired Systems
The difference in the performance between the cyclone and fluidised bed gasifiers is explained as cyclone gasifier char is continuously separated and removed, and the ash included in this char carries significant amounts of the potassium and sodium. The residence time is much lower for cyclone gasifiers, up to 1 second than the minutes in a typical fluidised bed. In the fluidised bed gasifier no solid material is removed from the bed except the particles entrained in the product gas flow. Gabra et aZ.[13] also carried out a chemical equilibrium model analysis that showed that the performance of alkali separation and retention can be expected to increase for elevated pressure compared to atmospheric gasification tests. Fig. 2 shows the effect of pressure on various alkali compounds. 3tJ -4KCl -t-NaCl
I
KZUD
-+ - NaZC12
25
h
2O
v
P P
15
E 0 I -
2 7
10
5 I
I
+ " 4
0
1
4
7
10
Pressure (bar) Fig. 2. Bagasse gasification-alkali concentration (ppm) with pressure, Gabra et al. (200 1).
The equilibrium calculations show that under gasification with equilibrium conditions the dominant alkali gas species are monochlorides KC1 (g) and NaCl (8) and, to a lesser extent dichlorides, &C12 (g) and Na2C12 (g). Formation of K and Na and hydroxides (KOH (g) and NaOH (8)) was an order of magnitude less. The calculations showed that the gas alkali phase in the producer gas from bagasse gasification increased with temperature or with a decrease in the equivalence ratio, Fig. 3.
134
C. Syred et al.
Gabra et al. [13] also showed that between 500°C and 600°C the alkali is in the solid phase. For the gasifier operation temperature of 800-900°C the alkali was predominately in the liquid phase. As the temperature is raised the alkali phase increasingly becomes gaseous. For the range considered the alkali phase is not greatly affected by the equivalence ratio. However equilibrium calculations over a greater range showed that the liquid alkali phase increases with increasing equivalence ratio and the gaseous phase decreases. Gasification trials were carried out at Cardiff University and determined the separation of alkalis through an inverted cyclone gasifier fired on two types of pulverised biomass using Plasma Excited Alkali Resonance Line Spectroscopy (PEARLS) developed at Tampere University of Technology. These results are discussed later.
d I @ d 6 +@ d
e@ $0"
Q
% @
++ eQ s+ $Q
Temperature 'C Fig. 3. Variation of alkali compounds in the producer gas from gasification of bagasse with temperature and equivalence ratios, Gabra et al. (2001).
3. Cyclone Gasification and Combustion Studies have shown that research into the use of biomass fuels for electricity production is needed. These studies have indicated that direct fired wood powder gas turbines are economically attractive for co-generating power plants up to 20 MW. Such plants
Use of Biomass in Small Direct Fired Systems
135
incorporate a gas turbine and steam turbine in series to maximise the electricity output [ 121. However problems are encountered with erosion from particles and deposition of alkali compounds, which could cause damage to the turbine blades as discussed in section 2. These problems can be increased by high inlet temperatures, which are desired to increase cycle efficiency. Direct fuelling of a gas turbine with wood powder [ 141 however showed no erosion or corrosion but severe deposits at inlet temperatures greater than 800°C. A possible solution to this process is the two-stage combustion of wood powder where the first stage is a cyclone gasifier/separator. In the cyclone the volatile and some of the fixed carbon are oxidised at low temperatures. The corrosive ash remains solid in the char particles and are separated from the flue gas in the cyclone. The wood powder can also be injected into the cyclone with steam as transport media. The exhaust gas can feed directly into the modified secondary cyclone combustor. The cyclone combustor will produce stable strongly swirling flow that will enhance mixing and burnout rates. Schematics of the system are shown later in Figs. 4 and 5.
Fig. 4. 3D view of cyclonic gasifier.
136
C. Syred et al.
Separation and retention of ash is enhanced using a device termed a vortex collector pocket (VCP) and evolved from cyclone separator work [15-171. In these devices the main cyclonic flow drives a weak subsidiary flow in a small auxiliary vortex chamber, with no net gas exchange, which slightly overlaps the main cyclone chamber. Particulates are concentrated into a thin boundary layer region. On encountering the VCP, the particulate matter passes directly into the VCP, where it rapidly decelerates due to wall friction and a change of direction where after it drops out into bins. VCP’s are used in both the gasifier and combustor. A novel vertically orientated, inverted cyclone gasifier was thus designed to gasify pulverised biomass. The design was based on a combination of principles including those involved in the general design of dust separators, where high angular momentum fluxes are used to separate dust suspensions and those involved in cyclone combustor design via the use of internally generated high centrifugal force fields, secondary flows and well stirred reactor conditions to mix fuel and air which promote good gasification characteristics [ 181. A 2 dimensional view of the test rig is shown in Fig. 4. The main cyclone gasification chamber consists of 4 sections each with 2 rectangular access ports, covered with plates to allow access, for e.g. optical measurement techniques or thermocouples and sample probes. The bottom main section includes the tangential inlet and the round axial outlet to the attached exit section. The geometric swirl number for the present design with no inserts (restrictions in the tangential inlet) is 3. However, with the use of inserts the geometric swirl number can be varied from 3 to 7.5. The exit- and ash bin section are attached to the main chamber as highlighted in Fig. 5. Two vortex collector pockets (VCP), as developed in principle by Biffin [15-161, and one central collector pocket (CCP) are incorporated to provide the necessary ash and dust removal capability. One vortex collector pocket is attached to the top main section and a second vortex collector pocket is incorporated in the exit section, attached to the central collector pocket leading to the ashbin section. A section attached to the tangential inlet allows the premixing of natural gas before it enters the gasification chamber (Fig. 5). This section had a venturi tube before the natural gas inlet and a pressurised air supply. An additional air supply from a fan can be used via a secondary tube attached to the inlet section. The spark ignition system is mounted at the intersection of both inlet tubes to provide an ignition source for premixed combustion conditions independent of the air supply selected. A hopper was attached to the venturi section to enable the input of wood powder for the gasification tests or inert particles for the separation trials. Wood fuel was fed to the system using a hopper attached to a screw feeder driven by an electric motor. Wood was dropped onto a vibrating table that provided a constant and reliable fuel feed rate. Wall thermocouples were positioned in each gasification section and at the exhaust. Thermocouple position is denoted using the non-dimensional position z/D,, where z represents the axial distance from the gasifier outlet/ exit (z = 0 is at the gasifier exit rim) and D, is the gasifier exit diameter. “IN” refers to inlet side thermocouples: “OUT” for those at the outlet/exhaust side. WR type thermocouples were used as appropriate.
Use of Biomass in Small Direct Fired Systems
137
Detailed velocity, temperature and pressure measurements were taken to characterise the cyclone gasifier under isothermal and combusting conditions [ 19-211. Gasification trials with a number of different fuels were undertaken to fully characterise the system, these results are reported in the following section.
Fig. 5. Experimental set up of cyclonic gasifier for separation and gasification trials.
3.1. Separation Tests and Gasification Trials The cyclone gasifier was warmed up using premixed natural gas with a mixture ratio of 1. This enables sufficient temperatures to be reached to allow gasification of the wood. Operating experience showed an exhaust temperature of approximately 600°C is sufficient to start gasification. This occurred after 10 minutes. The natural gas is then turned off, the air flow rate is turned down and the feeding of the wood fuel begins. The separation behaviour of inert particles, chosen to match gasification conditions, was studied in isothermal flow with four different types of particles to investigate how the system would act under gasification conditions, Separation in both the VCP and the CCP was investigated and showed that overall the system removed via the two sets of pockets all material above 63 microns, with significant particle collection down to 5 microns, in reasonable correspondence to what was found under gasification conditions. Isothermal velocity measurements [ 19-211 found that the minimum losses of tangential velocity at the cyclone inlet and the best conservation of tangential velocity in the main cyclone body were achieved with a geometrical swirl number of 5.1 regardless
138
C. Syred et al.
of the inlet flow, resulting in tangential velocity of up to 20% higher than the inlet velocity, thus used for all this work. 3.2. Gasification Trials
Experimental studies were carried out on three different fuels: 1. A: Commercial Austrian sawdust: dmean- lmm, 70% from 1 to 0.5 mm 2. B: Commercial British sawdust: dmean- 1 mm, 3>d>0.125 mm 3. S: Commercial Swedish Wood powder: 2>d>0.063 111111,dmean=OSmm Mean size distributions of the fuels were determined using sieves of different sizes. Each fuel has somewhat different size distribution, resulting in slightly different operation of the gasifier. For instance this gives a slightly unstable start-up condition for Fuel A when the fuel is first introduced to the rig and the gas is turned off. This is most likely due to the finer particles are initially needed to quickly provide the combustion heat for stable start-up operation. With the absence of finer particles more time is needed to change the wall temperature patterns of the cyclone from preheat to gasification processes. Experimentally a wide range of wood powder feed rates and equivalence ratios were used, resulting in a range of exhaust gas temperatures. All tests resulted in stable steady conditions. Detailed investigations of four conditions for typical conditions are shown in the following section. Table 4 below shows the gasification conditions investigated. Table 4. Gasification conditions.
Trial No Fuel Type Total Air Flow Rate [l/min]
Wood Fuel Feed Rate [kg/h] Equivalence Ratio [-3
a
b
C
d
Fuel A
Fuel A
Fuel S
Fuel S
550
660
550
660
26
31
26
40
0.243
0.205
0.243
0.190
A large number of trials were carried out with fuel A and S at different equivalence ratios to characterise the cyclone gasifier, exhaust temperatures are shown in Fig. 6 as a lowest exhaust gas temperature achieved at the smallest equivalence ratio). Exhaust gas temperatures above 1OOO°C do not represent realistic gasification conditions, regardless of the mixture ratio. Thus the lowest equivalence ratio at any feed rate is selected for further investigation. Exhaust gas temperatures were generally slightly higher for fuel A but fluctuations were higher with fuel S. This was due to fuel S being composed of smaller particles that are more reactive and hence release heat for the gasification reaction more quickly. The lowest equivalence ratios were obtained for higher fuel feed rates (Fig. 6). Wall temperatures increased as the equivalence ratio was increased. This effect was more pronounced at lower fuel feed rates and not present at all for fuel S at 40 kgh. The
139
Use of Biomass in Small Direct Fired Systems
inlet and exhaust sidewall temperature distribution are different with cyclic variations of about 50°C. However, local fluctuations of wall temperatures were very low. 1300 1250
E
I2O0
@ 1150 J
E
1100 1050 1000
f
950
f
900
U 26 kg/h, Fuel S
850 800 500
550
650
600
750
700
800
850
900
mr Flow Rate [llrninl 1300
oe-
1250 1200
@ 1150
=
1100
n 1050 u)
2
1000
c
3 2
950 900
t 4 0 kg/h, Fuel S
+26 kglh. Fuel A
850
t37 kg/h. Fuel A 800 0.15
0.17
0.19
0.21
0.23
0.25
0.27
Equivalence Ratio
0.29
0.31
0.33
0.35
[-1
Fig. 6 . Exhaust gas temperatures as a function of airflow and equivalence ratio.
Detailed experimental characterisation of the flow pattern within the cyclone gasification chamber was undertaken by Fick et al. [19-211. Two dimensional LDA measuring techniques were used. The isothermal and combusting flow patterns were shown to be generally independent of Reynolds and swirl number, however, negative or positive peak values were affected. It was found that regardless of the inlet flow rate, minimum losses of tangential velocity at the cyclone inlet and the best conservation of tangential velocity in the main cyclone gasification chamber were obtained for a geometrical swirl number of 5.1. The LDA data for axial and tangential velocity distributions measured for isothermal and combusting (natural gas) conditions was used as guides for the velocity distribution to be expected for the relevant gasification cases, Figs. 7 a) and b). Fig. 7 c) shows the measured temperature distribution under combustion conditions (the system was probed with fine wire thermocouples). An air flow rate of 1900 I/min and a gas flow rate of 100 l/min (mixture ratio-2) were chosen to simulate the
140
C. SyTed et al.
conditions for the gasification trials at realistic wall temperatures of around 850OC.With a Reynolds n ~ m b of e ~30055 in the cornbusting case (Fig. 12 c)) and an inlet velocity of 6.17 ads (for a total inlet Row rate of2000 Vhh-i),the Reynolds number for the relevant ~ a s ~ ~ c aconditions ~ i o n at about 650 llxnin decreases to 12755 based QPI an inlet velocity of 2 m/s for Fig. 4 a) and b). Fig. 7 a) and b) highlight the axial and tangential velocity d i s ~ ~ ~fort these ~ Q ~~ ~s ~ d i ~ i Q ~ s .
Fig. 7. Derived mean a) axial, b) tangential velocities for gasification, c) tempera^^.
141
Use of Biomass in Small Direct Fired Systems
3.3. Energy Balance An energy balance was performed for the four test cases, shown in Table 5 below. Table 5 . Summary of energy balance.
Exhaust gas velocity hot [ d s ]
7.80
9.44
7.88
9.60
Cal. Value exhaust gas [MJ/m3]
0.78
2.99
4.26
5.91
36
59
70
76
Efficiency [%]
Highest calorific value was achieved with trial d), including (by volume). 52% of nitrogen, 13% of carbon dioxide and remaining products, methane, carbon monoxide and hydrogen (35%). Losses for trial d) are lowest at 24%, followed by c) at 30%, then b) at 41% and a) at 64%. The higher calorific values and lower equivalence ratios for the fuel S result in lower losses. At a lower equivalence ratio condition b) could certainly reach higher efficiencies. However conditions a) and c) are approaching the lower operating limit of the gasifier wall heat loses are excessive. Trial c) represents the highest efficiency possible for such a low load condition. Table 6 shows the exhaust gas composition for the four cases.
3.4. ElementaVProximate Analysis: Separation Performance The separation of ash from the flow is very important to minimise damage to the turbine and additional hot gas clean up for use of the gas in the gas turbine. The VCP and CCP were designed to remove these particles from the flow as described earlier. A whole range of different tests were undertaken to investigate the overall separation performance of the gasifier as well as the collection efficiency for individual elements. To determine separatiodretention performance of the VCP and CCP a reference source to the ash was produced by forming 100% burnout ash from each fuel. This was done by burning a set amount of fuel in a stainless steel pot. The size of the 100% burnout ash for fuel A was typically 3 times smaller than the fuel and 5 times smaller for fuel S. The ash collected in the VCP and CCP was generally larger than the 100% burnout ash. For case d the CCP residual is of very similar size to the fuel with some 10% of the particles bigger than
142
C. Syred et al. Table 6 . Exhaust gas composition.
2mm. As fuel S contains no particles larger than 2 mm there is clear indication that agglomeration of the particles is occurring. The axial and tangential velocity distribution (Figs. 7 a) and b)) can explain this occurrence. Fuel particles move upstream and gasify close to the cyclone wall from r/De=0.7 to r/De=l.O. The particles not being separated by the VCP situated at r/De=l.O and z/De=3.33 join the downstream flow which leads into a reverse floe region from r/De=O to r/De=0.5 and z/De=1.53 to z/De=1.00. With low tangential velocities also in this region, low shear stresses are present assisting agglomeration in this region. This effect is however beneficial to the separation performance since the larger particles are more likely to be separated by the CCP collector pocket situated further downstream. Although particles collected in the VCP are generally smaller than the CCP ones, they are typically 30% to 70% heavier. This is because the centrifugal force field in the VCP region of the flow had tangential velocities identical to the inlet air velocity; the heavier particles are more likely to be separated. The remaining lighter particles in the flow are then separated in the CCP although heavier particles not captured by the VCP are again more likely to be separated. The VCP and CCP collected different percentage of the retained ash for each separate trial, highlighting the importance of both VCP and CCP collector pockets to provide good separation performance for a wide range of fuels and operating conditions. The overall gasification performance of the rig was characterised, with carbon burnout of 98/99% burnout. The collection efficiency for Al, Ca, K, Mg, Mn, Na, P and Si with respect to the input of those elements evaluated from the elemental analysis of 100%
143
Use of Biomass in Small Direct Fired Systems
burnout ash for fuel A and fuel S is shown in Fig. 8. The results are presented independent of the fuel feed content, i.e. 100% burnout ash content. Fig. 8 shows that, regardless of the air flow rate and the fuel load, there are trends only dependent on the fuel. For fuel S, trials c & d, typically 50% of all elements are separated with the VCP and CCP ash. For fuel A, trial a & b, the trend is also very consistent but with significant differences in separation performance between different elements. Only 10% of A1 was removed compared to well above 90% of Mn, doubtless due to volatilisation of Al. Large differences in separation performances were seen between the two fuels for some of the elements and it is clear that the separation rate is much more dependent on fuel type than any other operating condition. The retention rates of Alkali’s, Na and K, however were independent of the fuel type and identical for the moIre relevani: cases, trials b & d, with around 50/60% retention of Na and K. 100 90
F-
80 70
0
$
60
s 2 so I
5 ._
i 5 2-
s
40 30
20 10
0
A1
Ca
K
Mg
Mn
Na
P
Si
Element
Fig. 8. Collection efficiency for Al, Ca, K, Mg, Mn, Na, P and Si in VCP & CCP (%wt. of fuel feed)
As carbon burnout and the ash separation performance are generally very high, low particulate emissions in the exhaust stream can be expected. This was only the case however for the more relevant operating conditions (at lower equivalence ratios) trials b & d (here exhaust particulate emissions were down to 0.044 g/s, 0.4% of the inlet fuel, with between 70 to 80% of ash and unburnt char being collected). High particulate emissions were obtained for trials a & c, presumably due to the lower flow rates, hence velocities and cyclonic separation. The tar content of the particulate emissions in the exhaust gas was very low, less than 5% for trial b & d, indicating very good carbon burnout throughout the system. Very good separation and gasification performance was obtained with the inverted cyclone gasifier under atmospheric conditions at higher fuel feed rates. The lower fuel load 26 kg/h was found to be far too low for this test rig. The inlet velocities of 2 d s or
144
C. Syred et al.
less for the proposed conditions were on the lower operating limit of the test rig. Higher fuel feed rates and subsequently air flow rates/ inlet velocities are certainly possible and higher turn down ratios of 5 or even higher can certainly be achieved. This highlights the robustness of the design with clear benefits. 4. Cyclone Combustor
A secondary combustor was developed for integration into the small-scale biomass cogeneration process describe above. This combustor should be capable of running on the gasifiers’ exhaust gas whilst still be able to utilise natural gas or oil for start up and shut down procedures or as a pilot fuel during certain operational conditions. To minimise pressure loss across the system and maximise efficiency the use of complex hot gas clean up system should be unnecessary, due to the separation capability of the gasifier. The combustor will be indirectly fired to the gas turbine. Present designs of combustor that can run on multiple fuels are either direct fired or involve complex gas clean up systems. All these designs are essentially derived from conventional gas turbine combustor systems fired on conventional liquid fuels or natural gas. They are all designed to be tired on cleaned bio-gas. This arises because the turbines have sophisticated blades incorporating numerous fine cooling passages susceptible to blockage. Conversely the present combustor address a different problem involved with small scale power systems. Here gas turbine systems are generally of simpler construction with un-cooled turbine blades and can sustain modest levels of fine particulates less than 5 microns in size. Turbine inlet temperatures are up to 900°C. The combustor for this process can be more robust and is indirectly fired. The combustor should be capable of removing any particles remaining above 5 pm. 4.1. CombustorDesign
The cyclone type combustor is designed with several tangential inlets as shown in Fig. 9, an air inlet, a high CV fuel inlet for oil or natural gas and a low CV gas inlet. The tangential inlets create a strongly swirling flow that gives good mixing and burn out rates. The combustor is to be operated at a maximum thermal input of 550kW. The combustor is mounted vertically and has a cone section at the base to collect larger particles in the flow. The combustor is designed with a long chamber to allow the flame to move up and down with varying thermal input and quality whilst giving sufficient residence time for fuel burnout and thus low emissions of CO and volatile hydro carbons. The central section of the combustor is refractory lined allowing substantial heat storage capacity helping to create stable flames. There is a tangential off take on the combustor that recovers energy from the flow and also forces the exhausting flow tangentially across the VCP aperture, hence increasing separation capability. The combustor is to be used to fire a small gas turbine operating at an inlet temperature of 800°C. The exhaust gas of the combustor has higher temperatures than this, and is diluted by a co-flowing air stream that is passed through a jacket surrounding the combustor.
145
IJse of Biomass in Small Direct Fired Systems
This co-flow air lowers the temperawe of the exhaust gas and acts as a difhser to the Blow, lowering the pressure drop across the combustor. The VCP is placed just before the tan~ent~ off-take, a~ which removes any fme particle above 5 microns that could damage the mbine. Tbe VCP also removes the need for a cyclone separator to remove the particles, which would increase the pressure drop across the system further.
Tangential exit
rc’
Oil lniet
\e.
Air inlet 2 Air Inlet1
LCV vrood gas inlet
Fig. 9. Schematic ofthc cyclone combustor with 4 inlets at the bottom, a VCP to collect particles at the top and a tangential offtake to reduce pressure drop.
4.2 Combustor Flow Characteristics A cyclone type combustor was built with several tangential inlets, for ak, oil and gasifier LCV gas. The tangential inlets produced a swirling flow with good mixing, creating a stable flame. Fig. 18 shows the swirling flame inside the combustion chamber, (gas oil as fuel) with the top plate removed.
Fig. 10. Photo of swirling flame inside combustion chamber.
C. Syred et aL
146
4 3
cm ~0~~~~~~~
The omb bust or was initially modelled using the CFD package Fluent 4 , under atmospheric conditions but with velocities representative of those occurring under pressurised gas turbine operating conditions.‘ An unstructured t ~ e ~ ” d i m e n s i ogrid n a ~was applied for this analysis. The Reynolds Stress turbulence model was used as the combustor has strong swirling flow. Experience has shown that this model predicts swirling flow with s ~ ~ i f ~ c a nmore t l y accuracy thm other models. The LC’V gas species and gas oil were modelled using the mixture fractiodpdf approach thus a s s ~ i n g chemical equilibrium. The non-adia~dticpdf model was used as radiative heat transfer to wall boundaries would have to be considered. The local ~ e ~ o - c h e m i cstate a l is also related to entbalpy as well as the mixture fraction. The D T W radiation model was used t ~ o u g ~ o the u t cal~u~ation. Particles (assumed to be spherical) were i ~ ~ o d ~ into c e dthe LCV gas stream using the discrete phase function. This enables the particles to interact with the flow md predict particle trajectories, hence separation. The c o ~ ~ u s t was or modelled using the gas composition from the cyclone gasifier, trial b, Tables 5 and 6 . Two ~ a n ~ ~ ninlets ~ i atol the combustor were used, the LCV gas inlet and an air inlet. The combustor was designed with multiple inlets to allow high to low CV gas to be tested with two inlets used for the modelling described below, a LCV gas inlet and an air inlets. The combustor was modelled at 3Q% excess air to allow good mixing, produce a clean exhaust gas with low emissions and to allow wall temperatures to stay within accepkable limits. CFD results show using a typical LCV wood gas the combustor produced a strongly swirling flame. The vortex structure of the flow field caused an extended residence time producing complete combustion of the &el. Experimental tests demonstrated the swirling flame in Fig. 10. Velocity contours across the combustor are shown in Fig. 11. The LCV gas and air enters the combustor ~ ~ e n tthrough i a ~their ~ ~ respective inlets. Velocity contours taken at a plane through the inlets, Fig. 12 a), show high tangentiai velocity at the inlets and close to the wall.
Fig. 1 1 . Velocity contours across combustor d s .
Use qfRiomass in Small Direct Fired Systems
147
Tangential velocities decrease towards the centre md thus the flow is o f forced vortex form. Areas of high tangential velocity near the wall correspond to regions of high t e ~ p e r a ~ ~Fig. e s ,12 b). The flow has good mixing and produces a stable swirling flame. The majority of the swirling flow stays near the chamber walls. As the flow moves downstream. the velocity of the flow increases in the centre, indicating the presence of complex r e c i r c ~ ~ apatterns ~ i o ~ and secondary flows. The LCV inlet is designed with a relatively small area to increase mixing with the air. The tangential exit produces a near ~ i velocity ~ profile o thus ~creating more stable inlet c o n d ~ t ~ ofor n ~the gas turbine.
Fig. 12. a) Velocities and b) temperatwe contours thro~ghplane o f combustor inlet.
~ e ~ p e r contours, a ~ e Fig. 13, show that combustion is initiated in the wall region near the inlets and then spreads across the whole top 50% of the unit, p r e s ~ a b l ydue to the secondary flow effects. The exit temperature profile is quite uniform and when mixed with dilution air will form a good flow to fire into the turbine. Average exit temperatures are around 1350 K. Minimising pressure drop across a system is important, especially in small-scale plants, a level not exceeding 0.2 bar is needed here. Fig. 14 shows the pressure drop across the combustor is typicaliy less than 0.025 bar. This could M e r be reduced by placing a 4" difkser on the tangential off-take as would d i ~ ~ the ~ nhaus g us^ gas with co-flowing air, to lower temperatures suitable for the turbine inlet. Particles were introduced into the flow through the LCV gas inlet. Fig. 15 shows the tracks of 5 microns particles through the combustor. The VCP removes the particles -firom the flow before they exit in the exhaust. Thus the p e ~ o ~ a n of c e the VCP is adequate to remove particles &om the flow without the addition use of cyclone separators d o ~ s t r of e ~the combustor. The VCP acts by collecting particles from the wall b o ~ layer d ~where they are concentrated by the centrifugal force field, they pass though the aperture between the VCP and the main cyclone chamber into the VCP where they are slowed by fiction and drop to the bottom for collection.
148
.........
Fig. 13. Temperature profiles across combustor inlet and outlet.
"500 diameter (ma) Fig. 14. Pressure drop WFOSS the combustor
Use of Biomass in Small Direct Fired Systems
149
Fig. 15. Particle tracks of 5 micron particles through combustor and into VCP
5. Conclusions
The inverted cyclone gasifier developed by Cardiff University is suitable for gasifying biomass under a range of conditions with varying equivalence ratio between 0.19 and 0.24. The design is robust with good mixing and 98-99% burnout rate. The VCP and CCP removed particles from the flow, with high separationhetention rates. Up to 50% of the Alkali’s Na and K were removed from the flow. A good quality LCV gas was produced with maximum calorific value of 5.91 MJ/m3. The LCV gas produced is suitable for firing directly into the secondary combustor. The combustor produced a strongly swirling flame with good mixing and burnout. An area of secondary flow and recirculation occurred where combustion was enhanced. The combustion patterns showed good mixing with near full burnout and hence low CO and hydrocarbon emissions. Velocity and temperature profiles at the combustor exhaust were nearly uniform, forming stable conditions for the turbine inlet. The VCP positioned just before the combustor exit effectively removed particles above 5pm from the flow. The system produced gas suitable for indirect firing of the gas turbine. The pressure drop across the system was relatively low at a predicted maximum level of 0.2 bar.
Acknowledgements The authors gratefully acknowledge the support of the European Union in this work, both through the Euroflam Scheme and most directly through contract number JORC398081-
150
C. Syred et al.
start date 01/06/98. The work would not have been possible without this support. The support of the UK EPSRC is also gratefully acknowledged. Professor Syred gratefully acknowledges the award of a Royal Academy of Engineering Global Research Award.
References [ 11 Lees I. (2001) Idris Jones Lecture. Cardiff Castle, Wales, UK. [2] Loram G. (2001) Turning straw into power at Ely. Energy World286: 14-15. [3] Pilavachi PA. (2000) Power generation with gas turbine systems and combined heat and power. Applied Thermal Energy 20 (15-16): 1421-1429. [4] Reed TB. (1985) Principles and technology of gasification. Advances in Solar Energy: An annual review of research and development 2: 125-174. [5] Engstrom F. (1998) Hot gas clean-up bioflow ceramic filter experience. Biomass and Bioenevgy 15 (3): 259-262. 161 Wilkes C. (1999) Sulphur deposition in a gas turbine natural gas fuel control system. Joint Power Generation Conference FACT 23 (1). [7] Leyens C, Wright IG, Pint BA. (2000) Hot corrosion of an EB-PVD thermal-barrier coating system at 950°C. Oxidation ofMetals 54 (516). 181 Simms NJ, Oakey JE, Nicholls JR. (2000) Development and application of a methodology for the measurement of corrosion and erosion damage in laboratory, burner rig and plant environments. Materials at High Temperatures 17 (2): 355362. 191 Neilson C. (1998) Gas turbine modifications for biomass fuel operation. Biomass and Bioenergy 15 (3): 269-273. [lo] Williams RH, Larson ED. (1996) Biomass gasifier gas turbine power generating technology. Biomass and Bioenergy 10 (2-3): 149-166. [ l I] Salo K, Mojtahedi W. (1998) Fate of alkali and trace metals in biomass gasification. Biomass and Bioenergy 15 (3): 263-267. 1121 Fredriksson C. (1999) Exploratory experimental and theoretical studies of cyclone gasification of wood powder. PhD Thesis, Lulea University of Technology, Lulea, Sweden. [13] Gabra M, Nordin A, Ohman M, Kjellstrom B. (2001) Alkali retentiodseparation during bagasse gasification: A comparison between a fluidised bed and a cyclone gasifier. Biomass and Bioenergy 21 : 46 1-476. [14] Hamrick JT. (1991) Development of wood as an alternative fuel for gas turbines. Battelle Memorial Institute. Report PNL-7673iUC-245. [15] Biffin M, Syred N. (1985) Vortex collector pockets to enhance dust separation in a gas cyclone. Filtration and Separation 22 (6): 365-372. [16] Biffin M. (1984) Improved cyclone dust separators for hot gas clean up. PhD Thesis, University of Wales, Cardiff, UK. [17] Syred N, Biffin M, Dolbear S, Wright M, Sage P. (1986) Evolution of new concept for compact cyclone dust separators. Proc. Con$ on Gas Cleaning at High Temperatures EFCE, Event No 340, Inst. of Chemical Engineering: 17-30.
Use of Biomass in Small Direct Fired Systems
151
[ 181 Morgan DJ. (1990) Characteristics of non slagging cyclone combustors for solid fuels. PhD Thesis, University of Wales, Cardiff, UK. [19] Fick W, Griffiths AJ, Syred N. (1999) Cyclone gasification of pulverised biomass for operation of gas turbines in cogeneration plants. Isothermal LDA Velocity Measurements, Report no. 2435, Cardiff School of Engineering, Cardiff University, UK. [20] Fick W, Griffiths AJ, Syred N. (1999 b), Cyclone gasification of pulverised biomass for operation of gas turbines in cogeneration plants. LDA, Temperature, Pressure and Isothermal PIV, Report no. 2528, Cardiff School of Engineering, Cardiff University, UK. [21] Fraser T. (2004) CFD modelling of an inverted cyclone gasifier. PhD Thesis, Cardiff University, UK.
This page intentionally left blank
EMISSIONS REDUCTION BY CO-FIRING BIOMASS OR WASTE WITH COAL IN A PRESSURIZED FLUIDISED BED COMBUSTION COMBINED CYCLE POWER PLANT Y. HUANG, J.T. MCMULLAN, D. MCILVEEN-WRIGHT, S. MCCAHEY NICERT, University of Ulster at Jordanstown, Newtownabbey BT37 OQB, Northern Ireland, UK One potentially attractive power generation option to reduce harmful emissions and to reduce the waste disposal problem is co-firing biomass or waste with coal in the pressurised fluidised bed combustion (PFBC) system. In this study, mixtures of coal and biomass or waste, such as straw, wood chips and sewage sludge are fired in the PFBC, and the resulting gases and sensible heat are applied in a combined cycle. In order to investigate the effects of cofiring biomass or waste with coal on the performance of the PFBC system from operational, technical, economical and environmental considerations, detailed simulations were carried out with biomass or waste additions of up to 40% of the total thermal input. The results showed that the main parameters affecting the overall efficiency were the co-firing ratio and the biomass and waste properties. It was also found that the steam cycle output was more sensitive to the fuel used than the gas turbine cycle. As expected, the increased fraction of biomass or waste significantly reduced net COz emissions, and had a beneficial influence on SO, emissions. Co-firing wood chips resulted in lower NO, emissions and co-firing straw or sewage sludge slightly increased NO, emissions.
1. Introduction
Combustion of fossil fuels will remain the primary energy source in the world for the foreseeable future. It is also recognised that the conversion of fossil fuels into energy for electricity generation unavoidably leads to many ecological problems. The major pollutants formed in the combustion process are considered to be COz, SO, and NO,. One of the more promising methods for reducing the environmentalimpact of power generation in the future is the co-firing of biomass or wastes with coal in a PFBC [l], in a combined cycle configuration. The benefits of using this technology are related to both the reduction of environmental emissions and possibly the reduction of a waste disposal problem [2]. Unlike coal, biomass and wastes such as sewage sludge, wood chips and agricultural wastes in general contain a very small amount of sulphur. Hence, adding biomass or waste to the fuel stream in a coal-fired power plant can result in a proportional reduction in SO, emissions. In addition, replacing fossil fuel by carbon dioxide neutral biomass (that is absorbed during plant growth) is also an approach to decrease C02 emissions. Although the environmental advantages of biomass wastes are well known, the use of biomass wastes as a fuel source is still primarily restricted to small scale heat and power generation [3]. The problems associatedwith biomass or waste application in large capacity power plant are due to low quality feedstock, i.e., high moisture content, low energy density, impurities and low calorific value, resulting in handling and operating difficulties. However, as a mean of reducing coal consumption and improving environmentalperformance,co-firing of biomass or waste with coal can reduce the impact of these problems. The main purpose of this paper is to investigate the impact of variations in the cofiring ratio of biomass or waste with coal on the technical and environmentalperformance of an advanced PFBC system and to determine whether the biomass or waste addition had any 153
154
Y. Huangetal.
noticeable influence on combined cycle operations. To achieve these objectives, the approach to conducting case studies is as follows. A reference coal and three biomass or waste fuels were selected. The fundamental process modules making up the PFBC combined cycles were modeled and the necessary information for the process was collected including the properties of coal and biomass or waste. Following the modeling of the PFBC combined cycles and definition of the design parameters, numerical simulation of the process and a range of sensitivity studies were carried out, as described later, using the ECLIPSE process simulator [4-51. 2. The Pfbc Power Plant
The PFBC system selected here is shown in Fig. 1. It is based on a commercially available P800 module, which is the most recent version developed by ABB Carbon [6-71. Accordingly, the boiler system is designed to match the required steam conditions of 241 bar and 565 "C at a full load flow of 720 tonnes per hour with steam reheating to 48 bar, 593 "C. The steam turbine condenser pressure is 0.041 bar. The overall electric power output of the plant is expected to be about 360 MWe. Raw coal blended with limestone is fed to the preparation plant and crushed to a correct size distribution. Limestone sorbent is used if necessary, to allow for the relatively high sulphur fuels. The mixture is converted to a paste by mixing with water, normally maintaining 25 percent water by weight [8]. This paste is pumped into the pressurized fluidised bed combustor. Air is compressed to approximately 16 bar by a two stage compressor which is powered by the gas turbine driven by the hot flue gas. The outlet from the low pressure stage passes through an intercooler, to limit the outlet temperature from the high pressure compressor to 300 "C. The air from the compressor passes through the outer annulus of a coaxial aidflue gas pipe before it enters the combustor. This arrangement cools the wall of the inner pipe, which carries the flue gas to the gas turbine. Inside the combustor vessel, the air flow provides the oxygen for the co-firing of coal and biomass and enhances thermal contact between the gas and the solids. The fuel paste is burnt directly under pressure in the fluidised bed. The operating temperature of the bed is around 855°C. This temperature is well above the minimum coal and biomass combustion temperature and provides sufficient margin to prevent melting of the ash. It also minimises the products of another potential pollutant, nitrogen oxide (NO,) generated from either the combustion air or the fuel-bound nitrogen in the fuel. The sulphur compounds which are produced from the relatively high sulphur content of the coal during the combustion process are captured by direct sulphation of the calcium carbonate in the sorbent material, which is mixed with the coal in the fluidized bed and thus eliminates the need for additional flue gas desulphurisation (FGD) scrubbing system. The steam which is generated and superheated in tubes immersed in the fluidised bed is sent to the turbine stop valve. It is then expanded through the high pressure turbine which also drives a generator. The steam turbines have facilities for steam extraclion and allow for
Emissions Reduction by Co-Firing Biomass or Waste with Coal
155
transfer of steam to the feedwater heaters. The steam from the high pressure turbine is reheated before passing through an intermediate pressure and a double flow low pressure turbine. The hot flue gases are expanded in a gas turbine, which powers air compressors and provides further electricity generation. Bed ash is removed from the bottom of the bed to a bed ash cooler, where some sensible heat is recovered by low pressure feedwater. Particulate removal is achieved using parallel sets of two-stage cyclones [9] before the gas turbine, and final cleaning using a cold-side electrostatic precipitator. From the gas turbine exit, the flue gas passes to a waste heat boiler, where sensible heat is absorbed by the feedwater via a set of economizers. The plant includes all of the necessary power, water and waste treatment systems. Natural draught cooling towers are used to provide cooling water for the turbine condenser.
Fig. 1. Schematic of the pressurized fluidized bed combustion combined cycle.
3. Fuels The coal chosen in this study is the American Federal coal which serves as a reference coal. Federal coal has a medium calorific value of 35.64 MJ kg-' (HHV, daf), a low moisture content of 6.3% (ar) and a low ash content of 6.6% (db), but a higher sulphur content of 2.6% (daf). The biomass wastes selected during the case studies are straw, wood chips and sewage sludge, of which the calorific values range from 18.73 MJ kg-' to 22.93 MJ kg-' (hhv, daf), the moisture contents vary from 5.4% to 33.3 % (ar), the ash contents change from 0.9% to 21.8% (db), the sulphur contents are between 0.1% and 0.9% (daf), chlorine content (ranging from 0% to 0.4%) and nitrogen content (ranging from 0.1% to 5.1%). What should also be
Y. Huang et al.
156
considered in fluidised bed applicationsis the effect of mixing the biomass derived ash with the coal ash. In particular the potentially high alkali metal content of biomass ash can reduce the ash softening temperature and lead to agglomeration in the fluidised bed. Table 1 contains basic information about the feedstocks used in the co-firing simulation. Table 1. Fuel analysis. Description
Base case
Case 1
Case 2
Case 3
Fuel name
Federal coal
Straw
Wood chips
Sewage
Water % (ar)
6.3
10.6
33.3
5.4
Ash % (db)
6.6
4.6
0.9
21.8
HHV MJkg (daf)
35.6
19.9
18.7
22.9
Carbon
84.0
48.8
51.0
53.9
Hydrogen
5.7
7.1
6.0
7.9
Nitrogen
1.5
1.3
0.1
5.1
Sulphur
2.6
0.2
0.1
0.9
Chlorine
0.1
0.3
0.0
0.4
Oxygen
6.1
42.4
42.9
31.9
4. Process Simulation
To ensure that the investigation and comparisonswere performed on a consistent and reliable basis, all the cases were simulated using the ECLIPSE personal computer based process simulation package. ECLIPSE was developed for the European Commission over the period since 1986 by the Energy research Centre of the University of Ulster [4-51. A technical assessment study is performed in a logical sequence, in which the first stages involve the preparation of a process flow diagram for the system to be analysed, the addition of the technical informationused for the models and the completion of a converged mass and energy balance. When the mass and energy balance has been completed the next stages involve the economic analysis. These analyses provide all the data required to complete the assessment study.
5. Results and Discussion Table 2 summarises the major technical and environmentalperformance results for the PFBC combined cycles with the range of co-firing ratios and feedstocks given in Table 1. The sensitivity ofthe efficiencyto changes in the biomass types and co-firing ratios, as presented in Fig. 2.
Emissions Reduction by Co-Firing Biomass or Waste with Coal
157
Table 2. Mass and energy balance results.
Biomass name
Straw
Co-firingratio (thermalbasis) Fuel flow, kg s-' (ar) Thermal input, MW (HHV) Gas turbine output, MWe Steam turbine output, MWe GT share of gross output, %
I 0% 1 10% I 20% I 30% I I 26.95 I 29.20 31.44 33.69 I 840.10 I 840.4 840.00 840.00 I 78.10 I 80.27 I 82.47 I 84.67 I I 297.20 I 292.64 I 287.72 I 282.54 I I 20.81 I 21.53 I 22.30 I 23.05 I
40% 35.94 840.00 86.87 276.74 23.89
13.61
13.58
13.53
13.46
13.44
361.69
359.33
356.66
353.75
350.17
Net efficiency, % (HHV)
43.1
42.8
42.5
42.1
41.7
C 0 2 g kWh-' (total)
730
738
746
754
762
COz g kWh-' (excluding biomass)
730
661
592
523
453
185
216
249
Total auxiliary usages, MWe Net plant output, MWe
SO, mg (Nm3)-'
(02
6%)
NO, mg (Nm3)-' (026%)
-' (026%)
HCI mg (Nm3)
Main steam mass flow, kg s-'
187.7
I
181.2
184.6
I I
177.6
I I
173.6
334.1
326.3
Heat duty (SH & RH), MW
=-I-= I 182.0
178.8
Ash cooler heat duty, MW
6.0
5.8
5.7
5.5
5.4
Heat duty (HRSG), MW
105.0
106.3
107.0
107.9
108.2
Exhaust flue gas flow, m3 s-'
401.9
405.2
410.4
414.8
422.1
Exhaust flue gas heat loss, MW
102.2
108.6
115.0
121.9
131.5
Exhaust temperature, OC
122.0
122.3
123.5
124.5
127.5
Hot gas flow (GT inlet), m3 sC1 Heat duty (Boiler & Econ.), MW
1 Biomass name
1) Co-firing ratio (thermal basis)
I
I
0%
I I
78.8
10%
79.5
80.2
168.0
I
Wood chips 20%
I
30%
I
I 40%
Fuel flow, kg s-' (ar)
26.95
31.06
35.13
39.22
43.31
Thermal input, MW (HHV)
840.1
840.00
840.0
840.00
840.00
1
158
Y. Huang et al. Table 2. (Continued)
Biomass name
Wood chips
Gas turbine output, MWe
78.10
82.20
86.91
91.33
95.73
Steam turbine output, MWe
297.20
287.93
277.75
268.05
258.17
GT share of gross output, %
I
20.81
I
22.21
I
23.83
I
25.41
I
27.05
13.61
13.68
13.72
13.81
13.85
361.69
356.45
350.93
345.57
340.05
Net efficiency, % (HHV)
43.1
42.4
41.8
41.1
40.5
C 0 2 g kWh-' (total)
730
75 1
773
795
816
C 0 2 g kWh-' (excluding biomass)
730
663
596
528
46 1
SO, mg (Nm3)-' (026%)
175
157
140
123
106
Total auxiliary usages, MWe Net plant output, MWe
1 NO, mg (Nm3)-' (026%)
I
-' (026%)
126
I
114
I
102
I
90
I
78
121
108
97
84
72
Main steam mass flow, kg s-l
187.7
181.1
174.4
167.9
161.7
Hot gas flow (GT inlet), m3 s-'
77.4
79.1
80.9
82.6
84.3
Heat duty (Boiler & Econ.), MW
334.1
320.2
305.4
291.1
276.6
Heat duty (SH & RH), MW
182.0
175.6
168.8
162.5
156.5
HCl mg (Nm3)
1 Ash cooler heat duty, MW 1 Heat duty (HRSG), MW (1 Exhaust flue gas flow, m3 s-'
I I 1
6.0 105.0 401.9
I I
I
5.6 107.8 411.6
I I
I
I
I 108.3 I 108.4 I 426.3 I 443.2 1
453.3
5.3
5.2
5.0 110.8
Exhaust flue gas heat loss, MW
102.2
116.4
131.5
146.9
162.6
Exhaust temperature, "C
122.0
124.7
129.9
135.5
141.6
Biomass name Co-firing ratio (thermal basis)
Sewage sludge 0%
10%
20%
30%
40%
Fuel flow, kg s-' (ar)
26.95
29.21
31.46
33.72
35.97
Thermal input, MW (HHV)
840.10
840.20
840.20
840.30
840.10
Gas turbine output, MWe
78.10
80.16
82.18
84.24
86.29
Steam turbine output, MWe
297.20
292.85
288.59
283.54
277.85
Emissions Reduction by Co-Firing Biomass or Waste with Coal
159
Table 2. (Continued)
Biomass name
Sewage sludge
GT share of gross output, %
20.81
21.49
22.16
22.90
23.70
Total auxiliary usages, MWe
13.61
13.60
13.59
13.54
13.51
361.69
359.41
357.18
354.24
350.63
Net efficiency, % (HHV)
43.1
42.8
42.5
42.2
41.7
C 0 2 g kWh-' (total)
730
734
738
743
750
C 0 2 g kWh-' (excluding biomass)
730
661
590
520
450
SO, mg (Nm3)-' (026%)
175
167
158
150
141
NO, mg (Nm3)
(026%)
126
138
149
160
172
HCI mg (Nm3)
(026%)
121
160
199
237
276
Main steam mass flow, kg s-'
187.7
184.7
181.6
178.0
174.9
Hot gas flow (GT inlet), m3 s-'
77.4
78.4
79.2
80.2
81.1
Net plant output, MWe
~
Heatduty(Boiler&Econ.),MW
Heat duty (SH & RH), MW
Ash cooler heat duty, MW
I 1
334.1 182.0
I
I I
326.2 178.7
I
I I
319.0 175.8
I
I 311.5 I I 172.3 I
I
305.4 168.9
I
6.0
6.4
6.7
7.0
7.2
Heat duty (HRSG), MW
105.0
106.3
107.5
108.2
108.7
Exhaust flue gas flow, m3 s-'
401.9
406.9
411.7
418.3
425.9
Exhaust flue gas heat loss, MW
102.2
108.6
115.0
121.9
129.1
Exhaust temperature, "C
122.0
122.2
122.8
124.5
126.9
In general, as shown in Fig. 2, the overall efficiency, as a function of co-firing ratio, decreases with an increase in the biomass or waste proportion. In comparison with the base case (100% coal firing), case 1 (co-firing of straw and coal) displays a 1.4% (HHV)reduction of net efficiency when the co-firing ratio varies from 0% to 40% on thermal basis. Also similar trends were seen in case 2 (co-firing of wood chips and coal) and case 3 (co-firing of sewage sludge and coal), where the overall efficiencies were lowered by 2.6% (HHV)and 1.3% (HHV)points respectively. It can be seen that there is a notably different effect on the overall efficiency depending on the type of biomass or waste. The average rate of efficiency reduction from wood chips was about twice of the rate obtained from straw and sewage sludge.
Y Huang et al.
160
In general, the main reason for the efficiency reduction by the biomasses and wastes is the low energy density of the feedstocks, their high moisture (straw and wood chips) and the ash content (sewage sludge). The magnitude of the efficiency reduction reflects the relative significance and quantities ofwater and ash present in the feedstock. Since the PFBC system uses a wet fuel feeding, an even greater quantity of water is fed to the combustor in the case of the low energy density feedstocks than with coal. As a result, it is not surprising that more sensible heat is lost in the flue gases. Also, the higher ash content of the sewage sludge results in more sensible heat leaving the combustion chamber. Therefore, there is an increased amount of unrecovered energy from the flue gases and the ash removal system. However, the moisture content has a much more significant effect on the overall efficiency than the ash content.
g 5 5 .
’%
62 2
0
43,5 43,O 42,5
-
42,O 41,5 41,O 40,5 40,O 0
-
1 1
10
20
30
40
Biomass and waste fraction % (thermal input)
- - 0 - .Straw
--O- Wood chips +sewage
sludge
Fig. 2. Overall efficiency.
The effect of co-firing ratios on the combined cycle operating characteristics has been found and the results are shown in Table 2 and Fig. 3. Firstly, because the biomass or waste has a lower energy density than coal, when the co-firing ratio is increased up to 40%, the mass flow of fuel feed stream (including limestone) is enlarged by 33% for straw, 60% for wood chips and 33.5% for sewage sludge, compared to coal only firing. This is a disadvantage which requires more fuel storage area and fuel processing and handling equipment. Secondly, increasing the co-firing ratio increases the flue gas massholume flow, resulting in a larger fraction of gas turbine output. At the same time, the waste heat flow is enlarged. Although the electrical output from the expander section of the gas turbine is increased, in this case, less heat is available from the PFBC for the steam cycle, which decreases the steam output quite markedly. Compared to coal only firing, case study 2, when the co-firing ratio is set to 40%, the gas turbine output is increased by 17.6 MWe and the steam turbine output is decreased by 39.0 MWe. Consequently, the steam mass flow and the heat duty of the evaporating unit and superheating and reheating units are reduced by 14%, 17% and 14% points, respectively. There is about 10.7MW net heat duty reduction in the evaporating unit,
Emissions Reduction by Co-Firing Biomass or Waste with Coal
161
which is compensated for by increasing the feedwater temperature. In other words, it is necessary to increase the heat duty of heat recovery steam generator (HRSG). On the other hand, since the amount of waste heat in the flue gas path is increased, more feedwater is needed to reduce the flue gas temperature. As a result, if there is not sufficient heat exchange in the flue gas path, an additional increase of the exhaust temperature is unavoidable. Lastly, as mentioned earlier, since the co-firing ratio affects the gas turbine output, the gas turbine selected should have enough margins in capacity to cope with the changes in power output. With regard to the total emissions of COz from the stack, an increase in the co-firing ratio reduces the overall efficiency and therefore gives a higher emission compared with 100% coal firing (730 g kWh-'). For example, emissions are approximately 762 g kWh-' for straw, 816 g kWh-' for wood chips and 750 g kWh-' for sewage sludge when the co-firing ratio is 40%. However, if the net emissions to the environment are calculated, ie. the straw and wood chips are considered to be neutral due to the absorption of CO2 from the atmosphere during their growth, then the co-firing processes, as shown in Fig. 4, demonstrate a much lower level of CO2 emissions. Referring to case study 1, in total gaseous CO2 emissions, of the 762 g kWh-' was emitted by coal and straw at 40% co-firing ratio, about 309 g kWh-' was contributed by the straw. The result is a net avoidance of 38% COzemissions compared to pure coal combustion.
'
28,O
-
27,O
-
1
25,O 24,O 23,O 22,o 260
3 @ 0 g
2
-
21,o
%
6
20,o 0
20
I0
30
40
Biomass and wastefraction % (thermal input)
- - 0 - - Straw +Wood chips ~
~~
Fig. 3. Gas turbine output share.
+sewage
sludge
Y. Huang et al.
162
800 700 600
500 nnn 7""
1
0
20
10
30
40
Biomass and wastefraction 96 (thermal input)
- - 0 - . Coal+straw - - ik - .Excluding straw C02
-D-Coal+wood chips
+Excluding
wood chips C02
I
Fig. 4. Net C02 emissions.
As shown in Fig. 5 , since biomass or waste mostly contains considerably less sulphur than coal, an increase the co-firing ratio always makes the SO, emissions decrease proportionally. However, the sulphur rich sewage sludge gave less reduction of SO, emissions compared with straw and wood chips. If the limestone addition is adjusted to give the same SO, emission level, the option of co-firing of straw or wood chips and coal gives an improvement in efficiency of about 0.2% points at 40% co-firing ratio.
3
.3
120
8 3 2
100
I
0
10
20
30
Biomass and waste fraction % (thermal input)
- - 0 - .Straw
+Wood chips
Fig. 5 . SOX emissions.
-d-Sewage sludge
40
Emissions Reduction by Co-Firing Biomass or Waste with Coal
6 2
0
I0
20
30
163
40
Biomass and waste fraction % (thermal input) - - - 0-- - Straw
+Wood chips
+Sewage
sludge
Fig. 6 . NOx emissions.
Since the NO, emissions are highly dependent on the fuel nitrogen content and boiler operating conditions, increasing biomass or waste fraction in the fuel blends shows different trends, as shown in Fig. 6, but they remain at low levels. This is the contribution partly from the low flame temperatures associated with the pressurised combustion and partly from the lower fuel nitrogen levels in terms of nitrogen per unit energy fed. The emissions of carbon monoxide in the flue gas of pressurized combustors were low, which was less than Smg/MJ [9-101 due to the operating pressures [ll]. It should be noted that there is a higher chlorine content present in straw (0.3%) and sewage sludge (0.4%), which potentially gives a higher HCl emissions, 249 mg (Nm3)-' for straw and 276 mg (Nm3)-' for sewage sludge at 40% co-firing ratio. 6. Conclusions
The ECLIPSE process simulator was used to successfully perform technical and environmental studies for the co-firing of coal and biomass with different ratios in the PFBC system, where straw, wood chips and sewage sludge were chosen to blend with Federal coal (a reference coal). The process simulation and analysis of the co-firing in the PFBC power plant led to the following conclusion: The technical and environmental impact of co-firing of coal with straw, wood chips or sewage sludge in the PFBC system has been assessed theoretically. Increasing the co-firing ratio will, in all cases reduce the overall efficiency. However, the co-firing of wood chips with coal gives the lowest efficiency, due mainly to its higher moisture content.
164
Y. Huang et al.
The co-firing of biomass or waste with coal has a reasonable effect on the PFBC combined cycle. As expected, the influence on the gas turbine cycle operation is smaller than on the steam turbine cycle operation, although the gas turbine output does change significantly. Also, this change links with the feedwater preheaters and the economisers. In order to maintain the PFBC plant output, it is suggested that when blending with high moisture content biomass wastes, more heat transfer surface in the HRGS is needed. Proportional effects are shown on the reduction of both net C 0 2emissions (straw and wood chips) and SO, emissions with increasing co-firing ratio. NO, emissions from the cofiring of wood chips and coal are much less than the coal firing process, due to the lower he1 nitrogen levels in terms of nitrogen per unit energy fed. However, the straw and sewage sludge additions give a slight increase of NO, emissions. References [I] Lindberg T, Anderson J. (1997), Co-firing of biomass and coal in a pressurised fluidised bed combined cycle. Results of pilot plant studies. The 14th ASME International FBC Conference, Vancouver, May. [2] Van De Kamp WL, Morgan DJ. (1996) The co-firing of pulverised bituminous coals with straw, waste paper and municipal sewage sludge. Combust. Sci. and Tech 121: 317-332. [3] Pedersen LS et al. (1996) Full-scale co-firing of straw and coal. Fuel 75 (13): 1584-1590. [4] Williams BC, McMullan JT. (1988) Development of computer models for the simulation of coal liquefaction processes. In Imariso and Bemtgen (eds), Progress in Synthetic Fuels, pp. 183- 189. Graham and Trotman, London. [5] ECLIPSE Process Simulator. Copyright 1992. Energy Research Centre, University of Ulster, Coleraine, BTS2 1SA. [6] Veenhuaizen D. (1999) Private Communication, S.A. Jansson & J. Anderson. Progress of ABB’s, PFBC Projects, the 15th International Conference on Fluidised Bed Combustion, Savannah, Georgia, USA. [7] Gustasson 0. (1993) PFBC for combined heat and power generation, Fernwurme International-FWI 22: 210-217. [8] Anderson J. and L. (1997) PFBC: Competitive power with minimal environmental impact. Practical experience. IChE Conference November 3-4, London, UK. [9] Advanced Electric Power Generation Fluidized Bed Combustion, http :Ilwww .lanl.govlproj ects/cctc/factsheets/tidd/tidddemo. [ lo] Podolski WF et al. (1983) Pressurised fluidised bed combustion technology, NOYES Data Corporation, New Jersey, USA.
PRODUCTION OF BIOCOAL FROM CASSAVA STALK T. PUTHIKITAKAWIWONG, R. BOONSU, 0. JOOMPHA
Faculty of Science, Mahasarakham University, Thailand Thailand is an agricultural country. Each year the country produced not only agricultural product but also more than 50 million tons of agricultural residues. Cassava stalk is the fourth largest agricultural residues which accounted more than 4 million tons per year. Cassava stalk consider as useless agricultural residues. Production of biocoal from cassava stalk is the process to convert the useless residues into useful h e l . Production of biocoal from cassava stalk consisted of two main processes: production of charcoal from cassava stalk and briquetting of cassava stalk charcoal. The charcoal yield was 22.18 % and the suitable binder to charcoal ratio was 1 to 7. Biocoal from cassava stalk has more ash than Eucalyptus charcoal but fixed carbon and calorific value are lower than Eucalyptus charcoal. However, the calorific value of this biocoal is still higher than calorific value of firewood. So, biocoal from cassava stalk can be used as household cooking fuel instead of firewood and charcoal.
1. Background and Objective
Thailand is an agricultural country. The majority of the populations are live in the rural area. Agriculture is the main academic activity for the rural people. Which consider as country’s lifestyle. The residues from the agricultural products in the rural area are abundant. Each year the country produced more than 54 million tons of agricultural residues. The ten largest agricultural residues in Thailand are, paddy straws, bagasse, rice husk, cassava stalk, corn cob, coconut husk, coconut shell, cotton stalk, saw dust and peanut shell. Cassava stalk is the fourth largest agricultural residue in the country. The total amount of cassava stalk is around 4 million tons per year. This amount of cassava stalk is considered as useless waste. The aim of this study is to convert this amount of useless waste into useful fuel. Biocoal is one of the conversion process that suitable for this purpose. Biocoal is the combine process of carbonization and briquetting.
2. Biocoal Production The carbonization process converted cassava stalk into cassava stalk charcoal. The briquetting process converted the cassava stalk charcoal into cassava stalk charcoal briquetted, which is called biocoal. Binder is use during the briquetting process. The detail processes of biocoal production are separated as follows: carbonization process in Fig. 1 and briquetting process in Fig. 2.
3. Result of Study Result of study showed that the biocoal production from cassava stalk has two main processes. The result of each processes are as follow.
3.1. Carbonization Process This process converted cassava stalk into cassava stalk charcoal. Cassava stalk and property are showed in Fig. 3 and Table 1. 165
166
T. Puthikitakawiwong, R. Boonsu & 0.Joompha
Cassava Stalk
+ Drying
1 1
Sizing
Carbonization
+ Cassava Stalk Charcoal Fig. 1. Carbonization process
Cassava Stalk Charcoal
1 1
Grinding
Binder Mixing
+ Briquetting
Biocoal Fig. 2. Briquetting process
Production of Biocoalporn Cassava Stalk
167
Fig. 3. Cassava stalk. Table 1. Propemes of cassava stalk.
-
The charcoal kiln, cassava stalk charcoal and charcoal yield are showed in Figs. 4 and 5 and Table 2.
Fig. 4. Charcoal kiln.
168
Fig. 5. Cassava stalk charcoal. Table 2. Carbonization of cassava stalk.
I- - - -
Average
32.
~
~ Ppscess ~
~
U
@
~
~
This process briquetted the charcoal fine (mixed with binder) into the final product whish call briquetted charcoal or biocoal. The briquettirag machine sand biocoal product we showed ir! Figs. 6 and 7.
Fig. 6. Briquetting machine
~
I69
Fig. 7. Biocoai firom cassava stalk.
3.3. Binder to ratio Result o f study showed that the suitable binder to charcoal ratio was 1 to 7. The details o€ the study are showed in Table 3 . Table 3. Binder to ~ h a ~ cratio. ~al
The suitable binder to charcoal ratio are 1 to 6 and B to 7. Because of the costly binder, the least use of binder is the best choice. So that the best binder to charcoal ratio is 1 to 7.
3.4 ~
0~
~
~~ w~~ ~ Charcoal o ~ c
o
~~
~
~
~
s
The biocoal from these processes has the calorific value lower than Eucalyptus charcoal. Eucalyptus wood charcoal has calorific value around 29 MJkg. The calorific value of biocoal fram cassava stalk is around 23.5 MJkg. The calorific value of firewood is around 14 MJkg. So that the calorific value of biocoal fi-om cassava stalk is in between firewood and charcoal. This biocod can be used instead of firewood or charcoal in the nurd c 0 ~ ~ ~ ~ ~
~
170
T. Puthikitakawiwong, R. Boonsu & 0.Joompha
4. Conclusions
Production of biocoal from cassava stalk has two main processes, carbonization and briquetting. The yield of carbonization of cassava stalk was 20.8% dry basis. Oil drum kiln is use as charcoal kiln; it is still cheap, efficient and easy to operate. The briquetting process by screw type briquetting machine. The optimum ratio between binder and charcoal was found to be 1 to 7 respectively. The calorific value of biocoal is lower than the calorific value of Eucalyptus charcoal but is still higher than the calorific value of firewood. So, this biocoal can be use as cooking fuel instead of charcoal or firewood. The disadvantage of this study is the briquetting and grinding machines, which is very expensive for the rural community. References [ 11 Bhattacharya SC, Shrestha Rh4. (1990) Biocoal technology and economics. Regional
Energy Resources Information Center (RERIC), Asian Institute of Technology, Bangkok, Thailand. [2] Stassen HEM. (1997) Biomass energy. Biomass Technology Group (BTG), Enschede, The Netherlands. [3] Grover PD. (1981) Development of briquetted fuel (PARU) from agricultural and other biomass residues. Producer Gas System and Agricultural Application. [4] Panya Thanya W. (1998) Charcoal making technology for rural forest product. Royal Forest Department, Thailand.
A COMPARISON OF POWER GENERATION FROM BIOMASS IN A SMALL CFBC PLANT WITH BIOMASS CO-FIRED WITH COAL IN A LARGE CFBC D.R. MCILVEEN-WRIGHT"), B.C. WILLIAMS(*),J.T. MCMULLAN'~) (')Northern Ireland Centrefor Energy Research and Technology (NICERT), University of Ulster, Coleraine, BT52 ISA, Northern Ireland, United Kingdom. "Exus Energy, Templemore Business Park, Derry, BT48 OLD, Northern Ireland Biomass is one of the renewable energy sources which is not intermittent, location-dependent or very difficult to store. If grown sustainably, biomass can be considered to be C 0 2 neutral. The use of biomass for power generation is also considered to be important in increasing the electricity output from renewable energy sources. However, power plants dedicated to the use of biomass fuel are not in widespread use and the acceptance of this he1 and development of the infrastructure for biomass production and transportation remain in their infancy. If small ratios of biomass can be co-fired with coal in large-scale conventional power plants, without significant technical, environmental or economic penalties, it could lead to a greater demand for biomass and stimulate the industry. In this study a 80 MWth CFBC, fuelled by biomass only, and a large-scale 1000 MWth CFBC, cofired with coal and 8% biomass, and the same large CFBC system, fired only with coal, are modelled using the ECLIPSE process simulation package and their technical, environmental and economic properties analysed and compared. The co-firing of biomass with coal was found to have little effect on the large-scale CFBC system, when a small ratio of biomass is used. The large scale system was found to have higher efficiency, lower C02 emissions and lower break-even electricity selling price than the small biomass-helied CFBC. Co-firing of biomass with coal could be a promising way of promoting the production, use and acceptance of biomass as a fuel in electricity generation.
1. Introduction
The co-combustion of coal and biomass has received widespread interest for some time as a means of conserving coal reserves and reducing net CO2 emissions [ 11. Several other environmental advantages have been reported e.g. co-firing high-sulphur bituminous coal with 20% straw gave a net reduction in NO and SO2 emissions [2]; lower NO, emissions may be found during co-combustion, since there is high volatile content in biomass and biomass nitrogen preferentially forms NH3 to HCN which is formed preferentially by nitrogen from coal [3]; and the primary reactions of thermal decomposition of biomass fuels are not significantly affected by the presence of coal, which itself does not seem to be influenced by the release of volatile matter from biomass [4]. The use of biomass, which is considered to produce no net COz emissions in its life cycle, can reduce the effective C 0 2 emissions of a coal-fired power generation system, when co-fired with the coal, but may also reduce system efficienay and increase electricity selling price. An analysis of several power generation technologies, using 100% coal, 100% biomass and coal-biomass mixtures has been made to identify the economic cost of reducing C 0 2 emissions through the replacement of coal with different amounts and types of biomass [ 5 ] . More recently there have been further financial incentives for co-firing, such as the requirements for increasing the percentage of electricity generated from renewable sources, carbon taxes, the increasing cost of gate fees at landfill sites and the ban on putrescible wastes going to landfill. In this study co-firing of a high ash coal with 8% (by thermal input) biomass in a large scale power plant is compared with a smaller plant fired only by the same quantity of biomass, both using fluidised bed technology. Experiments on the co-firing of these 171
172
D.R. Mcllveen- Wright, B.C. Williams & J.T. McMullan
mixtures were carried out and the results used in process simulation software to model power plants employing fluidised bed technology and perform technical, environmental and economic analyses of such systems.
2. Circulating Fluidised Bed Combustion Plants In this paper a computer simulation, using the ECLIPSE process simulation software package [6], was made of a large scale 1000 MWth CFBC power generation system. The process flow diagram of the modelled system was based on the power plant at Gardanne, France, which has been tested for co-firing with coal and biomass. Simulations were also made of a smaller 80 MWth CFBC power plant, which was fuelled completely by biomass in this instance.
3. Typical 1000 MWth Coal-Fired Circulating Fluidised Combustion (CFBC) Power Station In a typical CFBC plant (see Fig.1) coal would first be transferred from the normal coal storage facilities where it is then pulverised in mills, before being pneumatically transferred, together with limestone, using preheated primary air to a balanced draught, circulating fluidised bed boiler. Secondary air is injected through a set of nozzles higher up the chamber walls. The high fluidising velocity forms an expanded bed with material carried out of the combustor. Cyclones separate the majority of the solids from the flue gas. These solids are returned either directly to the combustor or through a set of external heat exchangers which receive preheated fluidising air. The low operating temperature (850 "C) and the staged combustion of the coal helps to reduce NO, formation. Sulphur retention is achieved by adding limestone, so no additional flue gas desulphurisation is required. In the combustor, the walls are lined with tubes which remove the radiant heat, and maintain the furnace temperature at 850 "C. Approximately 40% of the bed material is removed periodically from the base of the combustor and heat is extracted for lowpressure boiler feedwater heating. The rest of the solids are carried forward with the hot gases and removed by bag filters. High ash resistivity makes cold side electrostatic precipitators unsuitable and bag filters have the added advantage of promoting further sulphur retention. Before reaching the bag filter the gases are cooled by transferring heat first to steam in the superheater and reheater tubes, then to condensate in the economizer, and finally by passing through air preheaters at the back of the convective pass section. Superheating is achieved in both the external heat exchangers and the convective pass section. The reheater tubes are also located in these external heat exchangers and final economising also occurs in the convective pass section. The cooled gases are exhausted to the atmosphere via the induced draught fan and stack. The steam from the superheater goes to the turbine stop valve and is expanded in the high-pressure turbine. The steam turbines have facilities for steam extraction and allow for transfer of steam to the regenerative feedwater heaters. Drains from the three high-pressure feedwater heaters are fed to the deaerator. The steam from the highpressure turbine is then reheated before passing through intermediate pressure and double flow low-pressure turbines. At the crossover from the intermediate to the low-pressure turbines steam is extracted for the deaerator. Drains from the three low-pressure feedwater heaters are fed to the condenser.
A Comparison of Power Generation?om Biomass
173
Fig 1. Diagram of 1000 MWth CFBC.
The steam from the low-pressure turbine is condensed and the condensate is pumped by the extraction pump through three low-pressure surface-type heaters and a parallel ash cooler to the deaerator. Here the incoming water is heated by direct contact with the bleed steam. The boiler feed pump forces the condensate through three high-pressure feedwater heaters and the economizer before reaching the boiler and completing the steam cycle. Recent trials have shown that, when around 543% of the feedstock is not coal i.e. consists of biomass andor certain wastes, no modifications of the coal-fired plants are necessary. In the large scale system proposed here, there probably would need to be some additional reception, size reduction, handling and storage facilities for the biomass, which has been taken into account in the desigdmodification of the power plant. 4. 80 MWth CFBC System
A typical CFBC system of this size (see Fig. 2) would have standard biomass feed preparation, storage and handling facilities from which the fuel would be transferred, together with limestone absorbent when coal is involved, to an atmospheric circulating fluidised bed combustor. Air is first heated in an air preheater and external heat exchangers and then fed to the base of the combustor. The high fluidising velocity of this air causes an expanded bed to form and which carries material out of the combustor into the recirculation cyclones. These cyclones separate the majority of the solids, which are then returned to the base of the combustor via external heat exchangers. Most of the ash is removed from the base of the combustor and the highly efficient cyclone filters remove
174
D.R.Mcllveen- Wright,B.C. Williams & J. T. McMullan
the rest. The low operating temperature helps to reduce NOx formation, and sulphur retention (when coal is used) is achieved by the addition of limestone.
!
Fig. 2. Diagram of 80 MWth CFBC
The hot gases from the recirculation cyclones are cooled initially in a heat recovery steam generator and finally in combustion air preheaters. The cooled gases are exhausted to the atmosphere via the induced draught fan and stack. The steam from the heat recovery steam generator goes to the turbine control valve and is expanded in the steam turbine. The steam turbine has facilities for steam extraction to allow for transfer of steam to the feedwater heaters and to the deaerator tank. The low pressure steam from the steam turbine is condensed and the condensate is pumped by the low pressure (LP) pump through the LP heater to the deaerator tank and the high pressure (HP) pump through the HP heater, before reaching the heat recovery steam generator and completing the steam cycle. 5. Simulation Results
The two power plants were modelled using the ECLIPSE process simulation software and technical, environmental and economic analyses made. The larger scale 1000 MWth CFBC was assessed when fuelled by a high-ash, medium sulphur Puertollano coal, and also when co-fired with 8% (by thermal input) with biomass. The 80 MWth CFBC was analysed when fuelled by the same amount (and type) of biomass as in the 1000 MWth plant.
A Comparison of Power Generation from Biomass
175
5.1. Fuel Analysis The calorific values, proximate and ultimate analyses of these fuels are shown in Table 1. Table 1. Fuel properties.
I
I
I
Volatiles Fixed Carbon
Coal Feedstock
Biomass
Puertollano Coal
Pine Residues
Proximate Analysis % wiw 24.9 37.3
I
I
16.2
Moisture
5.5
Ash
32.3
0.5
Total
100
100
I
I
72.6 10.7
Ultimate Analysis % daf C H
77.33 5.31
4.94
N
1.93
0.90
I
S
I
I
0 Total
I
5 1.57
1.29
I
14.15
I
0.00 42.58
100.00 19.06
LHV (ar) MJikg HHV (daf) MJikg LHV (daf) MJ/kg
I
I
100.00
I
I
20.20
18.21
19.01
30.64
22.75
29.28
21.41
6. 1000 MWth System, Coal Only and Co-fired with 8% Biomass
6.1. Technical Results The technical and emissions results for the simulations of the 1000 MWth power plant are shown in Table 2. Table 2. Technical results for the 1000 MWth CFBC.
I I I
coaltype
1
Reheat?
excess air
Steamcycle Fuel Mix
FD fans (3) ID fan
I I I I
Peurtollano 20% 160bari538C Yes 100% coal
I I I I
Peurtollano 20% 160bad538C Yes 92% coal/ 8% wood
11086.2
10854.6
3882.5
3829.4
I I I I
D.R. McIlveen- Wright, B. C. Williams & J. T. McMullan
176
Table 2. (Continued) Slag Outlet
148.6
Bag Filter
7.8
7.4
Coal Crusher
22 1.2
135.5
Conveyors (4)
2042.9
2132.8
HP & LP Pump
I
9774.3
I
Elect. Utilities
4899.21
9765.6
I
4894.91
Total Usages
32062.8
31760.3
HP Turbine
108079.5
107985.1
LP Turbines (4)
179475.3
I793 18.5
IP Turbines (3)
149386.2
149255.7
I I
I I I
Net Electricity
I
Thermal Inuut HHV
I
140.1
Gross Electricity
Thermal Input LHV Efficiency. LHV
I I I
43694 1
I
436559.3
404878.2
I
404799.0
1000.00 1046.7 40.49
I I I
1000.00 1048.0 40.48
Efficiency, HHV
38.68
38.63
C02 g/kWb
873.7
866.3
C02 mgfNm3 at 6% 0 2
239935
241577
SO2 mg/Nm3 at 6% 0 2
243.4
250.8
NOx mg/Nm3 at 6% 0 2
340.38
352.95
CO mgfNm3 at 6% 0 2
59.82
62.01
4.04
4.04
0 2 (dry)vol %
I
I I I I I
6.2. Economic Results The economic results for the ECLIPSE simulations of the large CFBC system are shown in Table 3. Table 3. Economic results for 1000 MWth CFBC.
A Comparison of Power Generationkom Biomass
177
Table 3. (Continued) 20 1.29
CFBC
37.82
37.45
253.15
250.32
CFBC HRSG Sub Total
113.37
Steam turbine
I
I
199.35
Steam System & Cond
I
37.03
Cooling Water
I
15.03
Water Treatment Chimney
15.02
10.62
10.62
4.82
4.77
67.50
67.40 494.05 12
1208.2
1220.7
48.79
48.51
SI ($/kWe) BESP ($IMWh)
I I
36.99
489.1904
Sub Total Total
113.29
I 1
7. 80 MWth System, 100% Wood
7.1. Technical Results The technical and environmental results for the CFBC using 100% biomass are shown in Table 4. Table 4. Technical results for the 80 MWth CFBC.
I I
Fuel Mix
I
FD-Fan
I
ID-Fan
I 1 I
8% Biomass 739.5 118.7
Gas Cleaning
I
0.0
WoodConvey
I
205.8
AbsConvey
I
0.0
Coal Convey
0.0
Ash Convey
5.5
Plastic Convey
474.8
LP Pump
2.9
Size Reduction Total
I I
LPTurbine
I
Electric Process
IPTurbine
Electric Utility
I I I
0.0
HP Pump
I
I I
800
I I I I
2347.2 13457.9 14337.3 25448.0 450.6
I I I I
178
D.R. Mcllveen- Wright, B.C. Williams & J.T. McMullan
Steam Cycle bar/"C
921495
Thermal Input LHV
80
Thermal Input HHV Efficiency LHV %
31.25
Efficiency HHV %
29.40
C02 glkWh
1019
SO2 mgMm3 at 6% 0 2
0.0
Nox mgmm3 at 6% 0 2
348.3
CO mgMm3 at 6% 0 2
61.2
0 2 (dry) Vol %
5.5
7.2. Economic Results The economic results for the 80 MWth CFBC using 100% biomass fuel are shown in Table 5. Table 5. Economic results for the 80 MWth CFBC.
Total ($M) Specific Investment ($/kWe) BESP ($/MWh)
49.233 1970 59.83
A Comparison of Power Generationfrom Biomass
179
8. Comparisons 8.1. Efficiency The efficiency of the larger CFBC was found to change little from the case where it is fuelled solely by coal (40.49%, LHV) to the case where it is co-fired with 8% biomass (40.47%, LHV), as shown in Fig.3. The 80 MWth CFBC system was found to be much lower (3 1.29%) in efficiency, probably because the steam cycle conditions result in an intrinsically less efficient system (538°C with reheat and 160 bar for the larger system compared with 495OC and 92 bar for the smaller system). Electrical Efficiency 45
Fig. 3. A comparison of electrical efficiencies.
8.2. Specific Investment
The capital cost per unit of electricity generated, or specific investment, SI, of the larger CFBC increases slightly from 1208 $/kWe to 1220 $/kWe if the cost of additional equipment for biomass reception, storage, size reduction and handling is taken into account. The SI for the 80 MWth CFBC fuelled by 100% biomass is much higher, around 1970 $Awe, as shown in Fig. 4. Specific Investment 2000 180
160 1400
1000 MWth
00 MWth
Fig. 4. A comparison of specific investments.
180
D.R. Mcllveen- Wright, B. C. Williams & J. T. McMullan
8.3. Emissions The C 0 2 emissions of the larger CFBC (1000 MWth) were found to be around 874 glkWh for 100% coal, dropping to about 866 gkWh when it is co-fired with 8% biomass. The 80 MWth CFBC was found to emit around 1029 glkWh C02 when fuelled with 100% biomass, as shown in Fig. 5. The smaller CFBC emits more CO2 than the larger one due to its lower efficiency. If the biomass is grown sustainably, it can be considered to be carbon-neutral. The net C 0 2 emissions from the 80 MWth CFBC could be considered to be zero, when using 100% biomass fuel, and the 1000 MWth CFBC would have net C 0 2 emissions of around 803 g/kWh, when co-fired with 8% biomass, which is an 8% emission reduction over the 100% coal case.
Fig. 5 . C 0 2 emissions for both CFBC systems.
9. Electricity Generation Costs The break-even electricity selling price (BESP) for the larger CFBC system was found to be 48.79 $/MWh using 100% coal and 48.51 $/MWh when co-fired with 8% biomass. The coal and biomass costs were taken to be $52.251 daf tonne and $26.061 daf tonne respectively. The sensitivity of the BESP to variations in the cost of coal or biomass is shown in Fig. 6. The coal and biomass costs are varied by k 100% of their nominal values.
Fig. 6 . Variation of BESP with fuel price.
A Comparisonof Power Generationfrom Biomass
181
10. Conclusions 10.1. Efficiency
The efficiency of the larger 1000 MWth CFBC was only negligibly affected by changing the fuel from 100% coal to co-firing with 8% biomass. The smaller 80 MWth CFBC was much less efficient than the 1000 MWth CFBC (3 1.2% compared with 40.5%, LHV). Co-firing biomass in the larger CFBC is the more efficient method for using biomass to generate electricity. 10.2. C 0 2 Emissions
The less efficient 80 MWth CFBC emits more C 0 2 than the larger 1000 MWth CFBC per unit of electricity generated, but no SOX, when fuelled by 100% biomass. However, the net C 0 2 emissions for the 80 MWth CFBC system may be considered to be zero, if the biomass is sustainably managed. Co-firing with biomass lowers C 0 2 emissions and net C02 emissions of the larger CFBC. 10.3. Specific Investment
The 80 MWth CFBC has an SI (about 1970 $/kWe) more than 50% higher than that of the 1000 MWth CFBC system (around 1200 $/kWe), since the larger system is more efficient.
10.4. BESP The 1000 MWth CFBC system co-fired with 8% biomass has a lower BESP value than the 80 MWth CFBC using the same amount and type of.biomass. In addition, the BESP for the 1000 MWth system is only negligibly affected by variations in the cost of biomass, whereas the 80 MWth BESP is significantly affected, as shown in Fig. 4. Financial incentives for the use of biomass are currently available in some countries for co-firing, such as the requirements for increasing the percentage of electricity generated from renewable sources, carbon taxes, the increasing cost of gate fees at landfill sites and the ban on putrescible wastes going to landfill. These incentives have not been taken into account in this paper, as their long-term availability for co-firing applications is currently unconfirmed.
10.5. Summary In summary, the ECLIPSE simulations of the 1000 MWth CFBC co-fired with 8% biomass showed this system had a higher efficiency, lower C02 emissions, lower SI and BESP than the 80 MWth CFBC fuelled solely by biomass. The larger system was also much less sensitive to variations in the cost of biomass, which would have a significant effect on the economic viability of the small biomass-fuelled CFBC.
182
D.R. Mcllveen- Wright, B.C. Williams & J.T. McMullan
Co-firing small ratios of biomass with coal offers a promising way of promoting the production, trade and infrastructure of a biomass wastes or energy crops industry and the wider acceptance of biomass as a long term fuel for electricity generation. References
[I] Hein KRG, Bemtgen JM. (1998) EU clean coal technology - co-combustion of coal and biomass. Fuel Processing Technology 54: 159-169. [2] Pedersen LS, Nielsen HP, Kiil S, Hansen LA, Dam-Johansen K, Kildsig F, Chrstiensen J, Jespersen, P. (1996) Full scale co-firing of straw and coal. Fuel 75: 1584-1590. [3] Spliethoff H, Scheuer W, Hein KRG. (2000) Effect of co-combustion of sewage sludge and biomass on emissions and heavy metal behaviour. Proc. SaJ Environ. Prot. 78: 33-39. [4] Biagnini E, Lippi F, Petarca L, Tiognotti L. (2002) Devolatilization rate of biomasses and coal-biomass blends: an experimental investigation. Fuel 81: 1041-1050. [5] McIlveen-Wright DR, McMullan JT, Williams BC. (2003) The economics of reducing carbon dioxide emissions by the use of biomass co-combustion, Proc. Seventh International Conference on Energy for a Clean Environment ‘CleanAir 2003 ’, Calouste Gulbenkian Foundation, Lisbon, Paper 16.1. [6] Williams BC, McMullan JT. (1996) Techno-economic analysis of fuel conversion and power generation systems - the development of a portable chemical process simulator with capital cost and economic performance analysis capabilities. Int. J. Energy Research 20: 125-142.
CHARACTERIZATION OF SWEET, FIBRE AND BIOMASS SORGHUM POTENTIAL IN PORTUGAL AS AN INDUSTRIAL AND ENERGY FEEDSTOCK A.L. FERNANDO, M.P. DUARTE, J. MORAIS, J.F.S. OLIVEIRA
Grupo de Disciplinas de Ecologia da Hidrosfera / Unidade de Biotecnologia Ambiental, Faculdade de Cidncias e Tecnologia da Universidade Nova de Lisboa, Quinta da Torre, 2829-51 6 Caparica, Portugal: e-mail:
[email protected] orjjio@ct. unl.pt The purpose of this work was to investigate the influence of crop management on the sorghum biomass quality and productivity, in Portugal, in order to access its potential as an industrial and energy feedstock. To do so, the effects of different fertilization levels, the sowing date and the harvest date in the biomass quality and productivity, were studied. Productivity was affected by the level of nitrogen and by the sowing date. Highest productivities were obtained in the most N-fertilised fields and in fields sowed between 251hMarch and 1’‘ May. Among subspecies, Sweet and Biomass Sorghum presented, globally, higher productivities than Fibre Sorghum. In all the experimental years, highest productivities, in all the fields, were achieved at 193 5 9 days after sowing, regardless of all the different sowing dates. Biomass quality was affected by the crop productivities. Fields with lowest productivities resulted in a biomass with a higher nitrogen and ash contents, namely the fields sowed at 2nd June and the fields without N-fertiliser. In the experimental conditions at Caparica, sweet sorghum crop showed a better potential as an industrial and energy feedstock, and to get better results, economically and environmentally, fields should he sowed between 25’h March and 1’‘ May, with a high nitrogen input (120 kg N/ha).
1. Introduction Sweet, Fibre and Biomass Sorghum (Sorghum bicolour L. Moench) are annual C4 plant species with high water efficiency, high biomass yield and high dry matter accumulation rates on a daily basis [1,2]. Although sorghum is of tropical origin, the plant is well adapted in subtropical and temperate regions [2,3]. The greatest interest in Sweet, Fibre and Biomass sorghum is biomass production. Fibre sorghum’s stalks are fibrous, not juicy [4]. This involves using Fibre sorghum as an energy plant or as pulp for paper industry. Due to its high sugar content Sweet sorghum is gaining attention as a potential alternative crop for the production of bioethanol or sugar-sirup for industrial productions [4]. Stalks are crushed for juice extraction to separate the sugar from the bagasse. After extraction the sugar can be concentrated for storage or it can be converted into fuel ethanol [ 11. Bagasse can be dried for long term storage or can be processed immediately into bio-crude oil, gasoline, electricity, pulp for paper, compost or charcoal pellets [l]. The study of sorghum as an energy crop started in Portugal in 1991, with UBiA, in the scope of the project “Activity in the Sector of Biomass Production Sweet Sorghum Network” (JOULE Programme, JOUB 0036) supported by the European Union. Two others projects followed this one: “Sweet Sorghum, a sustainable crop for energy production in Europe: Agricultural and Industrial improvement, optimization and implementation” (AIR 1 CT92-004 1) and “Environmental studies on sweet and fibre sorghum sustainable crops for biomass and energy” (FAIR-CT96-1913), both supported also by the European Union, with the participation of UbiA. These three projects 183
184
A.L. Fernando et al.
permitted the study of sweet, 'fibre and biomass sorghum in Portugal, namely the morphologic and morphometric characteristics of the plants, their productivity and their biomass quality. 2. Materials and Methods
The experimental fields are situated in the Peninsula of S e ~ b a lin , the south border of the river Tejo, near the estuary and the Atlantic coast (latitude 38'40' N, longitude 9" W, altitude of 50 m). The experimental plots were established from 1991 to 1994 and from 1997 to 2000, in a clayey and alkaline soil, using a row spacing of 0.70 m with a distance within row of 0.10-0.12 m. Three different subspecies of sorghum were studied, sweet sorghum (Keller) (1991-1994 and 1998-2000), fibre sorghum (H128) (1997-2000) and biomass sorghum (MN 1500) (1997). During the experimental years, the fields were sowed between the end of March and the beginning of June: 1991 - 2Sth April; 1992 - 25'h March; 1993 - gth May; 1994 - 9" April; 1997 - 2"d June; 1998 - lstMay; 1999 - loth April. Top fertilization was applied after the emergence of the plants. For the sweet sorghum fields, from 1991 to 1994, two levels of N-fertiliser were applied: 75 and 100 kg N.ha-'. For the biomass and sweet sorghum, from 1997 to 2000, three levels of Nfertiliser were studied: O, 60 and 120 kg N.ha-'. For fibre sorghum, from 1997 to 2000, only one level of N-fertiliser was studied: 60 kg N.ha-'. P-fertilizer (120 kg P.ha-') and Kfertilizer (120 kg K.ha-') were applied in all the fields. All the fields were fully irrigated in order to compensate the water deficit of the soil, and to prevent water stress. At the end of the growing season the vegetable material was harvested and the total aerial dry weight, ash, nitrogen and phosphorus contents, the fibre content, the sucrose content and the gross heat of combustion, were determined in order to evaluate the productivity and the quality of the biomass. The chemical analyses were performed according to the following procedures: a) ash content: calcination at 550 50°C for two hours in a muffler furnace; b) nitrogen content: by the Kjeldahl method; c) phosphorus content: by the ascorbic acid method, after digestion of the sample; d) fibre: by the Belluci method; e) sucrose: sugar extraction by hot distilled water, sucrose determination by the UV method for determination of sucrose and D-glucose in food stuffs and other materials (Boehringer-Mannheim); f) gross heat of combustion: using an adiabatic calorimeter.
*
3. Results and Discussion 3.1. Biomass Productivity Fig. 1 shows the differences, in terms of the dry matter productivities, for sorghum, among fields sowed at different dates. According to these results there were differences among fields sowed at different dates. Better results were obtained in the fields sowed between 25" March and lst May.
185
Sweet, Fibre and Biomass Sorghum Potential in Portugal
Sowing as late as gth May resulted in lower productivities, but still, higher than 20 t/ha, not compromising the economical viability of the crop production. But, sowing as late as
I
45
I
I
40
f
35
b 30 U m
25
z
20
c
p
15
!10 U
E
5 0
25th
9th April
March
~
10th April
28th
1stMay 9th May
2nd June
April
Fig. 1. Sorghum productivities obtained in fields sowed at different dates.
2"d June resulted into extremely low productivities, and this can compromise the net economical and environmental gain of this crop. But, these significant lower productivities obtained in fields sowed at 2nd June, can also be a consequence of the climatic conditions observed in 1997, namely, a colder summer than usual with heavy showers. Fig. 2 shows the differences, in terms of the dry matter productivities, among fields with different nitrogen fertiliser levels. According to these results there were significant differences among productivities obtained in fields with different levels of nitrogen: productivity increased with the increase in nitrogen input. Fig. 3 shows a comparison, in percentage terms, of the dry matter productivities, between biomass sorghum (MN 1500) and fibre sorghum (H128) and between fibre sorghum (H128) and sweet sorghum (Keller).
I
35
.
30
25
P m 20 c 0
15
TE i
10
0
Low
High
Fig. 2. Sorghum productivities obtained in fields with different nitrogen application levels (0 = 0 kg Nha; Low = 60 or 75 kg N/ha; High = 100 or 120 kg Nha).
A.L. Fernando et al.
186
100 90 80 70 60 S 50 40
--
30 20
10 0
Biomass Sorghum
Fibre Sorghum
Fibre Sorghum
Sweet Sorghum
Fibre vs Sweet
Biomass vs Fibre
Fig. 3. Comparison of dry matter productivities between Biomass Sorghum (MN 1500) and Fibre Sorghum (H128) and between Fibre Sorghum (H128) and Sweet Sorghum (Keller) (in percentage terms).
According to these results, Sweet and Biomass Sorghum presented, globally, higher productivities than Fibre Sorghum. In all the experimental years, highest productivities, in all the fields, were achieved at 193 f 9 days after sowing, regardless of all the different sowing dates. Before 193 days, biomass is still in the process of growing and productivities were lower. After 193 days, the productivities lowered significantly mainly due to the loss of the leaves and to the degradation of the non-fibre components of the stems, specially the sugars. Depending on the sowing date, the 193 f 9 days after sowing corresponds to a period that range from 14" October to 31d December.
3.2. Biomass Quality Table 1 shows the results obtained concerning the moisture content, the ash, nitrogen and phosphorus contents, of the sorghum biomass obtained in the experimental fields, along the experimental years. Table 1. Moisture content, ash content, nitrogen content and phosphorus content of Sorghum biomass obtained in the experimental fields at harvest (0 = 0 kg Niha; Low = 60 or 75 kg N/ha; High = 100 or 120 kg Niha). Sowing dates Moisture (%)
Nitrogen (% dry matter) Phosphorus (% dry matter)
Low
High
64i7
~
23rdMarch 9" May
6.2k0.8
4.1k0.8
3.610.9
2ndJune
8.2i0.3
6.2+0.2
4.8i0.2
~
Ash (% dry matter)
0
23rdMarch 2"d June
231d March
~
91hMay
0.3 f 0.2
Znd June
0.9 f 0.2
231d March 2"d June
0.2 i 0.1
~
Moisture content decreased along the growing season, ranging from 57% to 71% at the time of harvesting. Moisture content at harvest was not influenced by the sowing dates
Sweet, Fibre and Biomass Sorghum Potential in Portugal
187
and by the level of N-fertiliser and there were no differences among the subspecies of sorghum. However, moisture content, at harvest, was influenced by the climatic conditions: drought autumns leaded to lower moisture contents and rainy autumns to higher moisture contents, in the biomass. Ash content decreased along the growing season attaining the lowest values 193 f 9 days after sowing. No significant differences were observed among ash contents of sweet, fibre and biomass sorghum. However, significant differences were observed among plants obtained from fields with different levels of N-fertiliser. In fact, plants from fields with no application of nitrogen showed significantly higher ash content than plants from fields with low (60 or 75 kg Nka) and high N-application (100 or 120 kg N/ha). In the fields with no N-fertiliser application, plants can only accumulate and absorb nitrogen from the soil reserves. As a result, the production of proteins, nucleic acids, chlorophyll and other biomass molecules, by the plants, is more limited, and, a lower synthesis of glucides through photosynthesis can be observed. Consequently, lower productivities and lower organic matter contents might take place, as it did in the experimental fields at Caparica. Fields sowed between 23rdMarch and gthMay, didn't show significant differences in terms of the ash content of the biomass material, at harvest. But, plants from fields sowed at 2"d June, showed a significant higher ash content than fields sowed earlier. Probably, due to the late sowing date, but also to the disadvantageous climatic conditions observed in this year, the percentage of glucides synthesised was lower than in other experimental years, thus explaining the lower organic matter content and also the lower productivities. Nitrogen content decreased along the growing season getting lowest values 193 f 9 days after sowing. No significant differences were observed among nitrogen contents of sweet, fibre and biomass sorghum and among plants obtained from fields with different levels of N-fertiliser. Fields sowed between 231d March and gth May, didn't show significant differences in terms of the nitrogen content of the biomass material, at harvest. But, plants from fields sowed at 2"d June, showed significant higher nitrogen content than fields sowed earlier. Again, the lowest development of the plants from these fields resulted in a higher accumulation of minerals. As for nitrogen and for ash contents, phosphorus content decreased along the growing season getting lowest values at harvest. No significant differences were observed, in terms of phosphorus content, among sweet, fibre and biomass sorghum and among plants obtained from fields with different levels of N-fertiliser. No significant differences were also observed, among plants obtained from different sowing dates. Sorghum is characterized by low nitrogen and phosphorus requirements leading to less nitrogen and phosphorus fertilization inputs. Table 2 shows nitrogen and phosphorus offtakes observed in the experimental.
A.L. Fernando et al.
188
Table 2. Nitrogen and phosphorus offtakes observed in the experimental fields (0 75 kg Niha; High = 100 or 120 kg Niha). Sowing dates 231dMarch
Nitrogen offtake (kg/ha)
~
91hMay
2ndJune 231dMarch
Phosphorus offtake (kg/ha)
~
9" May
2"d June
=
0 kg N/ha; Low
=
60 or
0
Low
High
27
98
115
50
40
83
46
57
63
15
13
17
According to these results, sorghum depletes the soil nitrogen reserves in ca. 27 kglha, when no nitrogen is applied to the soil. Application of low levels of nitrogen fertilizer into the soil (60 or 75 kg Nha) can also contribute to a depletion of the soil N-reserves. In the fields were a high level of N-fertilizer was applied, 100 or 120 kg N/ha, the Nexported by the plants was approximately the same what was applied, not causing a depletion of N-resources but also, not contributing to the contamination of soils and waters with an excess of N-compounds. P-applied (120 kg P.ha-') was considerably higher than what was uptaken by the plants. Although phosphorus presents low mobility in soils, in the perspective of a sustainable development, P-fertilizer should be applied in balanced amounts. This means that, in the experimental fields at Caparica half of the Papplied would be satisfactory for the crop growth. Nitrogen fertilizer use efficiency was obtained as the slope of a linear regression between the aerial dry matter obtained at harvest and the nitrogen fertilization applied (Fig. 4). I
35000 30000 25000
20000 15000 10000
0
20
40
60
80
100
120
Nitrogen fertilizer applied (kglha)
Fig. 4. Nitrogen fertilizer use efficiency of Sorghum plants obtained in Caparica.
According to Fig. 4, the nitrogen fertilizer use efficiency was very high, 173 kg dry biomass per kg N-fertilizer. Fig. 4 also shows what was previously observed in Fig. 2: that there were significant differences among productivities obtained in fields with different levels of nitrogen fertilizer.
189
Sweet, Fibre and Biomass Sorghum Potential in Portugal
3.3. Potential Energy Production The potential energy produced during combustion of the harvested biomass was evaluated as the productivity (t/ha) x gross heat of combustion (GJ/t) (Table 3). Table 3. Gross Heat of combustion of Sweet, Biomass and Fibre Sorghum and the Potential energy production (GJiha) of Sorghum biomass obtained in fields with different nitrogen application levels. Gross Heat of combustion (GJit)
0 kg N h a
60 kg Niha
120 kg N h a
Sweet Sorghum Biomass Sorehum
12.3 12.9
176 60
326 93
484 118
Fibre Sorghum
11.1
Potential Energy Production (GJiha)
210
In terms of gross heat of combustion, there were no statistical significant differences among subspecies of sorghum, although fibre sorghum presented a lower value. There were also no significant differences among plants obtained in fields with different levels of nitrogen and among different sowing dates. Then, in terms of the potential energy production, differences among the fields were observed only due to the differences in the productivities. According to this, highest values were obtained in the most fertilized fields (120 kg Nha) and with sweet sorghum. Biomass sorghum presented the lowest values but these results were obtained only in one experimental year, in 1997, when the sowing date was at 2”dJune, and when productivities were remarkably low. 1997 year results were not considered for the calculation of the potential energy produced by fibre sorghum. Considering the results obtained for sweet sorghum with a N-fertilizer application of 120 kg Nha, 9000 1 of gasoline, 10000 1 of diesel oil or 15 t of coal, could be saved. Table 4 shows the potential energy produced during conversion of sorghum sugars into fuel ethanol. This potential energy was determined considering that the heating value of ethanol is 26.9 MJ/1 and that the theoretical ethanol yield (ma) can be calculated from the formula [I]: Total sugar content (“3 dry matter) x 6.5 (conversion factor) x 0.85 (process efficiency) x total biomass (tha dry matter). Table 4. Saccharose + Glucose content of Sweet, Biomass and Fibre Sorghum and the Potential energy production (GJha) from Sorghum biomass sugars, that can be obtained from fields with different nitrogen application levels. Saccharose + Glucose (“hdry matter)
0 kg N h a
60 kg Niha
120 kg Niha
Sweet Sorghum
31
91
129
21 1
Biomass Sorghum
36
20
36
60
Fibre Sorehum
35
Potential Energy Production (GJiha)
102
190
A.L. Fernando et al.
As for the gross heat of combustion, in terms of the saccharose + glucose content, there were no statistical significant differences among subspecies of sorghum and there were no significant differences among plants obtained in fields with different levels of nitrogen and among different sowing dates. Then, in terms of the potential energy production, differences among the fields were observed only due to the differences in the productivities. According to this, highest values were obtained in the most fertilized fields (120 kg N/ha) and with sweet sorghum. Biomass sorghum presented the lowest values but these results were obtained only in one experimental year, in 1997, when the sowing date was at 2"dJune, and when productivities were remarkably low. 1997 year results were not considered for the calculation of the potential energy produced during conversion of fibre sorghum sugars into fuel ethanol. Considering the results obtained for sweet sorghum with a N-fertilizer application of 120 kg Niha, 4600 1 of gasoline, 5000 1 of diesel oil or 7.4 t of coal, could be saved. It should be mentioned that, beyond this high potential energy production from sorghum sugars, a significant amount of lignocellulosic material is left as bagasse. This bagasse can be used for pulp production or the cellulosic fraction can be fermented, thereby increasing ethanol yields, and consequently energy yields, by at least 40%.
3.4. Potential Pulp Production
The potential pulp produced after the harvest of the biomass was evaluated as the productivity ( a a ) x fiber content ("3 dry matter)/l00 (Table 5). Table 5. Potential pulp production of Sorghum biomass obtained in fields with different nitrogen application levels. Potential Pulp Production (tiha)
Fibre (%
matter)
0 kg N/ha
60 kg N/ha
120 kg Nlhb
Sweet Sorghum
44
5.7
10.8
15.6
Biomass Sorghum
38
1.7
2.1
3.1
Fibre Sorghum
51
9.4
In terms of the fibre content, fibre sorghum presented a higher value than biomass and sweet sorghum, although differences among subspecies were not statistically significant. There were, also, no significant differences among plants obtained in fields with different levels of nitrogen and among different sowing dates. Then, in terms of the potential pulp production, differences among the fields were observed only due to the differences in the productivities. According to this, highest values were obtained in the most fertilized fields (120 kg N/ha) and with sweet sorghum. Biomass sorghum presented the lowest values but these results were obtained only in one experimental year, in 1997, when the sowing date was at 2ndJune, and when productivities were remarkably low. 1997 year results were not considered for the calculation of the potential pulp produced by fibre sorghum.
Sweet, Fibre and Biomass Sorghum Potential in Portugal
191
4. Conclusions
Productivity was affected by the sowing date and by the level of N-fertilization. Better results were obtained in the fields sowed between 25'h March and lst May and productivity increased with the increase in nitrogen input. Among subspecies, Sweet and Biomass Sorghum presented, globally, higher productivities than Fibre Sorghum. In all the experimental years, highest productivities, in all the fields, were achieved at 193 9 days after sowing, regardless of all the different sowing dates. In terms of moisture, ash, nitrogen and phosphorus contents, there were no significant differences among sweet, fibre and biomass sorghum. Moisture and phosphorus contents were, also, not affected by the sowing date and by the Nfertilization. In terms of ash content, fields with lowest productivities, namely fields where no N was applied and fields sowed after 2"d June, resulted in a higher accumulation of minerals. For the nitrogen content of the plants, no significant differences were obtained among plants cropped from fields with different N-fertilizer. In relation to this parameter, there were also no significant differences among plants obtained from fields sowed at different dates, except for the plants obtained from fields sowed at 2"d June that presented a significant higher value. In the perspective of a sustainable growth and development, nitrogen fertilizer was applied in balanced amounts not contributing to the contamination of water resources with an excess of N-compounds, like, for example, nitrates. But fields where no nitrogen was applied or where 60 kg N/ha were applied, led to the depletion of the soil N-reserves. Calculus of the nitrogen use efficiency, with the experimental results obtained at Caparica, showed that 173 kg dry biomass can be produced per kg N-fertilizer. Compared with other crops, this value is higher, characterizing Sorghum as a low N-input crop. More essays with a higher N-fertilizer application, 180 and 240 kg N/ha, for example, should be experimented, in order to access, if after 120 Kg N/ha, the N-use efficiency show a decay. P-fertilizer application was higher than the amount uptaken by the plants, showing that sorghum can be characterized by low phosphorus requirements, thus leading to less P-fertilizer inputs. Considering the potential energy production and the potential pulp production, highest values were obtained in the most fertilized fields (120 kg Nha) and with sweet sorghum, due to the higher productivities obtained in these fields. So, in the experimental conditions at Caparica, sweet sorghum crop showed a better potential as an industrial and energy feedstock, and to get better results, economically and environmentally, fields should be sowed between 25th March and 1st May, with a high nitrogen input (120 kg Nha).
*
192
A.L. Fernando et al.
References [ 11 El Bassam N. (1998) Energy Plant Species. James & James (Science Publishers) Ltd. [2] Duarte MP, Fernando AL, Guimarfies H, Amparo V, Alves L, Oliveira JFS. (2001) Study of sweet and fibre sorghum crops in Portugal. Effect of climatic conditions and sowing date on the final productivity and on the quality of biomass. In S. Kyritsis, A.A.C. Beenackers, P. Helm, A. Grassi, D. Chiaramonti (eds) Proceedings of the 1st World Conference on Biomass for Energy and Industvy, Seville, Spain, 5-9 June 2000, James & James (Science Publishers) Ltd, London, United Kingdom, 361-364. [3] Duarte P, Fernando A, Oliveira JFS. (1996) Analysis of correlation between productivity and biochemical parameters for the sweet sorghum in Portugal. In Proceedings of the First European Seminar on Sorghum for Energy and Industry. Toulouse, France, 1 - 3 April, 243-247. [4] Duarte P, Fernando AL, Alves L, Amparo V, Silva S, Guimaraes H, Oliveira JS (1998) Characterization of Sorghum potential as an industrial and energy feedstock - the influence of crop management. In H. Kopetz, T. Weber, W. Palz, P. Chartier, G. L. Ferrero (Eds) Proceedings of the 1Oth European Conference and Technology Exhibition Biomass for Energy and Industry, Wiirzburg, Germany, 811 June, C.A.R.M.E.N., Rimpar, Germany, 1046-1049.
PROCESS DYNAMICS OF FOSSIL STEAM POWER PLANTS INDUCED BY THE INTEGRATION OF TRANSIENT SOLAR HEAT
v. SCHERER(’), K. ROTH(’), M. ECK(*) (I)
Department of Energy Plant Technology, Ruhr-Universitaet Bochum, Bochum, Germany (2)GermanAerospace Center (DLR), Stuttgart, Germany
In the current paper a concept to integrate solar steam generated by parabolic trough collectors into a fossil fueled power plant is being investigated. Feed water preheaters are one favourable location to integrate the external heat into the water steam cycle. Since the addition of the external energy causes less steam consumption in the high pressure and low pressure preheaters, less steam is extracted from the steam turbine. This leads to an increasing output of electrical power, which is a profitable product at the energy stock markets. The occurring time dependent processes are studied on a numerical basis, using the dynamic process simulation software APROS from VTT. With APROS a model of the parabolic trough collectors (62 MW maximum thermal output) was built-up, adapted and parameterized with typical lay-out data. Using meteorological data from southern Europe the time dependent steam generation was simulated. The collector model has been connected with a comprehensive APROS-model of an existing fossil steam power plant with a nominal output of 393 MW. The results show, as expected, that the steam generated by the solar collectors varies directly with the irradiation present. In July the power plant output increases at noon by 17.5 MW from 393 MW to 410.5 MW with a maximum gradient of 0.3 MW/min.
1. Introduction and Purpose In order to reduce the emissions of greenhouse gases like COz and to promote renewable energy a specific percentage of the electricity has to be generated by renewable energy sources in some European countries. The legislation in Italy, for example, requires 2 % nowadays, rising to 4.4 % in 2012. A promising technique to meet these goals is to employ a hybrid power plant, where a renewable energy source is added to a conventional and fossil fueled power plant cycle. The great advantage of this concept is, that it can be realized as a retrofit to existing power plants. Solar thermal plants are favourable to be located in southern European countries, that are offering a high insolation. Such kind of power plants, for example with parabolic trough collectors heating a synthetic oil which transfers the heat via heat exchangers to a conventional steam turbine cycle, are operating since many years and have proven their technological reliability. Direct solar steam generation is a possible improvement of the parabolic trough power plants vaporizing water directly in the absorber pipelines and eliminating costly equipment like the heat exchanger and the oil pump. This solar generated steam can be integrated into the water steam cycle of a fossil fueled power plant [ 11. A schematic process diagram is shown in Fig. 1. By this means less steam is consumed by the high pressure and low pressure preheaters. One or more steam extraction lines can be closed and the steam can be used for a very fast and dynamic additional power generation in the steam turbine. The power plant’s efficiency is improved simultaneously. Since the integration of the transient heat sources will lead to substantial dynamic energy shifts within the power plant’s water-steam-cycle, one objective of the developed automation concept is to keep the boiler inlet feed-water temperature constant to avoid any effects on the complex boiler control. The description and analysis of these shifts together with a cycle evaluation using an unsteady process simulation tool are the content of the current paper. 193
V. Scherer. K. Roth & M Eck
194
External (transient) Heat
Fig. 1. Schematic steam power plant with external heat source.
2. Process Simulation Software APROS
The time dependent processes are studied on a numerical basis, using an unsteady process simulation software. Such software tools are based on physical principles and empirical correlations and solve the one-dimensional unsteady conservation equations for mass, energy and momentum [2]: mass balance:
dAp -+-
aApv
az
at
momentum balance:
dApv ~
dAph 0
energybalance:
at
at
=o
+
dApv2 ~
aZ
dApvh
-I--
az
=
dAp += s, az
s,
(3)
On the basis of a detailed software evaluation [2,3,4], reviewing more than 15 simulation tools, the unsteady simulation software APROS from the finish supplier VTT was selected. The Advanced PROcess Simulation environment APROS is one of the very few software tools, that provides a 6-equation model for calculating the conservation equations for water and steam separately. This is of certain interest for the calculation of so called zero flow conditions appearing during start-up and shut-down when the flow velocities decrease to zero. Most of the other tools provide 5-equation models only, neglecting the momentum differences between the liquid and gaseous phase. Further advantages are the various interfaces providing several alternatives for connecting external software to the simulation programme like the IAPWS-IF97 for the thermodynamic properties of water and steam [6]. A long list of reference users working with APROS especially in Europe is another advantage. The number of available power
Process Dynamics of Fossil Steam Power Plants
195
plant components is very large and nearly every plant type from fluidized bed boilers to even coal gasification combined cycles including automation and electrical systems can be modelled.
3. Parabolic Trough Collectors and Solar Irradiation In the current paper parabolic trough collectors with direct solar steam generation are employed to supply steam to the preheating section of a conventional power plant. Although other solar thermal concepts feature higher concentration factors and higher system efficiencies the parabolic trough technology has been chosen because it is a proven technology which has demonstrated its reliability in large scale applications (80100MW) [l]. Tracking the sun from sunrise to sunset the cylindrical parabolic mirrors concentrate the sun’s radiation on the black absorber tubes along their focal line transforming radiation into heat. In these absorber tubes water gets vaporized and superheated to temperatures of more than 400°C.
Fig. 2. Parabolic trough collector LS-3 (schematic) [7].
Fig. 2 shows a typical parabolic trough collector. The aperture area is about 545 m2 with a length of 99 m and the reflectivity is more than 97% [7]. Favourable locations within Europe to operate solar power plants are southern countries like Spain, Italy or Greece. Because of the Italian legislation having a special interest in promoting renewable energy sources the location Greater Bari in southern Italy was chosen. The yearly global irradiance is about 1600-1750 kWh/m2a as it can be seen in Fig. 3. [8]. The yearly direct normal irradiance, that is used by the parabolic trough collectors, is up to 2200 kWh/m2a. In order to design and simulate a parabolic trough collector the monthly average values of irradiation were chosen at first.
196
V. Scherer, X Roth & M Eck
In Fig. 4 the average dkect normal irradiance of the last ten years in greater Baai can be seen for the months January to December. The highest maximum imadiance o c ~ w s in July at noon (884 W/m*) and the lowest is observed in January (349 W/ma).The parabolic trough collector plant which is the basis BOP the current study is not an existing one but is an ~ p ~ r o to a csimilar ~ existing plants and ~nvesti~a~ions in the field of direct solar steam generation [I].
Fig. 3. Global imdiance in Europe per year [S].
Fig. 4. Average direct normal itradiance in Greater Bari.
Process Dynamics of Fossil Steam Power Plants
197
4. Reference Plant and Parabolic Trough Model In order to examine the feasibility of integrating external solar heat into the water steam cycle of a power plant, a model based on an existing reference plant was defined. This reference plant is a typical fossil fuelled conventional steam power plant with an output of 393 MW and 7 .feed-water preheating stages. Using APROS a comprehensive model of this power plant was built-up, adapted and parameterized. Especially the parameterizing is a very time consuming step. A great amount of lay-out data like geometries, isometries, geodetic elevations, thermodynamic data, valve characteristics and automation concepts have to be integrated into the model. In Fig. 5 a more detailed APROS view of the plant’s high-pressure preheater section is illustrated as an example. The automation and measuring concept is not shown. The power plant has been subdivided into more than 500 single components. For each component the one dimensional unsteady differential equations for the conservation of mass, momentum and energy are solved. Heat transfer, heat capacity of solid walls and two phase flow phenomena are taken into account. On the basis of this model several different kinds of plant configurations with or without an external heat source have been simulated. Careful calibration of the model has been carried out to meet the steady-state design and guarantee values. A parabolic trough collector was modelled in APROS. Design values calculated by steady state simulations and taken from the literature were used [ 1-91. The generation of steam with a heat capacity of 60 MW at noon in July was set as boundary condition for the collector design. The simulation is done with parabolic collectors of the type LS-3 (see Fig. 2) with a length of 100 m. Ten of them are added up to a 1000 m collector line. According to this there are 15 lines in parallel necessary to provide a peak load of 60 MW. The collector efficiency is assumed to be constant at 67 %. This simplification is of sufficient accuracy for the simulations carried out in this paper but will be corrected in future work. The collector feedwater pump control guaranties that the water pumped through the absorber tubes gets vaporized and superheated to a constant temperature of 380°C.
5. Integration of the Parabolic Trough Collectors and Power Plant Simulation The concept of integrating the parabolic collectors into the cycle is illustrated in Fig. 6. The temperature level of the solar steam is high enough to be integrated into the highpressure preheater section and to preheat a part of the boiler feedwater. Thus the preheater’s steam consumption is reduced and the steam remaining in the turbine can be used for an additional power generation. The condensate leaving solar preheater 1 is still at a temperature level to preheat a part of the feedwater in the low pressure preheating section (solar preheater 2). The main part of the feedwater is heated in the preheaters 1 to 4 by extraction steam conventionally.
198
V. Scherer, K. Roth & A4 Eck
Bypasses with solar preheaters are integrated into the water steam cycle of the power plant as shown in Fig. 6. In the solar preheater 1 the superheated steam gets condensed via heat transfer to the feedwater. The feedwater is switched into the bypass by the main control valve MCV. The solar heat and correspondingly the feedwater mass flow through solar preheater 1 is permanently varying with a maximum at noon caused by the changing irradiation in the course of the day.
Fig. 5. High pressure preheating section in APROS.
Fig. 7 illustrates that the water at the outlet of the parabolic trough (PT) gets warmed up in the early morning (t=5h), is then vaporized (t=6h) and finally superheated to a constant temperature of 380°C. It is obvious that the PT outlet temperature follows quite closely the irradiation curve. In parallel to the irradiance the feedwater mass flow through the parabolic trough increases till noon and then decreases. Because of mass release effects during the vaporization phase in the morning the mass flow characteristic through the absorber tubes shows a peak. Effects like these in combination with long dead times resulting from long absorber tubes and low flow velocities complicate the development of reliable automation concepts for the water mass flow through the absorber tubes in the parabolic trough collectors [9]. Therefore a pilot control concept based on measured irradiance data to pre-adjust the water mass flow has been applied. The steam generated in the PT is to be integrated into the preheating section of a conventional steam power plant. A control strategy which is illustrated in Fig. 8 has been developed that allows the addition of the external heat into the preheating cycle without affecting the feed-water temperature at the boiler inlet (25 1"C) and therefore without affecting the complex boiler control.
Process Dynamics of Fossil Steam Power Plants
I
199
Time
Fig. 6. Integration of the parabolic trough collectors into the cycle.
This control strategy consists of 2 control loops employing the feedwater temperatures as controlled variables: In the first loop the feedwater temperature at solar preheater 1 (SP1) exit is measured by the temperature sensor TC1. The PID-controller opens the main control valve (MCV) and the feedwater is directed into the bypass successively in order to keep the given temperature set point. Because of the lower irradiance in the morning the parabolic trough supplies hot water instead of steam which leads to lower feedwater temperatures. In order to avoid any mixing of “cold” bypass feedwater with the mainstream, it gets discarded by the control valve CV as long as it is below the set point temperature. In the evening the irradiance decreases and the bypass feedwater is discarded again.
Solar Preheater 1
Parabolic Trough
\
/
J
Solar P r e h e a t e r 2
PT Pump
\ /J
1
1
1
Fig. 7. Irradiance, temperature and mass flow characteristics in July.
V. Scherer, K. Roth & M. Eck
200
,
y....................... W
.......
cv
~
0
.......................,, 8
P A ,$ .
.
. . .
~
Fig. 8. Bypass control concept.
Another temperature sensor TC2 can be found in the feedwater line downstream of heat exchanger PH7 and is part of the second control loop. This temperature measurement is used to control the steam extraction valves V1 and V2 to keep the mainstream feedwater temperature constant. The control concept in the low pressure preheating section is working similarly. Fig. 9 illustrates the mass flow characteristics in the high pressure preheating section during the course of a day in July. As it can be seen in Fig. 9, up to 179 kgls of feedwater can be guided into the bypass to be preheated by the solar generated steam. In consequence the mainstream feedwater mass flow decreases from 320 kgls to 141 kgls. Since the reduced feedwater mass flow needs less steam the controller closes down the extraction steam valves V1 and V2 as it is shown in Fig. 10. The processes appearing in the low pressure preheating section are very similar.
+Mass
5
7
9
11
13
Row through HP bypass
15
Time
Fig. 9. Mass flow characteristics.
17
19
21
20 1
Process Dynamics of Fossil Steam Power Plants
20
1-
5
7
11
9
13
15
17
I9
21
Time
Fig. 10. Extraction steam mass flows into PH6 and PH7.
Closing the steam turbine extraction lines lead to a significant increase of the steam power plant’s power output as shown in Fig. 11. The power increase is following the solar irradiance characteristic nearly without any time delay - this is an essential result of the investigations carried out. The combination of integrating an external heat source into the cycle and closing the extraction steam lines leads to a rise of the steam power plant output from 393 MW up to 410.5 MW at noon in July.
1200
1000
$
800
.-
C
600 C
m
s
-
400
200
0 5
7
9
I1
13
15
17
19
21
Time
Fig. 11. Irradiation and power output characteristics.
This corresponds to a power increase of approximately 4.5 percent (17,5MW). Compared to the results in July, the potential power increases in the other months are lower, for example in January the maximum additional power rises about 6.7 MW. On the basis of
K Scherer, K. Roth & M. Eck
202
this simulation data an efficiency factor can be calculated, defined as a ratio of the parabolic trough thermal power output integrated into the cycle and the achievable electrical power output increase at the steam turbine's generator. This efficiency factor is about 28 % at noon. In addition to the use of average irradiance data for the simulation a single day in July showing strongly varying irradiation caused by clouds was selected as a test case. The simulation results can be seen in Fig. 12. Because the irradiance is temporarily too low to preheat the feedwater in the solar preheater 1 to 25 1"C, the feedwater is not led back to the mainstream but discarded. In these periods no additional power is generated in the steam turbine. Hence significant output losses are to be noticed in the example presented. However, the results favourably show that even fast changes in solar irradiation can be managed with simple control strategies. 404
400
-
.E
P
c
0 L
396
g -8 m c
I-
* 6
8
10
12
14
16
18
20
22
Time
Fig. 12. Simulation of a single day in July.
6 . Conclusion The integration of external solar generated steam into the water steam cycle of conventional steam power plants is a promising concept to promote renewable energy and as a consequence to reduce the COz emissions. The simulation results based on meteorological data of a location in Italy show a power increase following the solar irradiance characteristic nearly without any time delay. This can be managed even with simple control strategies. Since the peak of electricity consumption is usually to be found at noon, the additionally produced power matches very well the electricity demand of industrial countries. The efficiency factor defined as the ratio of the parabolic trough thermal power output integrated into the cycle and the achievable electrical power increase at the steam turbine's generator amounts 28% maximum.
Process Dynamics ojFossil Steam Power Plants
203
Acknowledgements
Building up the reference plant model in APROS was carried out within the Arbeitsgemeinschaft Turbomaschinen 2 (AG Turbo 2) and was financially supported by the BMWiiBMWA Germany (Ministry of Labour and Economics) and ALSTOM Power Support Mannheim. The authors would like to acknowledge the provision of detailed power plant data by the project partner ALSTOM. References [ l ] Eck M. (2001) Die Dynamik der solaren Direktverdampfung und iiberhitzung in
Parabolrinnenkollektoren.ed. VDI-Verlag, Dusseldorf. [2] Roth K, Scherer V, Behnke K. (2002) Wirtschaftliche Nutzung von transienten Waermequellen im Wasser-Dampf-Kreislauf. Proc 8. Statusseminar der AG Turbo, Koeln. [3] Roth K, Scherer V, Reinig G, Gebhardt B, Behnke K. (2003) Steigerung der Leistungsdynamik von Dampkaftwerken durch Einbindung externer transienter Warmequellen in die Speisewasservonvarmung. Proc VDI Fachtagung Fort schrittliche Energiewandlung und - anwendung, Stuttgart, 1746: 303-3 16. [4] Scherer V, Roth K, Behnke K. (2003) Enhancing the dynamic performance of electricity production in steam power plants: Integration of transient waste heat sources into the water steam cycle, Proc 7th International Conference on Energy for a Clean Environment, Lisbon. [ 5 ] Silvennoinen E, Juslin K, Hanninen M, Tiihonen 0, Kurki J, Porkholm K. (1989) The APROS software for process simulation and model development. Espoo, Technical Research Centre of Finland, Finland. [6] Accuracy of APROS’ (1998) Fast steam tables for real time applications. VTT, Technical Research Centre of Finland, Finland. [7] Deutsches Zentrum f i r Luft- und Raumfahrt; Institut fiir technische Thermodynamik; Germany, www.dlr.de/tt. [ 8 ] Suri M, Huld TA, Dunlop ED. (2005) PVGIS: a web-based solar radiation database for the calculation of PV potential in Europe. International Journal of Sustainable Energy 24 (2): 55-67. [9] Eck M, Steinmann WD. (2000) Dynamic behaviour of the direct solar steam generator in parabolic through collectors: A simulation study. Proc SolarPACES, Internat. Symp. on Solar Thermal Concentrating Technologies 10: 101-106.
This page intentionally left blank
DIRECTIONS FOR RENEWABLE ENERGY IN CANADA’S SMALLEST
PROVINCE A. TRIVETT
Department of Engineering, University of Prince Edward Island, Charlottetown, PEI, Canada Prince Edward Island’s economy relies almost entirely upon energy that is imported. Up to 200MW of electrical power is imported through an underwater cable from the province of New Brunswick, while the bulk of our energy is imported by tanker truck and ship in the form of oil, gasoline and propane. Political, economic and environmentalpressures are forcing us to re-think the sources and supply routes of our energy. A recent provincial government policy draft has suggested that Prince Edward Island, Canada, should be generating 10%-15% of its electrical energy from wind power by 2010. Since we are already generating close to 10% of our average power from wind, several industry representatives have proposed that this target is too low. It can be argued that we have sufficient wind resources to be planning the generation of 10-20% of the energy for the ALL the MARITIME PROVINCES from PEI wind resources, which would mean more than 200MW of electrical capacity from wind turbines. Doing so could make PEI a net energy exporter, but concern over the variable nature of wind power makes the local utility very concerned over the viability of this option. At the same time as we debate the large-scale renewable energy strategy, there is a new demonstration project to produce hydrogen fuel from wind generated electricity being planned. The hydrogen will be used in a small, remote community to provide energy storage in order to replace wind-derived energy at times of low wind speeds, as well as to supply wind-derived fuels for vehicles. This paper will review some of the energy realities of a small Canadian island, outline the PEI Wind-Hydrogen Village project, and discuss the practicalities of making major energy supply changes in a rural setting.
1. Energy Supply in Canada’s Smallest Province
As a nation, Canada is a net exporter of energy. We are the largest trading partner of the United States, and a major commodity which we supply to our neighbour is energy. In 2002, Canada exported 6% of its total electricity generation to the United States, while reciprocally importing only 3%. 55% of this electricity was produced from hydro electric dams. The remainder came from nuclear power or conventional coal and oil combustion
HI. While as a whole, Canada has a wealth of energy, and this is being utilized from a wide variety of sources, the picture is quite different for some of the regions and provinces which make up the Canadian Federation. Prince Edward Island is the smallest province in Canada, consisting of no more than 140,000 souls [l]. We are situated on combustionturbine
3%
(,
receipts
deliveries
3%
6%
7
steam nuclear 11%
r-
conventionalsteam
55%
22%
Fig. 1. Total electrical energy generation in Canada for 2002 [ 11.
205
A . Trivett
206
what is fondly thought of as the "garden in the Gulf ', an island in the Southern portion of the broad Gulf of St. Lawrence. The island is dominated by fishing and farming industries, with a small, but growing manufacturing sector. There is no commercial energy resource industry, despite recent high-profile exploratory wells drilled in hopes of finding natural gas or oil. At the present time, the prospect of a home-grown oil and gas supply is very slim. The electrical supply of our energy is mostly from imports through an underwater cable to the power grid in New Bmnswick. Fig. 2 shows the breakdown of electrical energy generation for the province of PEI. The Statistics Canada source does not include wind generation in its database, so the wind component of our power base is estimated using data from the number and size of wind turbines at the PEI Energy Corp site in North Cape. Steam generation is a very small component of the total due to the electrical utility's policy of maintaining a 60MW generation plant in Charlottetown as a backup facility. This plant is operated primarily on extremely cold winter days when the supply from New Bmnswick is in highest demand locally in New Bmnswick. A 38MW gas turbine facility is also maintained for backup power generation. PEI 8 NB Energy Generation
'%
Total conventional steam ge"Wall0" 2%
f
Total hydro generation 11 4%
)
Total rkceipls wind generation 9 1% 0 4%
Fig. 2. Electrical energy generation on PEI: the vast majority of power is purchased from off-island sources in New Bmnswick. The graph to the right shows the break-down of the combined New Bmnswick and PEI electrical generation [ 11.
It is clear from Fig. 2 that the vast majority of our energy is not under control of our governmental or corporate entities. When looking at the source of this power from New Brunswick, it is apparent that the mix of energy comes from primarily oil burned in one of several stationary combustion plants. The fuel for the generation plants is from oil imported by ship into the port of St. John, New Brunswick. 2. The Total Energy Demand
Electrical power generation is not the complete picture, however. The use of electrical energy from the power grid is not as great as the demand we have for fuels used separate from the electrical grid. The energy from electrical grids is only a minor part of the total.
Directions for Renewable Energy in Canada's Smallest Province
207
Fig. 3 shows the total use of energy from all sources, for all uses. The consumption of transport fuels dwarfs our use of electricity. Energy Consumption on PEI 30000
25000
Fig. 3. Total energy consumption on PEI. The component of electrical consumption includes all grid power imported directly from New Brunswick, which, from the previous figures, 60-70% originates from burning
fossil fuels [l].
A closer look at our use of petroleum products is important to understand what actions can be taken in the future. The Island has no public transport system that is island-wide. In recent months, a very limited bus system has been offered within the capital city, Charlottetown. Ridership has not been very high in the early stages. The only means for travelling for the vast majority of residents is by private automobile. Fig. 4 shows the seasonal changes in the use of different fuels. Gasoline and Diesel fuel are in much higher demand in the summer months as a result of our very large influx of tourists between June-September. Tourists to the island travel by automobile almost without exception. The businesses on PEI are oriented towards the tourist trade to a large degree, and thus the demand for fuel from the businesses coincides with the peak tourist season, and in some businesses, energy intensive activity leads the tourist season by a few months as businesses prepare for the summer. In contrast to the high summer demand for transport fuel and business-related demands, we have a very cold winter climate and this creates a high demand for home heating fuel from November-March. At the same time, fewer residents on the island and
208
A . Trivett
less travel reduces the consumption of gasoline and diesel fuel. The combined effect of these two factors tends to smooth some of the seasonal variations. The peak season for total energy demand is, as a result, in the spring months, March-May.
Mar-97
Mar-98
Mar-99
Mar-00
Mar-01
Date
Fig. 4.The seasonal variation in usage of different fuels on PEI. The cold climate results in a very high demand for home heating oil in the winter. Our large tourism industry results in high demand for gasoline and diesel fuel from May-September [ 11.
3. Alternatives for the Future
3.1. Wind Power At the same time as liquid fuels are very expensive, the public perception of wind power is positive. Maritime Electric, the private utility which distributes all of the electricity to PEI, offered customers a chance to purchase blocks of “green” power from the PEI Energy Corporation’s wind turbines at North Cape in 2001. 389 Residents and 32 business customers voluntarily bought all of the 1097 5OkWmonth blocks for an additional 35% premium to support the use of wind power. This was widely seen as a strong indication of support. The Island has a proven wind resource, with capacity factors of up to 42% at North Cape, and potentially similar factors at other sites across PEI. This resource can be a benefit through electricity exports. There are presently 16 Vestas V47 turbines installed by the PEI Energy Corporation at North Cape. A prototype V90 was installed by Vestas Canada in the fall of 2003. Future expansion of this site is expected. Shallow waters surrounding PEI may make future expansion of ocean wind farms a practical alternative if land use issues become a concern.
Directionsfor Renewable Energy in Canada's Smallest Province
209
3.2, Wind and Hydrogen The greatest challenge for the adoption of wind power is the inherent variability of the supply. Numerous alternatives exist for complementary renewable energy sources that can balance the variable nature of wind energy. The alternatives have not been researched, and this must be considered to be a high priority for the engineering community. To this end, many groups worldwide are investigating the use of hydrogen (H2) as a component in new energy strategies. Implicit in environmentally-mindedH2 policies is that the fuel, Hydrogen, must be produced from zero-emission renewable primary energy sources. Demonstrating potential for generating significant amounts of H2 from renewable resources such as wind or solar power is an important aspect of these programs. We see a future potential for H2 not only as a replacement for battery storage of wind energy, but as a means to use wind power for existing transport fuels. We have proposed to develop a small coastal community of 50-100 people demand in the range of 75-150 kW of electrical power (using PEI average estimates). If we factor in the fuels used for transport and home heating, this may amount to approximately 300-500kW of power to supply all their needs. Such a community would benefit from the installation of 3-5 lOOkW wind turbines. The turbines and associated H2 equipment capital cost must be compared with the alternative cost to the residents for continuing to purchase all their fuel and energy from outside of the community. In this light, 100 residents who may currently spend 1500-3000$per year, per person on all fuels (a figure which is on the low extreme of what is currently spent) could, in theory, finance installation of 1.5-5M$ of infrastructure for self-generated electricity and fuels. To be justified, the equipment would operate over a 10-15 year lifetime. If such a system is shown to be technically practical, the economics may also prove to be attractive to local residents. Such a community will serve to address not only the supply of electricity from renewable sources, but also the stabilization of supply from wind generators in addition to a means of replacing fossil fuels in transportation. Since, as was shown previously, the transportation and home heating use of fuels outweighs the use of direct electricity, a solution which serves all three needs is one worth studying. 4. Conclusions
The fuels we use are supplied to Prince Edward Island via a tanker which docks in Charlottetown approximately three times per month. Recently, while speaking to 10 and 11 year old school children at a local public school, I asked them if they knew where their energy came from. After some interesting guesses, I asked how many had noticed the particular tanker which docks regularly within sight of the school. All the children recognized the ship, having seen it numerous times, but none connected its regular routine with our driving habits and our energy use.
210
A . Trivett
A significant challenge for energy policy in Canada is making a direct connection between the source of supply and our consumption in the public. Being part of a seemingly infinite supply of electrical power and liquid fuel, regardless of price, can delay public action towards sustainable practices. For the first time in two generations, the price of gasoline has reached l.OO$/litre, a psychological price barrier never before crossed. All other petroleum based fuels are at record-setting high prices. Despite the opportunities for a shift in direction towards more sustainable energy future, there is no concerted effort from all parties to develop a coherent plan. There has been no ongoing research into energy alternatives since the late 1980’s. Projects, such as the North Cape wind farm, and a district heating plant in Charlottetown burning municipal waste are isolated projects. This is slowly changing. We have seen in the past year recognition from the provincial government of the value from R&D in renewable energy. The challenge for our future is to integrate the most appropriate energy solutions for a small Northern Island. Acknowledgements The author wishes to acknowledge support form the University of Prince Edward island office of Research, and well as the Dean of Science. References [l] Statistics Canada: Table 128-0002 - Supply and demand of primary and secondary energy, http://cansim2.statcan.ca/cgi-win/cnsmcgi.exe. [2] Statistics Canada: Table 127-0001 - Electric power statistics, monthly, http://cansim2.statcan.ca/cgi-win/cnsmcgi.exe. [ 3 ] State of the Environment, Published by the Prince Edward Island Department of Fisheries, Aquaculture and Environment, June 2003, available online at http://www.gov.pe.ca/photos/original/fae-soe-report.pdJ:
EXTERNALITIES AFFECTING THE VIABILITY OF WIND POWER FOR HYDROGEN PRODUCTION N.KASSEM Royal Institute of Technology, Dept. of Energy Technology, S-100 44 Stockholm, Sweden Off-peak electricity from wind power farms is a valuable and cheap source of energy, which is hlly competitive, with fossil fuel and nuclear power generation. Wind power is one of the lowest-priced renewable energy technologies available today depending on the wind site, the type of wind turbines, and project financing. Coupling of wind turbines to high temperature steam electrolyzers would have the potential of providing low cost hydrogen source. For offshore wind farms to be economic, they have to be of sizes greater than 150 MW using large turbines (> 1.5 MW). The study uses methods of risk analysis to assess the economic viability of wind power and novel electrolytic processses to evaluate the simultaneous effect of multiple input uncertainties on system output. The results show that stochastic variations in wind speed, electricity prices, and interest rate have direct impact on investment capital requirements, and hence on the cost of hydrogen.
1. Electrolytic Hydrogen from Renewables
Hydrogen is an alternative energy carrier, which would significantly contribute to the growth of sustainable energy systems. It is considered a renewable fuel if produced from renewable energy resources such as hydropower, solar-and wind-energy. The opportunities for innovation and economic growth in hydrogen energy utilization are virtually untapped, and many industrial nations are dedicating their efforts to establish a leading position in this expanding field. A great deal of R&D work is conducted worldwide to utilize hydrogen as a new energy source in district heating, power utilities, and transportation sectors. Hydrogen can be produced in virtually unlimited quantities from renewable resources. Thus it is a primary element to sustainable growth in power generation and transportation systems. Conventional technologies for hydrogen productions [ 1-31 include primarily partial oxidation of heavy oils, steam reforming of light hydrocarbons, coal gasification, and water electrolysis. Electrolysis is an energy intensive process, where the cost of electricity has a direct impact on hydrogen economy. Although not economically competitive, electrolytic hydrogen provides high product purity and flexible operation modes. Electrolysis plants may operate over a wide range of production capacities, thus they are more convenient for electricity load leveling either at the generation site or at any location directly supplied from the grid. Wind power is an emerging technology with great potentials for electricity generation and has been grown steadily, over the last 10 years. The main growth, however, has been in Europe, where government policies and high conventional energy costs favor the use of wind power. Almost 3 1 GW, of wind power were already installed worldwide by the end of 2002 [4]. The major challenge in the process of wind energy assessment is how accurately the air- flow and hence the error in wind energy can be predicted and how the total capital investments are estimated. Failure to account for uncertainties would result in deterministic estimates that tend to overstate performance and underestimate costs. Accounting for uncertainties in emerging technologies, requires applying methods of uncertainty analysis in any life cost assessment. This study provides an assessment of the risks and pay-offs associated with electrolytic hydrogen production using electricity from offshore wind farms and heat from the combustion of bio-fuels. 21 1
N. Kassem
212
2. Offshore Wind Power
Wind power is a potential renewable energy resource in many countries with substantial installed capacities, Germany (12GWe), Denmark (2.88GWe), Spain (4.8GWe), and India (1,2GWe). The U.S. Department of Energy (DOE) recently launched the Wind Powering America initiative with goals to increase federal use of wind energy to 5 % by 2010, and to satisfy at least 5% of the nation's electricity demand using wind power by year 2020. As part of its strategy, the EU-Commission has launched a target to increase the Union's installed capacity of wind power from 7 GWe in 1998 to 75 GWe by year 2010 [4]. Wind power is a source of clean, nonpolluting electricity, and at favorable wind sites, it is fully competitive with fossil fuel and nuclear power generation. The available wind resources at any given site are crucial in wind energy implementation. The minimum economic value for the average wind speed at a given location is about 6-8 d s , a condition that is satisfied in many places in the United States, Europe, and Asia. In Antarctic, however, the average speed for 75 YOof the year is about 3-4 folds of this value, which make wind power more cost-competitive. Finally, good wind sites are often located in remote locations far from areas of electric power demand. Most of these problems have been greatly reduced by properly sited wind farm or through technical development. Main drawbacks of wind power are attributed to the fact that it is an intermittent electricity generator and hence it does not provide power on an "as needed" basis. The potential power yield, expressed in Watt varies with the cubic power of the wind speed as given by Eq. 1. E, = 0.125 zr Ow2p V 3
Watt
(1)
In this study, however, the uncertainty in wind speed is accounted for through including a
1
XI0
random number generator in the velocity component of Eq. (1). The wind variation for a typical site is usually described by a Weibull distribution which has a probability density fimction given by Eq. (2). The statistical distribution of wind speeds varies from place to place around the globe, depending on local climate conditions, the landscape, and its surface. The Weibull distribution may thus vary in its shape factor k, and the scale factor c. If the shape parameter is exactly 2, the distribution is known as a Rayleigh distribution. Wind turbine manufacturers often use this distribution to present the standard performance figures for their machines The economics of wind power are compelling and the cost of installing and operating wind turbines has been fallen dramatically by almost two thirds between 1981 and 1995. The typical capital costs of large onshore, small onshore, and offshore wind farms in year 2000 is estimated between 1000 to 1700 $/kW. The annual investment requirement for achieving 10% of the world's electricity from wind energy is about $3 billion in 1999. It is expected however, to reach a peak of $78 billion in 2020 (green peace, October, 1999). These are only a fraction of overall global energy investments of $170-200 billion per year in 1999. Wind energy is a domestic, reliable resource that
Wind Power for Hydrogen Production
213
provides more jobs per dollar invested than any other energy technology including fossil and nuclear fuels. Employment implications of the 10% target are quite significant; and would create 1.7 million jobs in manufacturing, installation, and operation worldwide. 3. High Temperature Steam Electrolysis
Well-developed and novel processes for hydrogen production, based on both fossil and non-fossil fuels, are frequently reviewed in the literature, Winter and Nitsch [ 13. In this study, however, the main focus is on the High Temperature Steam Electrolysis (HTE), detailed in [5], using bio-fuels supplementary thermal source. The total energy requirement (AH) for the electrolysis process consists of an electrical energy part (AG) and a thermal energy part (Q=T.ds) to provide the heat demand of the endothermic electrolytic reactions. The basic idea of HTE-technology is that a substantial saving in electric power requirement can be achieved by increasing the operating temperature. Up to 38% of the total energy demand can be satisfied by a cheaper energy source than electricity, thus resulting in a significant improvement in the overall efficiency of electrolytic hydrogen production. The standard free energy for water dissociation is A c O 2 9 8 =237.4 kJ/mol., equivalent to the standard potential E'29.9 ,1.230 V, while the standard enthalpy of the electrolytic reaction is -AH298= -286.4 kJ/mol, equivalent to standard potential of 1.483 V. The difference has to be supplied by other energy sources. The amount of heat needed (T m) for the endothermic electrolytic reactions can either be supplied by internal resistivity losses of the cell itself due to high current densities (300-500 mA/cm2) or by direct heat input from high temperature heat source. Compared with conventional electrolysis, the electrolytic decomposition of steam requires quite different technologies, whereby the liquid electrolyte is replaced by a special solid electrolyte having good ionic conduction properties. The solid electrolyte is usually in the form of a porous metal oxide ceramics, such as zirconium oxide stabilized with yttrium oxide (Zr02 - Y2 O,), which allow the oxygen ions to migrate by ionic conduction to the anode compartment. The operating temperature is governed by the electrolyte resistance and electrode polarization. However, the present achievements in thin-film technology allow the fabrication of electrolysis elements with electrolyte film thickness in the range of 10-50 pm. This is expected to lower the operating temperature of the HTE process by 100-200 "C. Under these conditions, the operating temperature is more likely to be determined by electrode polarization rather than electrolyte resistance. A basic requirement for the solid electrolyte ceramics is that its structure and composition would remain unchanged during electrolysis processes. The conditions for reliable HTE operations is discussed by Perfiliev [6] The electrode polarization, due to electric current flow, has significant effect on the electrode/electrolyte interface, where the oxygen activity decreases at the cathode while it increases at the anode. A schematic flow diagram of the HTE process is shown in Fig. 1. Saturated steam from renewable resources such as thermal solar power or waste heat, after being superheated in a heat recovery unit, is mixed with hydrogen recycle stream and preheated to 1100 "C (2012 OF). This takes place in ceramic conductor tubes in a bio-fuel -fired furnace before being fed to the electrolyser. At this temperature, only ionic conduction can cause oxygen ions to move across an ionic conductor. Single cells, produced from the ionic conductors arrayed in various geometrical forms, are connected
N. Kassem
214
in series or parallel to form a battery of cells. A stack of these batteries are hooked together to form a large-scale electrolyser. 4. Performance and Economic Models
Engineering models include both performance and economic models based on thermodynamic relationships, conservation equations, cost scaling, thermo-economic
Offshore Wind farm
Rectifiers
Seperator
0, Storage
0, compressor
Fig. 1. Schematic flow diagram for hydrogen production using renewable resources
relationships, and published cost curves for generic process areas. These models are mostly empirical in nature with the parameters being fitted using real or statistical data. The cost models must be compatible with performance model simulation results, which include key parameters needed to determine the total capital investment, running costs, and cost of hydrogen. In order to account for design and economic uncertainties, and to apply methods of risk analysis, a stochastic modeling capability is developed, which allows input uncertainties to be represented using probability distributions. The modeling capability provides an automatic random sampling of model inputs/outputs, and a systematic method to evaluate the simultaneous effect of multiple input uncertainties on key plant outputs. It is also possible to apply methods of statistical and sensitivity analyses to model outputs, and to illustrate the results as cumulative distribution functions (cdf). Like any other modeling tool, the outcomes of the stochastic methods are very much dependent on the type and quality of the data and judgments provided by the user. A summary of performance and cost models used in this study is given below.
A. Electrolvser Model [7]
+ C, + CR
(Cinv),
= CA
C,
=
C,
=
c, + 6 I
=
pru
c, c,
A
Wind Power for Hydrogen Production
215
B. Hydrogen liquefaction. & Compression [S] Fj, = 3 . 3 elf, F&
kW
= R T l n ( p , ) Q l q c I3600 kW
(G")lLq,c,p
= 25422(p)0.61
C. Hydrogen Storage [S].
(Ci,,)slo
= 9.17 1 exp ( f ( v > )
v = volume of H , storage
f ( v ) = nonlinear function of volume Models for estimating the cost of electrolytic hydrogen are available with varying degrees of detail and sophistication in the literature [7, 81. Generally, electrolysis plants are designed in modular form, which are not sensitive to the economy of scale. Design data for water electrolysis are readily available in the literature and can be adapted for the given model. Economic input data, however, are rather rare especially for large-scale operation of advanced technologies. 5. Economic Assessment
For power utilities and chemical industries, there are generally accepted guidelines regarding the methodology of developing cost estimates. The levelized cost of electrolytic hydrogen is determined based on the methods and assumption described in the TAG report of EPRI [9], which include typical cost estimate parameters. The cost of hydrogen includes the cost of compression, liquefaction, and storage, and the allowance for byproduct credits. The main cost items considered in this assessment are: 1. Facility Invest. Costs (FIC), i.e. equipment and machinery. 2. Indirect Costs (IDC) including engineering, installation, contingency, and administration costs. 3. Direct Costs (DC) including Feedstock, operation and maintenance (O&M) costs, and utilities. 4. General Administration: (GA) including insurance, taxes and inflation costs, royalties, and depreciation cost. All cost estimates are evaluated by applying a GNP price deflector factor of 1.32, for 1990198, obtained from economic indicators published in IEA 1999 edition. The inflation rate is taken as 2% and all costs are expressed in 1999 US$. With limited cost data, cost scale factors ranging between 0.5-0.8, are used to determine the approximate capital cost of the new plant. Indirect cost (IDC), which includes installation, piping, building instrumentation, auxiliaries, and structures are estimated at 2.0-2.5 of delivered equipment cost. Engineering, start-up cost, plant contingencies, and interest during construction were taken as 35 % of FIC whereas the working capital is estimated at 3% of FIC. The outside battery limits was estimated at 35 % of the inside battery limits. The contingency represents additional costs that are expected to occur and yet it is not explicitly included in main parts of cost estimates. It is a single large and least
216
N. Kassern
documented expense item. Contingency factors, estimated by conventional approach, are inadequate to quantify the risk of cost increases in new technologies and tend to be underestimated as discussed by Milanese [ 101. By-product credits, mainly oxygen, were taken into account and being estimated at $45 ton-' oxygen. The power generated by the wind farm is based on a standard 1.5 MW turbines, which have a capital investment cost of $1000-1400/kW installed.
6. Uncertainty Analysis Prediction of the future commercial-scale performance and costs of a new alternative technology are often based on limited data from small-scale version of the plant. These predictions are generally expressed in terms of single-point deterministic estimates based on the best guessed values of key performance, design and economic variables, while ignoring whatever uncertainty associated with them. Such estimates are inherently uncertain and generally provide misleading estimates, which can have serious implications on project planning and resources allocation Thus, the ability to conduct uncertainty (risk) analysis is of special importance in the context of ongoing assessment of emerging technologies, where technical and economic figures-of-merit are not wellestablished. The extent of uncertainty will vary from one parameter to another depending on the state of development of the selected technology, the level of detail of relevant estimates, and future market values. The extent and severity of uncertainties reflects the degree of confidence the designers might have towards input assumptions of the applied model. Uncertainty in key performance and economic parameters e.g. wind energy potential, feedstock prices, contingency factors, and interest rates would lead to uncertainties in key output variables such as overall efficiency, plant capital cost, cost of electricity, or cost of hydrogen. Studies of advanced power generation, prepared for EPRI by Hager and Heaven [ l I], include uncertainty analysis, which consider only uncertainties in cost related input parameters, in particular capital costs. The type of distribution chosen to characterize uncertainty reflects the amount of information available on a given variable. Probability distribution defines the range of values, which the process variables can assume, and the likelihood of occurrence of each value within the defined range. If data are not available, uncertainties can be expressed in terms of probability distributions based on the judgment of technical and economical experts. Compared to conventional sensitivity analysis, the development of estimate of uncertainties requires detailed thinking about the possible outcomes and their relative likelihood. This eventually would lead to anticipating a broader range of problems or system payoffs, that otherwise might have been overlooked. For practical problems with many variables subject to uncertainty, the combinatorial explosion of possible sensitivity scenarios becomes unmanageable. This type of analysis is inadequate and provides no insight into the likelihood of obtaining any particular results that might be of use in the decision making process. The main objective in using such tools is to assess the risks and the payoffs of the new concepts compared to conventional ones. To account for uncertainties in wind power based hydrogen technologies, a stochastic modeling tool is required. Two types of such tools are frequently used in the literature, namely the Latin Hypercube Sampling (LHS), where the distribution of a given input is divided into intervals of equal probability, and one sample
Wind Power for Hydrogen Production
217
is taken from within each interval at a random manner. The second tool is the MonteCarlo Sampling (MCS) method, where samples from the input parameter distribution are taken randomly from the whole range. Detailed descriptions of both procedures are given by Ang and Tang [12]. Considering the advantages and disadvantages of both methods, the MCS method has been selected as a suitable stochastic modeling tool for conducting the required uncertainty analysis in this study. The, the stochastic modeling tool provides explicit and quantitative measures of uncertainty in key figures-of-merit, thus helping the analyst to focus the efforts on reducing their effect.
7. Results and Discussion The main focus of the techno-economic assessment carried-out in this study is to predict the uncertainties associated with the cost of electrolytic hydrogen from renewable resources. The main parameter values used in this analysis are given in Table 1. Table 1. Main parameter values ~
500 $/mz fm = 0.17 k = 2.0 p = 100 $/kW qc= 0.75 c,
=
c , = 5000 D,=64m c = 0.125 F=0.05 $/A Q = 10-70 metric ton H2/hr
A stochastic modeling capability using Monte-Carlo sampling technique has been applied to monitor the effect of input uncertainties on a number of performance and economic figures-of-merit of the combined offshore wind farm and high temperature electrolytic hydrogen plant. Production rates in the range of 10-70 metric ton/ hour of hydrogen ware considered suitable for the study, while the wind turbine capacity factor (CF) is assumed to be 0.4. The main input uncertainties considered are: wind speed, cost of electricity, and the interest rate. The uncertainty in wind speed is described by a Weibull distribution, Eq. (l), with parameters k and c being given in Table 1. The uncertainty in electricity and biofuel costs, and the interest rate are represented by normal distributions having mean and standard deviations as shown on the diagrams The probability density (pdf) and cumulative distribution functions (cdf) for the selected key input uncertainties are shown in Fig. 2 . The engineering models were run repeatedly, using values, generated from those distributions using Monte-Carlo sampling method. For multiple input parameters, the values of all input are sampled simultaneously during each computational cycle. The number of repetitions is equivalent to the sample size, which is selected based on the desired precision of the output distribution. The results are in the form of numerical values that can be treated statistically as sets of experimental values and are given in the form of cumulative distribution functions (cdf), as shown in the following figures. 7.1. Figures-of-Merits
Evaluation of the electrolytic hydrogen technology requires the access to quantified measures of the projected performance and cost parameters. These measures include
218
N. Kassem
k = 2.0
f
0.4 0.05
0.2
0
10
5
15
20
Wind s p e e d m/s
cs =
5 10 Wind s p e e d
m/s
0.01
0.03
15
20
0.00
0.4
ft
0.4
0.2
0.2 0
0
0.02
0.01
0.03
0.04
0
0 0 '
2
4
6
I
8
. 0
lntrest rate %
ft
0.2 0 0.8
0.4
= 2%
0.2
1
1.2
W p invest K$/kW
0.04
z 2
1
V 4
6
6
Interest rate %
p = 6% -3
0.02
cost of elect. COE $/kWh
Cost of elect. COE W k W h
1.4
2
4
6
8
W p invest. $/kW
Fig. 2. Probability functions of input uncertainties.
performance and economic figures-of-merits (FOM), which can be determined using the engineering model outputs. For each set of randomly selected input values, a set of FOM evolves at each Monte-Carlo simulation cycle. This procedure is iterated repeatedly until sets of sufficient size emerge with statistical properties that are meaningful to the decision-maker. Probability distribution functions of some relevant figures- of-merits are shown in Fig. 3. Here, the top left plot represents the cumulative probability of the generated power by a single wind turbine having CF= 0.4, and reveals that there is about 80% probability to get a power output up to 2 MW. This applies to the entire range of production rates. The plot on the top rights of Fig. 3 shows, however, the amount of wind energy generated per square meter of the swapped blade area of the turbine. It indicates an 80% probability to generate up to 800 W/m2 using 64 m-blade diameters. The next four diagrams represent the probability of capital investments for the main facility items namely, the electrolyser, wind turbine, liquefier, and bio-fuel (BF)-furnace. The diagrams, however, reveal that there is a great deal of uncertainty (big variance)) associated with the last three items, while the electrolyser shows a reduced degree of uncertainty (small variance). They also show that the electrolyser has the highest capital investment cost of about 4$/GJ H2 with a probability greater than 95%, while the BF-furnace has the lowest investment cost of about 0.06 $/GJ H2 with 50% probability. The capital investment of
219
Wind Power for Hydrogen Production
both the wind turbine and liquefier are quit close to 0.6 $/GJ H2 with a probability of 75% and 50% respectively. In case of wind power investment, the probability distribution is almost linear indicating a higher degree of uncertainty. Of all the four production items, the capital investment of the liquefier showed some degree of sensitivity towards the production rates of hydrogen. The bottom diagrams show the total capital requirement TCR and the return-on-investment (ROI) expressed in $/GJ H2 produced. The diagram on the left reveals that the certainty in TCR, which is a measure of the cost of hydrogen (COH) lies in the range of 12-18$/GJ H2 with 90% probability that the expected value is about 15 $/ GJ H2. The ROI on the right has a certainty range of 8-12 $/GJ H2 with an expected value of 10$/GJ H2 at 90% probability. Both diagrams indicate that the effect of production rate on TCR and ROI is insignificant. 1
I
500
Wind power M W
1000
1500
Wind Energy swaped W/m2
0.9,
0.7
0.5
Wind power invest. $/GJ H 2
8
1 5
10
Electrol. invst. $/GJ H 2
0.2 0
0
0.2 0.4 0.6 Liquef. invest. $/GJ H 2
0.8
Tot. capit. requir. T C R $/GJ H2
0 0.02 0.04 0.06 0.08 B F Furnace invest. $/GJ H2
Ret. on invest. ROI $/GJ H 2
Fig. 3. Probability functions of some figures-of-merits.
7.2. Multiple Sensitivity Analysis
To assess the technical and economic viability of new process concepts, the designer need to get more information on how input uncertainties would affect some key figures-ofmerit. To achieve this aim, the designer should conduct some sort of sensitivity analysis
N. Kassem
220
to identify input uncertainties and to characterize their effects on process outputs. In this study, however, multiple sensitivity analysis was conducted by allowing process inputs to change randomly in accordance with their probability distributions, while observing their impact on a number of key economic figures-of-merit. The results of this study are illustrated in Fig. 4. The two plots at the top represent the effect of interest rate and the cost of electricity (COE) on the cost of hydrogen (COH) expressed in $/GJ at different production rates. The COH showed almost the same behavior with respect to both input parameters, but showed no distinct sensitivity towards the economy of scale. A gradual increase in the interest rate in the range of 2-6% resulted in a linear increase of hydrogen cost from about $15 to $20 /GJ H2. Beyond the selected range of interest rate, there is a dramatic decrease or increase of the expected hydrogen cost. A similar behavior is resulted, when the cost of electricity is allowed to change in the range of $10-$30/ MW, which is illustrated by the plot on the top right of Fig. 4.
-- I
2
4
6
5'
8
10
20
Cost of elect.
Interest rate %
30
I
0.1
'
I
2
4
6
8
I
-
Lu
2
Interest rate %
Interest rate %
z: 1
0.8,
0.1
n
I
prod. rate 10-70 ton/hr H2
A I
2
4
6
Interest rate %
0
C O E $1 MWh
8
0.01
1
2
4
6
Interest rate %
Fig. 4. Sensitivity analysis of some key figures-of-merits
1
a
22 1
Wind Powerjor Hydrogen Production
The next four plots illustrate the impact of the intere te on the facility investment costs . The diagrams show a gradual (FirC), in $/GJ H2, of the main fom production fac linear increase of about 25% in FIC as the interest rate increases in the range between 2In a ~ ~ i t i othe n ~facility cost (FIC) o f all items, except for the elecrolyser, showed ct sensitivity towards the economy of scale. ~ncreas~ng the production rates in the range between 10-70 metric tonshour, indicated by the direction of the arrow, would result in reducing the i n v e s ~ e n cost t by almost 50%. The electrolytic unit, however9is n o ~ a c~o ~n sy~ c t in e ~modules, thus i s not sensitive to the econorny of scale.
20
1s 10 S
0
7 6
8. a 4
a 3
8 2 1
10 20
30 40 SO
60 70
0
10
M e t r i c tonlhr H2
a
30
40
50
60 70
M e t r i c tonlhr H2 10
10 ‘
20
E
6 4 2
0 5 0
20
30 40
50
60 70
Metric ton/hr H2
10
20
30 40 SO
60 70
M e t r i c ton/hr H2
Fig. 5. Relative magnitudes of various investment costs.
The relative order of ~ a ~ of ~various ~ economic d e ~ ~ e s - o € - m eexpressed ~t, in $/GEE, for the p r o d ~ c ~ ~ofo hydrogen n in liquid (LEI21 and gaseous (GH2) forms are i ~ ~ ~ sin~Fig. a t 5e. ~Itemized capital such as facility investments (FIC), indirect versus the production rates in the first f v ~ o costs (BDC), and direct costs (DC) are stack plots at the top. Here the dominant item is the direct costs, since it include the cost of e ~ e c and ~ other ~ c feedstock. ~ ~ The capital costs o f the faci~~ties are shown in r n i d ~ ~ e plots and clearly indicate that the electrolytic facility has the most dQminantcost f o ~ ~ ~ by wind farm and ~ i ~ u e f a c t ~costs. o n Finally the two bottom plots compare the ~ e e ~ s t o c ~
~
N. Kassem
222
(electricity, bio-fuel and steam) costs indicating that the cost of electricity is the dominant cost, especially for liquefied hydrogen.
8. Conclusions 0
Wind power has a great potential for electricity generation and is fully competitive with fossil fuel and nuclear power generation; Improved design of wind turbines would enhance the economic viability of offshore wind farm as a source of cheap electricity; Characterization of uncertainty by their probability distribution provides more insights into performance and economic estimates compared to deterministic ones; Interactions among uncertainties in key performance and cost parameters are propagated through the model to yield an explicit indication of output variable uncertainties; The capital investment cost of the electrolytic hydrogen plant, and electricity cost requirement are the most dominant cost items in electrolytic hydrogen production.
Nomenclature A
AF CA Cinv COE COH CR CS Dw Ew
DC FIC GRR IDC I P Pr
Q RC ROI TCR
Total Electrode area Annuity factor Area related costs $/MWh Capital investment costs $ Cost of Electricity Cost of Hydrogen $IGJ Rectifier related costs $ Auxiliary related costs $ Wind turbine blade diameter m Wind power Watt Direct costs Facility investment costs $ Gross revenue required MM$/yr Indirect costs Total Current Amper Power kW Pressure ratio Flow rate kg H 2 h Running costs Return on investment MM$/yr Total capital requirement MM$/yr
U V WE WP c k cdf Ca > Cs
fm V
X
Cell voltage V Windspeed m!s Wind Energy Wh Wind Power kW Weibull factors Cumulative DF Constants Figure-of-merit storage volume m3 random number
Greek symbols p Constant parameter $lkW 6 Constant parameter $/A p air density qc Compressor efficiency Subscript Ele liq CmP
Electrolyser H2 Liquefier H2 compressor
Wind Power for Hydrogen Production
223
References
[ 11 Winter CJ, Nitsch J. (1989) Hydrogen as an Energy Carrier. Springer-Verlag, Berlin. [2] Hammerli M. (1984) When will electrolytic Hydrogen become competitive? Int. J. Hydrogen Energy (IJHE) 9: 25. [3] Basye L, Swaminathan S. (1997) Hydrogen Production Costs - A Survey. Report Sentech, Inc Bethesda, MD. [4] The European Wind Energy Association, EWEA, Briefing report, October 2003. [5] Doenitz W. et al. (1980) Hydrogen production by high temp. electrolysis of water vapor. Int. J. Hydrogen Energy (IJHE) 5 : 55-63. [6] Perfiliev, M. (1994) Problems of high temperature electrolysis of water vapour. Int. J. Hydrogen Energy (IJHE) 19 (3): 227-230. [7] Stuki S. (1991) The cost of electrolytic hydrogen from off-peak power. Int. J. Hydrogen Energy (IJHE) 16 (7): 461-467. [8] Syed M. et al. (1998) An economic analysis of three hydrogen liquefaction systems. Int.J. of Hydrogen Energy (IJHE) 23 (7): 565-576. [9] EPRI TAGTM(1986) Technical assessment guide, 1: Electricity supply, EPRI rep. P-4463 SRI, Palo Alto. [ 101 Milanese JJ. (1987) Process industry contingency estimation. Publ. N-2386-PSSP, Rand Corp., St. Monica, CA. [ 111 Hager RL, Heaven DL. (1990) EPRI GS-6904, Palo Alto, CA. [12] Ang AH-S, Tang WH. (1984) Probability Concepts in Engineering Planning and Design. 2, Decision, Risk and Reliability. Wiley, New York, NY.
This page intentionally left blank
NON-TECHNICAL BARRIERS TO LARGE-SCALE WAVE ENERGY UTILISATION A.J.N.A.SARMENTO'',~),F. NEUMANN'~), A. BRITO-MELO'~) (I)
Instituto Superior Tecnico, Lisbon, Portugal,
(21
Wave Energy Centre, Lisbon, Portugal
In order to enable a large-scale implementation without violating the principles of sustainability, it is of high importance to avoid conflicts that partly emerged from the context of the large-scale implementation of other RE technologies, mainly with respect to environmental and societal sensitivities. In addition to the lessons learnt from other technologies, there is a number of specific aspects for the utilisation of wave energy that have to be taken into account with sufficient care in the early phases of development. In the present paper the approach of the WEC (Wave Energy Centre) to seize these aspects, is presented, based on results of a case study on these non-technical barriers to large-scale development of wave energy basically arising from environmental sensitivity aspects and competing uses of the ocean. The utilisation of wave energy is generally deemed environmentally benign and only results of several years of operation at significant scale may confirm the validity of this evaluation. Therefore, emphasis in this phase is given to the aspects of conflict of use, which can easily be underestimated in the implementation of this renewable energy. Although the study has been elaborated for the situation in Portugal, several findings are expected to have generic character.
1. Introduction
The present energetic situation world-wide and in particular in EU countries demands a high priority to intensify and diversify the exploration of endogenous energy resources in the near future. According to geographic and geologic characteristics, the resulting renewable energy (RE) mix can assume very different compositions, among which hydropower and wind energy already cover notable shares in a number of countries. Among the RE resources that have not yet contributed in a significant scale, ocean wave energy is likely to play a major role in the f i h r e for a number of regions, such as countries at the European Atlantic coasts including the UK, Ireland, France, Spain and Portugal. Despite the vast energy potential of ocean waves and various attempts to tap this for the last 30 years, technical drawbacks and high development and installation costs were the main obstacles for the implementation of this RE technology. Meanwhile, a number of systems have reached the prototype stage and some promising achievements indicate that it is likely to overcome technical difficulties and yield sufficient maturity for large-scale utilisation in the medium term. On the background of the priority that is presently given to RE implementation on the community political level, also the prospects for economic viability of wave energy implementation are encouraging. In order to enable a large-scale implementation of wave energy compatible with the principles of sustainability, it is of high importance to avoid mistakes that were partly committed in the context of the large-scale implementation of other RE technologies, mainly with respect to environmental and societal sensitivities. In addition to these lessons learnt over the last two decades, there is a number of specific aspects concerning the utilisation of wave energy that have to be taken into account with sufficient care in the early phases of development. It is expected that large-scale implementation will take place by deploying offshore submerged or floating devices at depths around 50 m, grouped in wave farms with specific configurations and dimensions depending on the type of technology. In contrast to wind energy, wave energy has not converged yet to a unique technology. Many different concepts of devices have been proposed so far and presently some have 225
226
A.J.N.A. Sarmento. F. Neumann & A. Brito-Melo
achieved the prototype status and are under tests at real sea conditions. An emerging industry in wave energy domain has accompanied this development with strong enthusiasm, and from the viewpoint of technological development and economical considerations, large-scale implementation in Portugal and other countries may be realistic by 200712008. In Portugal there is a substantial “know-how“ in the field of wave energy. R&D on wave energy utilization was initiated in 1978 at Instituto Superior Tecnico (IST), joined by INETI in 1983. The work conducted by this team in the numerical and experimental fields lead to the construction of the 400 kW onshore OWC pilot plant at the island of Pico, Azores [ 11. This construction was funded by the EC, EDA (Azores utility) and EDP (Portuguese mainland utility), and was technically supported by a number of Portuguese companies. In 1997, IST established collaboration with the Dutch company Teamwork Technology, developer of the offshore submerged device “Archimedes Wave Swing (AWS)”, whose 2MW prototype is presently being deployed offshore Povoa de Varzim (www.waveswing.com). The Portuguese %now-how“ in the field of wave energy, together with the existence of a good natural resource at the Atlantic Portuguese coast and good infrastructures, make Portugal a favoured country for playing an important role in the industrial development of wave energy conversion, compared to its European cogeneris. This poses a number of subsequent aspects that have to be considered in view of such development, based on a planning that weighs the eventual impacts or conflict of uses with other sectors of activity. There are several aspects to have in mind that may constitute potential barriers to large-scale implementation of wave energy. Among these are certainly still a number of “know-how” barriers related to technological experience or availability of specific expertise and processes. However in the present study priority is given to the aspects not directly related to technological development, which are assumed to be overcome in the next years.
2. Competing Uses of the Maritime Space Having in mind the increasing maturity of wave energy technology, it is opportune to investigate the main concerns and issues over the Portuguese coast to large-scale development of wave energy basically arising from competing uses of the ocean space. In the maritime domain there are several players and entities with different interests and competences, so that it is important to try to understand and, if possible, to conciliate the different viewpoints. The experience from offshore wind energy has demonstrated that this is a difficult task, which has to be approached as early as possible. The lack of sensitivity of the entities involved in a large-scale wave energy project may in some cases generate conflicts and some concern by the communities more directly involved in the traditional activity sectors of the maritime domain. This may constitute a barrier to the realisation of a project, and to the acceptance of the technology. In countries with significant coastline, the near-shore area has been a traditional domain of the fishermen communities and the interference with these communities can turn out to be problematic for the development and implementation of the project. This study identifies the main problems, provides a perception of relevant issues by interested parties and gives recommendations for preventive mitigation measures. On the other hand also potential benefits and opportunities of large-scale wave-energy development are briefly discussed. Successfully overcoming the non-technological barriers enables to
Non-Technical Barriers to Large-Scale Wave Energy Utilisation
227
provide a framework to facilitate the licensing process of offshore wave energy increasing investor confidence and market penetration. The methodology applied in the study has consisted in defining the different areas in the maritime domain subject to restrictions and thus identifying a set of uses susceptible of conflict with the implementation of wave energy projects, focussing on the activities around the bathymetric line of 50 m where it is expected that large-scale development will concentrate. To achieve this objective, a number of entities with activities or jurisdiction in the maritime field has been consulted, namely those related with the environment, fisheries, authorities, submarine cables and prospecting1 exploitation of petroleum. The restricted areas were compiled in a map of the coastline of continental Portugal, including main bathymetric lines and port locations (see Fig. 1). 2.1. Fishermen’s Society
In Portugal fishery represents in economic terms an important sector with special social relevance. The implementation of large-scale wave energy projects in an area, which is traditionally regarded as domain of the fishery community requires a careful approach of the involved fishermen community. The barriers imposed to the development of wave energy may be mainly a result of the fishermen acceptance to its installation. In order to identify eventual conflict of uses with the fishery activity, the maritime authorities along the West coast of Portugal have been contacted. The opinions collected in these meetings indicated that the eventual conflict is related mainly with the area of activity of the different types of fishery: the traditional fishery done by small local vessels close to the shoreline in shallow waters (in a distance less then 3 miles from the coast) do not achieve in general the 50 m water depth except in the Southwest of Alentejo and in the Vicentine coast where the bathymetric of 50 m is much more closer to the shoreline. Meanwhile in this area the devices may be placed more distant from the coast as the slope of the continental platform is very soft and the 60 m water depth is about 4 miles from the coast. The industrial fishery is only allowed at a distance beyond 6 miles from the shoreline, which in the Portuguese coast occurs in depths of more than 50m, except for the area about 50km south of Oporto. According to the maritime authorities, the wave energy farms installed in water depths of 50 m will generally not interfere significantly with the main types of fishery and thus should not represent any serious conflict with the activity of the fishermen. However it is necessary to consider navigation corridors between the farms for the ships to go round the area delimited by the wave energy farms, and to be sure that they are neither placed in hot-spots from the fisheries point of view, nor in navigational corridors between hot-spots. Fishermen assume in general an attitude of apprehension and doubt to the introduction of a new activity in the sea. In regard to wave energy they assume a defensive opinion and consider the wave energy farms as a barrier to the fishery activity, recognising however the benefits of wave energy utilisation. The general attitude is not against the installation of large-scale wave energy projects, as long as adequate lengths between the farms are maintained and security and insurance aspects are taken into account by the promoters of the project. The question of attribution of responsibilities in case of an accident is a relevant aspect for the fishery communities. Finally they show clear interest in being informed of the plans for future projects and to be involved in the licensing process.
228
A.J.N.A. Sarmento, F. Neumann d; A. Brito-Melo
R
.EGEND Coastline 50mdepth 100 rn depth 200 m depth 6 miles distance
~
iESTRlCTEDAREAS I-,-"
;,g&j Nature reserves ZdCd
,g Submarine cables >TKT.,
' %_* 2
Militar use
~~
Port authorities
ZOMPETING USES Traditional fishery Industrial fishery
Fig. 1. Wave energy resource conditioning map for Portuguese nearshore.
Non-Technical Barriers to Large-Scale Wave Energy Utilisation
229
2.2. Port Authorities The maritime traffic is done in a large distance from the coast (typically more than 5-10 miles), where wave energy devices are not expected to be installed. The only conflict may be the entrance channels to the harbours. According to the maritime authorities, a sector of 135" at the entrance of the main harbours (Sines, Setubal, Lisbon, Figueira da Foz, Aveiro, Leixoes and Viana do Castelo) is recommended to be reserved just for navigational use.
2.3. Submarine Cables In the Portuguese coast there are submarine telecommunication cables protected areas in the near-shore of Lisbon and Sesimbra. Within these areas it is forbidden to perform any activity susceptible of damaging the cables. Therefore wave energy devices can not be installed in these areas. However they coincide with zones where it is not expected the utilisation of wave energy due to the fact that the shore is protected from the main incident waves and on the other hand due to the presence of the two main ports - Lisbon and Setubal (nearby Sesimbra) -with an intense traffic. 2.4. Military Use Some coastal areas are periodically reserved for military exercises. On the Portuguese coast these are announced by radio navigational warnings. These areas are delimited and represented in nautical maps. Although there are no restrictions to maritime use outside the exercise periods, it is apparent that wave energy devices should not be installed in these areas. However this does not constitute a relevant conflict of use with wave energy as one of the zones is located in the vicinity of the area in Sesimbra designated as submarine cable protected area; and the second one is a small area in front of the Vicentine coast (Southwest of Portugal).
2.5. Marine Resource Exploration There is some activity Concerning hydrocarbon exploration in Portuguese waters, although to date there are no prospection and extraction activities. The discovery of petroleum in Cadis Sea (Southern Spain) has valorised also the Portuguese coast and an increase in the exploration activities is expected. These activities may affect the installation of wave energy devices just in specific localisations. The prospection and extraction processes would not be significantly affected by wave energy installations, as it is possible to do inclined perforations. Further the device locations may be deviated if an important petroleum field is discovered. It is recommended that plans for large-scale offshore wave energy implementation is synchronised with the activities of the petroleum industry, in order to avoid potential conflicts. It should also be noted that there can arise a number of technological synergies concerning the use of offshore installations that may be explored by such an approach. 2.6. Nature Reserves
In the maritime domain there are some areas delimited as natural parks and reserves. At the bathymetric line of 50m the following areas were identified: 0 Berlengas-rectangular area including Berlengas archipel;
230
A.J.N.A. Sarmento, F. Neumann & A . Brito-Melo
Southwest of Alentejo and Vicentine coast-2 km wide corridor along the entire extension of the coastline. In these areas the installation of wave devices it is not recommended. The limit of the 2km corridor on the Southwest of Alentejo and Vicentine coast reaches the 50m depth contour, but as referred before when discussing the fishery activities, the slope of the continental platform is very soft and the devices can be installed in a depth of 60m (3 to 5 miles offshore), neither interfering with the natural reserve, nor with the traditional fishery. Although other natural reserves in the Portuguese shoreline do not reach the 50m water depth contour, they need to be taken into account in the underwater crossing and installation of electric cables from the wave energy farms through land. However a large scale utilisation of wave energy will possibly require an underwater transmission electrical line, extended from North to South along significant parts of the coast, thus reducing the number of connections to the shore.
3. Environmental Concerns Beyond the conflict of use with nature reserves, which indeed originate from environmental concerns, the general consequences of large scale utilisation of wave energy are of highest importance to the technology’s future potential. Lessons have been learnt from related technologies, such as wind energy, and it is essential to give importance to a number of potential concerns, such as visual impact and noise. 3.1. Visual Impact
The experience from offshore wind energy clearly identifies the visual impact as a main concern. The coastal areas are usually important for recreational uses and so the visual impact of shoreline projects should be considered. Offshore devices will have certainly less visual impact. The visual impact may be relevant in areas where the water depth required is attained at small distance visible from the coast. The visibility of the offshore devices from the coast will depend much of the signalisation with lights and therefore its effects must be considered. Some recommendations in order to improve the public acceptance have been proposed from wind offshore such as to avoid recreational areas, to consider a planning phase well open to the public with eventual involvement of the public and to discuss and analyse the configuration of the farms, number of devices and dimensions previously to any final decision. It was verified that many times the turbines are assumed as an intrusive element on the landscape and tests have been done by different suppliers in order to investigate the most adequate colour for the towers. However, the comparison with wind energy appears very conservative when taking into account the structural dimensions of wave energy devices. These are either submerged or surface-piercing but with substantially lower heights than wind turbines. It can therefore be expected that the visual impact of wave energy parks will be negligible for offshore wave energy and very reduced for near-shore exploration.
3.2. Noise Noise is mainly caused by the mechanical components or air-flow in case of OWC devices. The noise may affect the habitat of marine species and if audible by the
Non-Technical Barriers to Large-Scale Wave Energy Utilisation
23 1
population on the coast may reduce the public acceptance. Care needs to be taken that the noise emitted from devices does not cause interruptions to migratory pathways or breeding grounds [2]. In the onshore plant at the Islay, Scotland it was verified that the turbine was responsible for unpleasant noise which was audible above the background of wind and waves and a solution was adopted to reduce the level of noise. To demonstrate that the noise resulting from wave energy devices is an impact that can be mitigated to nonsignificant levels is a very important task for the future of wave energy and that is why the experience gained in individual projects is very important. It shall be noted that such impacts are considered to be negligible for offshore wave energy, as the noise levels 4-5 miles away from the shore will not seriously affect population on shore. 3.3. Emissions
Wave energy generation does not bring along Greenhouse Gases emissions or any other atmospheric pollutant. However the emissions may result from other stages of the life cycle of wave energy technologies (Resource extraction, resource transportation, materials processing, manufacture, transportation, construction, operation, maintenance decommissioning and product disposal). Several detailed studies with other renewables energies demonstrated that in general the emissions released during the manufacture of the materials the most important, depending on the industrial practices and regime of pollution control in the country [3], but similar to what happen with the conventional processes of energy production. 4. Further Barriers
Although the competing uses and the environmental impacts are considered to be the most important potential restraints to large-scale wave energy implementation in the future, there is a number of other factors that presently limit a more efficient development and a faster implementation. These are mainly related to the fact that the development of utilising the ocean as major energy resource is a relatively recent challenge to society, resulting in the following main barriers:
4.1. Marketing Deficits Marketing barriers require the characterisation of the potential market for shoreline and offshore systems at the Iberian Peninsula scale, but also, whilst at a less detailed levels, at the European and word scales. For such procedure, different wave energy technologies should be grouped by similar requirements concerning minimum or maximum water depth, ocean bottom and subsoil characteristics, maximum allowable tidal amplitude and currents, minimum wave resource for economical competitiveness and technological support and infrastructure. 4.2. Administrative Procedures and Legal Aspects
Administrative barriers are mainly related to permits procurement. The experience in Portugal, both with the Pic0 and the 4 W S plants shows that this is a very important area.
232
A .J.N.A. Sarmento, F. Neumann & A . Brito-Melo
In Portugal, and possibly in many countries, the regulations for the sea are less well established, no experience exists in licensing wave energy plants and a large number of public departments are called to participate in the permit procurement. This turns the permit procurement a very complicated and time-consuming process. In the case of the AWS permit about 10 different public departments were involved (the Defence, Environmental and Public Works ministries were directly involved with more than one department from each). Regarding this aspect, a similar approach like the new Strategic Environmental Assessment (SEA), as practiced in the United Kingdom, could straighten the procedure. In this methodology, the National waters are divided in several zones, for which the necessary permission and licensing procedures and the involved competent entities are clearly defined. Concerning the permissions and licensing to feed electricity into the electrical grid, the restrictions to power quality are quite clearly defined by Portuguese law, however the quantity and the timing for large-scale feed-in is not easily sizable. The work to be done includes the identification and comparison of legal procedures and constraints related to the access to the electrical grid, with the objective of removing the constraints and simplifying and standardising the procedures required for wave energy. The identification of the electrical grid connection points availability and characteristics in the areas of interest for wave energy is an important step towards enabling large-scale implementation. 4.3. Financial Barriers
Financial barriers are related to the feed-in tariff of the produced energy, the financial schemes to support the technology development (funds or special interest loans) and their evolution with the development of the technology. The situation in Portugal concerning the feed-in tariff is extremely favourable compared to other European countries, which is why a strong development can be expected during the next years. In particular in the United Kingdom, various players of the wave energy community stated that this is the most desired kind of governmental support, as it is calculable and rewards the actual cost efficiency of devices. However, in the present phase of pre-commercial development, additional measures to finance the development are essential. In a following phase, the potential involvement of large industrial players or risk capital is much more likely to occur than at present, which is why strong incentives are particularly important before reaching that point. They are also require to assist suppliers of services and components to target their expertise to wave energy, since the development of wave energy and the potential benefits of exportation of the technology can not be achieved without a competitive and competent cluster of suppliers. 5. Conclusion
The main non-technical barriers are related with competing uses of navigation activities, fishing resources and, to a small extent, petroleum prospectiodexploitation. Restricted areas such natural reserves, military exercises areas and submarine crossing cables need to be taken into consideration as well but are not critical in Portugal. Another aspect that may constitute a barrier to the development of wave energy projects is the public acceptance. It can play an important role for the critical path of large-scale projects and may impose serious conditioning to the time-schedule and realisation of a project. Public
Non-Technical Barriers to Large-Scale Wave Energy Utilisation
233
acceptance barriers can have various origins; often they are a mixture of societal aspects and competing uses, as identified in this paper. A major barrier could be the licensing process that although recently adoptions in the law were made, could play an important role in a country, in which this process takes several years for wind parks and mini-hydro plants, technologies with substantial operational experience. Generally, the environmental impact of the wave energy technologies presently relevant is considered to be much reduced, provided that the site selection is done prudently and a controlled planning underlies for sensitive locations. In spite of having a small to moderate expected level of environmental impact, the penetration of wave energy technology will highly depend on this aspect. Thus, it is important to assess the environmental impact of different wave energy technologies and the limitations that may result from it, both in what concerns the market and alternative technological developments. A substantial baseline for this will be provided by the pilot plants presently operational or in the deployment phase. Measurements on these plants are foreseen and will allow updating the present understanding and providing means for a adequate legislation, before passing from the demonstration to an industrial phase. On the other hand, there is to date no substantial operational experience that can confirm this evaluation. It is mandatory to investigate and monitor the first systems of significant scale with respect to potential impacts, in order to identify their extent and to adopt mitigation measures where necessary. To conclude it is perceived that the non-technical barriers resulting from conflicts of interest do not constitute major barriers for the large-scale development of wave energy; however they need to be taken into account in a strategic plan for large-scale wave energy development. The competing uses are indicated in a map of the coastline of continental Portugal (see Fig. I), including main bathymetric lines and port locations. From this figure can be seen that about 300 km of the Portuguese west coastline are potentially available to wave energy utilization.
Acknowledgements This work was partly sponsored by the Marie Curie Fellowship Scheme (contract EVK32000-55 135), the Portuguese Science and Technology Foundation (FCT; contract SFRH/BGCT/11789/2003) and Direccao Geral de Energia, Portugal (Portuguese Directorate General Energy).
References
[I] Falcao AFO. (2000) The shoreline OWC wavc power plant at the Azores. Proceedings of the 4thEuropean Wave Energy Conference, Aalborg, Denmark, pp. 42-48. [2] Thorpe TW. (1999) A Brief Review of Wave Energy. DTI, United Kingdom. [3] Wavenet - Results from the work of the European Thematic Network on Wave Energy, ERK-CT- 1999-2001 2000-2003, European Community, March 2003.
This page intentionally left blank
SOLAR POWER AGRICULTURE: A NEW PARADIGM FOR ENERGY PRODUCTION U. BARD1
Dipartimento di Chimica, Universita di Firenze. Polo ScientiJico di Sesto Fiorentino, 50019 Sesto Fiorentino (Fi) Italy,
[email protected] ~
The present paper argues that solar power technologies, such as photovoltaic panels and wind turbines, should be regarded as a form of agriculture in the paradigm named here “solar power agriculture”. These technologies share several features with traditional agriculture, for instance the basic process of energy conversion by collecting solar light over relatively large areas. Solar power technologies and agriculture also share a dynamic resource growth curve which is s-shaped in contrast with that of non-renewable fossil resources which is bell shaped. Conventional agriculture uses solar light to produce food, textiles and other products, whereas solar power technologies mainly use light to generate electric power. Framing renewable energy within a well known and accepted paradigm, that of agriculture, leads to major advantages in terms of public image, with the consequent unlocking of the financial resources needed to increase the use of renewable resources. The concept of solar power agriculture provides a bridge between the present situation in which renewable play a marginal role in the world’s energy production to a future situation when renewables will instead play a major role.
1. Introduction
The world is not (yet) running out of fossil fuels, but progressive depletion is making these fuels expensive as the result of factors which include increasing direct costs (e.g. the need to exploit smaller oil fields) and increasing external costs related, for instance, to the greenhouse effect generated by the products of combustion. Rising costs will cause the production of fossil fuels to reach a peak at some point in time and subsequently to decline as extraction becomes progressively less and less convenient in economic terms. Some projections [l-41 indicate that the peak for crude oil production could take place within the first decade of the 2 1st century, to be followed by the other fossil fuels peaking in later decades. The production curve of a mineral resource is often described as “bell shaped” [5]. In contrast, the curve for agricultural production is, in principle, s-shaped, i.e. it reaches a plateau when all productive land is exploited. Agriculture is sustainable, at least in theory. Historically, both soil and water may turn out to be a non-renewable commodity when misused [6-71, however it is at least a physical possibility for agricultural societies to reach a stable state. The subject that the present paper addresses is how fossil fuels can be replaced as a source of energy using renewable solar technologies in the “agricultural paradigm”, that is in order to abandon the “bell shaped’’ curve and reach, instead, a plateau of production at levels comparable to the present values. In this approach, called here ”Solar Power Agriculture” or, also “agri-energy” electric power produced on agricultural lands is considered as just another form of agricultural product, to be approached with the same social and economic approaches which are commonplace for conventional agriculture. Several previous reports (e.g. 8-16) have discussed the ultimate limits of solar technologies in terms of land requirement. These studies arrived to the conclusion that a fraction of the earth’s equatorial deserts would be sufficient to provide abundant energy 235
236
U. Bardi
for humankind's needs. However, large scale renewable energy plants in deserts do not appear to be on the verge of materializing. The main problem appears to be the need to attract the huge investments needed, both for the plants and for the related energy vectoring system. Here, a different approach is considered on the basis of the idea that renewable energy can to make significant inroads in power production only if its introduction is gradual and it starts from a relatively small scale. This approach leads to the idea of embedding solar plants within areas used for conventional agriculture. It is an approach, in fact, that has already been tested for wind energy in countries such as Denmark and Germany. Obviously, the possibility of expanding this strategy to obtain a significant fraction of the worldwide energy needs depends on a number of factors: 1. Technological factors: can renewables produce enough energy for the needs of humankind without competing with conventional agriculture for land needed for food production? 2. Cost: Even when embedded in conventional agricultural production, won't renewable energies remain too expensive? 3. Public opinion. Can the relatively large extension of land needed for solar energy be made acceptable to the public? The present paper will show that the answer to these questions is, in principle, positive and that the concept of solar power agiculture has a strong potential in order to speed up and favor the introduction of solar renewable technologies in the world.
2. Technological Factors Human beings have been exploiting solar energy for their needs since the time when agriculture was first developed, about 10,000 years ago. As a technology, agriculture was hugely successful and in time it diversified and expanded until, at present, most of the land "potentially suitable for agriculture" on the planet, about 130 million km2, [ 17-181 is exploited for human needs. The fraction of this land which is exploited for food production, i.e. the sum of pasture land and arable land corresponds to approximately 50 million km2that is about 38% of the total [ 17-181. Agriculture uses solar energy to carry out chemical synthesis, in turn based on photosynthesis, a biochemical process which uses solar light as the source of energy necessary to synthetize carbohydrates from water and carbon dioxide. Further stages of the biochemical process can produce a variety of organic substances. Mainly, agriculture is used to produce foodstuff and fiber for textiles. Traditionally, agricultural biomass has been also used as energy source for residential heating but in recent times it is increasingly exploited to produce electric power and, in some cases, fuel for traction. Other physical and chemical processes to exploit solar light for human use have been known for a long time but the large scale development of what we call "renewable energy technologies", started only during the first decades of the 20th century with hydropower plants. The last decades of the 20th century saw the extensive development of the technologies which go under the name of "new renewables". Among these, we can define
Solar Power Agriculture: A New Paradigm for Energy Production
237
as "solar direct" methods those which collect and exploit solar light: mainly photovoltaic (PV) and solar concentration. We may define as "solar indirect" methods those which exploit the effect of the solar light in heating the atmosphere or the ocean: wind power, small hydropower (SHP), wave, ocean currents, and others. There exist also renewable technologies which are not based on solar energy, i.e. geothermal and tidal which may play an important role in the future but which will not be considered in detail here as they can't be framed within the "agricultural" paradigm. The energy needs of humankind will vary in the future depending on population and on the human lifestyle. The present work does not attempt to address projections for the future growth of population, nor the need of equilibrating the energy use by different fraction of the population. However, considering the slowdown of population growth observed during the past decades, most authors indicate that a stabilization of the world population might take place before reaching mid 21st century for numbers not much higher than the present ones. Considering also the ample possibilities for rich countries to reduce the amount of energy utilized by measures of efficiency and conservation, an "order of magnitude" estimation of the energy needed for humankind can be obtained examining the amount of energy used nowadays. The parameter to be examined here is the "total final consumption" (TFC). This is a better parameter to consider in the present context than the other common one, "Total Primary Energy Supply" (TPES), since renewables normally provide immediately exploitable energy in the form of electric power. The world TFC is reported to be 8 . 4 ~ 1 0 ~ G W y e a r [20] of which electric power is about lx107 G W y e a r . The OECD countries have access to 53.5% of the world TFC [20] for a population of 17% of the world total. The amount of solar energy that reaches the surface of the earth is commonly given as 1x1012 GWWyear (see. e.g. ref. 16): a value more than ten thousand times larger that the present TFC. In addition, we can roughly estimate the wind energy generated by solar irradiation as ca. 2% of the total solar energy arriving on the earth surface [ 161 that is about 2x10'' GWhJyear, again very large in comparison to the target. The energy associated with the global water flow generated by thermal effects is very difficult to estimate, but it is another very large amount. These data are listed in Table 1, together with two more data for comparison: the amount of the total biomass produced every year on the planet, some 2x10" tondyear, or lx109 G W y e a r [21] and the total human metabolic requirement [2 11. Table 1. Comparison of energy flows and energy requirements.
Human total metabolic energy requirement (present) World totai electric power generation (present) World total Anal consumDtion. T K Inresent) World total biomass produced Solar energy arriving on the earth's land surface I Total solar enerov arrivino on the earth's surface
5x106 1x10~ sx1a7
11~10~
I
3x1Ol1
I
I l Y l n ~ ~I
238
U. Bardi
The actual possibility of exploiting the large amount of solar energy arriving on the earth surface depends on the efficiency of the technology used. Furthermore, we have to consider the quality of the energy produced by solar technologies and not just its amount. Passive solar collectors, for instance, are not suitable for power production. Most renewable technologies (e.g. photovoltaics and wind turbines) generate high quality energy in the form of electric power and can therefore be used for the direct replacement of fossil fuels in power production. It is also always possible to use high grade energy to efficiently produce low grade energy, such as domestic heating. Regarding vehicle fuels, at present hydrogen, which can produced from electricity and water, appears to be a possible choice as replacement for fossil fuels. A different situation exists with biomass, which can be used to fuel a thermal engine in order to generate electricity or simply burned to obtain heat. In practice, for the purpose of comparison, in the present discussion it can be simply assumed that the energy output of renewables in the form of electric power can be used to replace all the forms of energy that compose the world TFC of today. This approach has been used, among others by Pimentel [ 151. This said, we can calculate the fraction of land area (or “footprint”) needed to generate the amount of energy needed by humankind by renewable technologies. In general, solar irradiation on populated lands ranges from a minimum of ca. 900 kWh/m2/year (e.g. northern Europe) to values of the order of 2200 kWhIm2lyear in subtropical regions, with even higher values in equatorial areas such as the Sahara. An approximate worldwide average value for populated areas could be taken as ca. 1500 kWWm2Iyear. The OECD countries tend to be located somewhat farther away from the equator, so that an average value in this case should be lower. Commercial photovoltaic panels have an efficiency that today can routinely reach 14%. Experimental PV cells may do significantly better. The efficiency of the power delivered to end users by this kind of plants can be reasonably considered as 10%. In terms of land use, with such efficiencies, solar direct plants can bring to the end user about 120 kWh/m2/year in medium latitude areas (1200 kWh/m2/year). As an order of magnitude estimation, it is possible to consider an average value of 100 kWh/m*/year for the power delivered to end users. In the case of solar indirect technologies, it is not possible to define a land requirement in the same way as for photovoltaics. Wind turbines must be spaced at a certain distance from each other, but most of the land occupied remains available for agricultural use. The physical footprint of a wind turbine may be estimated to correspond to a land use of about a factor 100 smaller than PV panels for the same delivered power. Of course, offshore wind turbines do not occupy any land area, and the same is true for SHP plants and for all the schemes which use wave power or oceanic currents. Finally, as a direct solar technology, biomass is relatively inefficient. The upper limit of photosynthesis efficiency at the molecular level is estimated as 6% by Tiezzi [21] and as 4.5% by Patzek [22]. In practice, the efficiency of biomass production in plant organisms is much lower. From the data of table 1, it can be calculated that the “planetary efficiency” for biomass production is about 0.1%. In specific cases, the efficiency can be higher. Patzek [22] reports a calculation on the average potato yield in England where the photosynthesis efficiency turns out to be
Solar Power Agriculture: A New Paradigm for Energy Production
239
0.4%. According to Tiezzi [21] crops may have yields of the order of 1%, but only at the cost of extensive use of irrigation and of fertilizers derived from fossil fuels. Another source [23] reports that some ecosystems, such as tropical forests, can reach efficiencies of conversion higher than 1% whereas crops cultivation has an average efficiency of ca. 0.3%. Taking into account the need of using a thermal engine to produce power, the sustainable efficiency of biomass as a source of electric power cannot be considered as higher than ca. 0.1%, again as an order of magnitude estimation. Obviously, the generation of electric power is not the best energetic use for biomass, but this value will be retained as a means of comparing biomass conversion with other energy sources. A further important parameter affecting land use is the energy return on energy invested (EROEI). This parameter can be defined as the total amount of output energy produced by the plant over its lifetime divided by the input energy, that is the energy required to build, maintain, fuel (if needed) and eventually dismantle the plant, as well as energy needed for vectoring the energy produced to end users (see e.g. ref. 15). The concept of EROEI can be expanded to include also the external costs (e.g. greenhouse gases produced) of the plant, measured again - in terms of energy. This parameter defines the “ecological footprint” (EF) of an energy producing plant, which is larger (and may be much larger) than the actual footprint of the plant. For instance, for fossil fueled energy plants, the EF is large because it is calculated taking into account the area of forest needed to remove the amount of C02 generated in the combustion of the fuels [24]. On the basis of the concept of EROEI, the Ecological Footprint (EF) of a renewable technology can expressed in units of area needed to provide a certain yearly output of energy, taking into account the area needed for the input energy, obtained using the same technology and averaged for the same period. Simple considerations lead to express the EF as a function of the actual footprint (AF) as EF=AF x EROEI/(EROEI-1). The EROEI is a difficult parameter to calculate and the estimates reported in the literature are widely variable. Some significant values for the technologies considered here are listed in the following table. For comparison, a giant oil field in its prime is reported to have an EROEI of ca. 30-50, whereas the EROEI for partially depleted and/or smaller oil fields may be around 10 or lower [25]. ~
Table 2. Comparison of different renewable technologies in terms of energy return for energy invested (EROEI) and ratio of “ecological foot print” (EF) to “actual foot print” (AF)
The values of Table 2 are to be considered as estimations only, however they do indicate that for the “new renewables” (mainly PVand wind) the EF is of the order of 10% or less larger than AF and therefore can be considered to lie within the uncertainty of the estimation of the needed land area. For biomass, the area correction for a value EF/AF ratio of 1.5 cannot be neglected, but it must also be taken into account that the EROEI
U. Bardi
240
calculated by Pimentel [ 151 includes the loss due to the transformation of heat to electric power. Therefore, the approximate value of "0.1%' calculated before can be retained since it already took into account this inefficiency factor. It is to be stressed that this biomass is intended to be mainly harvested from natural forests. Crops or other agricultural products make little sense as sustainable energy sources because of the extensive need for fertilizers, irrigation, etc which makes their EROEI close to one or even lower than one as reported by Pimentel [29] for bio-ethanol. The last important parameter to be considered here is the need of energy storage for intermittent sources such as photovoltaic and wind power plants. It is known that when the amount of energy produced by intermittent sources exceeds ca. 20% of the total grid power, it becomes difficult to control the energy input to the grid without some kind of energy storage. There exist a large number of methods for electrical energy storage, several still in the development stage (e.g. hydrogen coupled with fuel cells) others are possible only as large scale solution (e.g. pumped storage). For the present land area estimation, it will suffice to note that storage does not involve large areas and so it does not need to be considered here. Once these data are available, the simplest approach is to examine the land use for each kind of technology and from that calculate the extent of planetary land needed to generate, for instance, an amount of energy in the form of electric power equal to the present TFC. The following table summarizes the considerations developed so far. Table 3. Comparison of different renewable technologies in terms of e fficiency of conversion and area occupied. Method
Organic biomass
Approx. efficiency of light of conversion (ratio incident solar eneigy to electric power delivered to users)
-
0.10/0
Solar direct (PV or solar thermoelectric)
5%-100/0
Solar indirect (wind)
n.a.
Order of magnitude deliverable powel pel area occupied [footprint) for an iiradiation of 1200 kWh/rn'/year. (kWh/m'/year)
-1
-
-
100 10000 (on suitable
sites only)
At this point, we may proceed with embedding the data found so far within the concept of agriculture. That means to estimate what fraction of the land is needed to generate energy comparable to the present day use. The relevant numbers here are listed in Table 4 (data from refs. 17 and 29). Table 4. Land areas of the world presently used for different purposes.
Millions square km . Land "potentially suitable for agriculture"
130
Land used for food production
50
Arable land (crops production)
1.5
Solar Power Agriculture: A New Paradigm for Energy Production
24 1
There is a certain degree of uncertainty in these data [18], but the values reported here can be considered as prudent estimates. The concept of “land used for food production” is the sum of the arable land plus the grassland used for pasture. It is reasonably safe to assume in both cases the land is accessible by human beings and equipment and that it exists in human-compatible conditions of climate and temperature. Hence this area is also suitable for setting up energy producing plants. From these data, it is possible to estimate the extents of land required for two targets taken as the energy equivalent to the present value of world electric power generation (target 1) and equivalent to the present value of world TFC. The results are listed in Table 5 . Table 5 . Land areas needed for energy production using different renewable technologies
Land area needed as percentage of land used for food production (50~10~ Km2) Technology generation
BIOMASS conversion to electric power SOLAR DIRECT conversion (PV or solar concentration)
0.2%
SOLAR INDIRECT conversion (wind)
0.0030/0
1.5% I
0.03%
The values reported in Table 5 are, obviously,just orders of magnitude and are also to be considered as conservative. The fraction of land needed for the stated targets could become smaller for technological improvements and also taking account that a considerable fraction of the plants could be placed in areas which are not “agricultural land”, for instance offshore or on the roofs of existing buildings. The data, anyway, lead to the conclusion that both solar direct and solar indirect plants embedded in agricultural land would have a very small impact on agricultural production even for the relatively ambitious targets outlined above. Indeed, the human metabolic requirement of energy is only 5x106 GWWyear and with the known efficiency of the best crops it could be managed in just a few percent of the world land area. The impact of energy plants on agriculture would be, therefore, smaller than that of phenomena such as erosion and soil degradation. It is also worth comparing the surface needed for energy production with that of urban areas. Here, the data available are very uncertain. Kitajma [30] estimates the fraction of “urban land” as ca. 2% whereas Goldsmith [31] reports data which indicate that the fraction of land covered with “permanent structures” may be as high as ca. 15%. The large discrepancy among these values is probably due to differences in the definition of “urban land”. However, Kitajima’s estimation, compared with the data of Table 3,
242
U. Bardi
indicates that at the land area needed to provide a town with sufficient energy is of the same order of magnitude as the urban area. In other words, a town should dedicate a land area of extent approximately equal to the area of the town itself to the production of energy. Micro-production of energy at the level of single roofs is a possible solution, but a much easier approach is to disperse this production over the agricultural land surrounding the town. There exist very large urban spreads in the world where this approach may be difficult, but in most OECD countries towns are surrounded by agricultural areas much larger than the towns themselves. It may also be remarked that it may be possible to make even solar direct technologies compatible with organic agriculture occupying the same areas. Right now, the shadow of a solar plant (photovoltaic or thermoelectric) makes it impossible to cultivate the area Underneath. However, photosynthesis uses only a fraction of the solar light spectrum and solar direct plants could be modified to let energy pass through in the right wavelengths only. The mirrors of a solar concentrating plant, for instance, could be coated in such a way to reflect only the part the spectrum not used by photosynthesis, letting instead pass the rest. The efficiency of the solar plant would be lowered, but it would be possible to cultivate the area underneath. This technology is probably unnecessary at present because of the small fraction of area needed by solar plants, but it is mentioned here as a possibility for the future.
3. costs Estimations of the costs of renewable energy are common in the literature, a review can be found, for instance, in [8]. In most of the studies published so far, the calculations arrive to values that favor fossil fueled plants over renewables. In general, electric power generated by wind turbines is only marginally more expensive than that generated by fossil fuels. For photovoltaic panels, the gap is large, of approximately a factor 5 (see, for instance, ref. 10). The “solar power agriculture” paradigm offers a way to look at the financial support for renewables which goes beyond the cost calculations reported in the literature. The misleading factor of these calculations lies in the fact of producing a single, static number. However, the cost of energy is dynamic and it varies over the whole lifetime of a plant which, for the case of renewable plants may be very long, typically of a few decades. We saw in the previous section that the “new renewables” produce a positive, and often highly positive, energy return (EROEI). Since energy is worth money, there follows that the financial return from these technologies will be positive in the long run. It may argued on whether other investments will produce a quicker return, but not on the fact that the money and resources placed into the deployment of renewable energy plants will be, eventually, paid back and will subsequently produce a profit. A case could be made that renewables are not only an effective investment but also the safest possible investment since they produce a commodity, energy, which is the basis of human civilization and hence will never suffer market crashes. However, it doesn’t appear that investors have picked up yet the potential of renewable energy. This may be due to the bad image of renewables which still lingers in
Solar Power Agriculture: A New Paradigm for Energy Production
243
the press and in the public opinion. In this respect, it may be worth citing here as an example an article by J. Dvorak [32]: “People still equate photovoltaics with the hippydippy 1970s. A recent article in the New York Times showed some hippies in Ukiah, California, with a solar panel outside their yurt-like home with a sod-covered roof. Presumably, the solar panel was used to power a lava lamp.” Apparently, the promotion of renewable energy done so far has been off target with the emphasis made on the concept that renewable energy is something “soft” as opposed to fossil fuels which, somehow, are “hard”. Switching to “soft” energy, apparently, would change our way of living, maybe carrying all of us to the Aquarian Age, to a higher state of consciousness or something like that. It isn’t clear how the kWh at a domestic outlet can be defined “hard” or “soft” and anyway these concepts don’t go well with investors who are more likely to think in terms of financial returns rather than in terms of holistic happiness. However, the economic return of renewables looks different when placed within a context, agriculture, which is known to be reliable, solid, and effective. Maybe agriculture never provided such a rapid return on investments as startup companies from Silicon Valley, but everybody understands that we can’t eat silicon. So, there is a logic in investing in agriculture and there are people interested in investments which privilege safety over speed and which also have a certain degree of “ethical” value. Here is where the “agriculture paradigm” as opposed to the “industrial paradigm” can be most useful. As long as energy production is managed by power utilities, as it is the case today, energy is produced within the “industrial paradigm” since power utilities are industrial companies which have a tradition of generating energy from the combustion of fossil fuels. In the industrial paradigm, the capital is provided by shareholders and companies will normally choose the road that leads to the fastest return on investments. In “solar power agriculture”, however, the situation is different and the focus moves from capital to land. The question that a stock market investor asks is “how can I increase the value of the capital I own?” while the question that a farmer or agricultural operator asks is “how can I increase the value of the land I own?” The answer to the latter question is to produce something on the land: it may be crops, it may be wine, it maybe wood, it may also be electric power in the paradigm if solar power agriculture. Large financial resources could be unlocked and utilized for the diffusion of renewable energy as soon as the concept of solar power agriculture becomes an accepted paradigm. For instance, the budget of the European Union for “rural development” in the period 2000-2006 is of approximately 50 billion euros. It is worth noting that in the 2003 document describing the EU agricultural policies [33], the world “energy” is not even mentioned, although there is one picture showing 19‘hcentury windmills. If governments realize that agriculture can be supported in an effective manner by means of the concept of solar power agriculture, part of these resources can be utilized for the diffusion of solar energy in farmlands; this benefits not only farmers, but all the sectors of society. It remains to consider how fast renewable energy could develop in order to cover the targets outlined here. This depends on the amount of financial resources that can be reasonably allocated to renewables and that may be estimated from the historical record. During the 1990-2000 decade wind power grew at an average rate of 22.4% per year and
244
U. Bardi
photovoltaic of 28.9% in the countries belonging to the OECD (Organization for Economical Cooperation and Development) [81. According to the BP statistical energy review [34] the growth of wind power worldwide was of 29% from 2001 to 2002, with the wind share of total global electricity supply increasing four-fold since 1996, to 0.4% in 2002. In those countries where wind technology can be considered as “mature”, the data [35] growth rates higher than 20% have been observed over at least one decade as for instance in Denmark (22.0% yearly) and in Germany (62.9% yearly). As a comparison, the highest rate of growth of crude oil production was 7% yearly for the period from 1930 to 1971 [I]. At a yearly growth rate of 20% and starting with 0.5% of the present supply, renewables could grow to produce the equivalent of the electricity produced today by fossil fuels in 30 years. This is, obviously, just an order of magnitude estimation. However it shows that renewables have a considerable potential to replace fossil fuels. 4. Public Acceptance and Environmental Concerns
Opinion polls show that most people are, theoretically, favorable to renewable energy (e.g. 36). However, when it is question of turning ideas into practice and actually build plants, the public opinion may not remain as favorable as opinion polls claim it to be. Basically, people object to renewable energies for two reasons: I. Renewable energies are seen as ineffective, that is they are not taken seriously a source of power. This concept leads to the related opinion that money and resources should be better spent in energy conservation measures. Renewable energies are seen as polluting, not in the same way as fossil fuels or 2. nuclear plants, but still causing visual pollution, damaging wildlife, producing noise, etc. These two arguments are closely linked to each other. Surely, people would be much more willing to accept the noise and the visual pollution of renewable plants if they were convinced that they are a real option to get rid of fossil fuels. Evidently, this is not the case since in recent times a remarkable backlash has been observed against wind power (e.g. ref. 37). The negative public attitude is not limited to wind turbines only. Mini-hydro plants are also accused to kill fish [8] and photovoltaic power is accused of using up too much land (see e.g. ref. 38). The emergence of such negative attitudes towards renewable plants is often related to policy mistakes, in this case in the way the new plants are proposed. Local communities often resent being invaded by power utilities with their gigantic wind towers. People feel (in most cases, correctly) that the value of their property will be reduced by the presence of large and visible energy plants in the vicinity. In these conditions, local communities cannot be blamed for opposing renewable plants. Here, the paradigm of “solar power agriculture” can reduce public opposition to renewable energies. By linking the new renewables with a well known and accepted technology such as agriculture, the gain in image is impressive. Solar power technologies, such as wind turbines and PV panels, cease to be seen as toys for hippies or bird-killing machines but become part of a respected and reliable production system which has fed
Solar Power Agriculture: A Mew Paradigm for Energy Production
245
mankind for some ten thousand years and is still feeding 6 billion and a half people on the earth. If, as it is the case, solar power agriculture is just another form of conventional agriculture, then it shares with it the same sturdiness, reliability, and social value. At the same time, the visual impact of the new renewables is considerably reduced in relative terms when it is looked at in the context of the impact of traditional agriculture. It may be argued that no activity of mankind on earth has had a more important impact than agriculture. Indeed, with about 40% of the land area used for food production, and a comparable fraction used for fiber and biomass production, we can say that traditional agriculture didn’t just have an impact on the environment; it radically changed the face of the planet. In many cases, actually, agriculture has been devastating for the environment, leading to irreversible salinization of desertification of large areas [7]. In comparison, turning less than 1% of the land from conventional agriculture andor forestry to solar power agriculture represents no major ecological impact. 5. Conclusion and Perspectives for the Future
The paradigm of “solar power agriculture” or “agri-energy” outlined here falls within some well known and accepted paradigms of agriculture and it goes one step beyond the idea that renewables are useful for supplying isolated areas with energy. Here, it is the opposite case: the existing power lines can be used for carrying the rurally produced electricity to towns and industrial centers at little or no additional costs. Carrying their products to town is what the rural world has been doing during the past 10,000 years or so. The concept of “solar power agriculture” permits to approach the problem of introducing renewable energy in the world supply in a completely different way than the presently accepted industrial paradigm. The use of renewables for isolated or remote areas is probably reaching saturation whereas the idea of placing large solar plants in deserted areas has not materialized yet for the high investments needed and for the lack of a suitable infrastructure for transporting the energy produced. However, once electricity (or other forms of energy) are placed in the same range of products as agricultural products, we immediately set a path of growth which is gradual, economically sound, requires low investments, and which can be put into practice in OECD countries without revolutionary changes. The present work has calculated that the fraction of land needed for present-day renewable technology to provide the world with an amount of energy comparable to the energy produced by fossil fuels would have only a minimal impact on food and textile agricultural production. The experience of the past decade shows that renewable energy production can grow at a yearly rate higher than 20%, much larger than the present trends of increase in energy production (around 2% year). If these trends could be maintained, even ambitious targets such as the complete replacement of fossil fuels by renewables are not impossible in less than one century.
246
U. Bardi
References [ 11 Association for the Study of Peak Oil (ASPO) 2004, www.oilpeak.net. [ 2 ] Bentley RW. (2002) Global oil & gas depletion: An overview. Energy Policy 30 189-
205. [3] Deffeyes KS. (2001) Hubbert's Peak, the impending world oil shortage, Princeton. 141 Campbell CJ, Laherrere JH. (1998) The end of cheap oil. ScientlJic American March: 60-65. [5] Hubbert MK. (1962) Nat. Res. Council Publ. 1000-D Washington DC, 54. [6] La1 R. (2003) Global potential of soil carbon sequestration to mitigate the greenhouse effect. Environment International 29 (4): 437-450. [7] Ponting C. (1991) A green history of the world, New York. [8] International Energy Agency (IEA) (2003) Report on renewables, www.iea.org. [9] Muneer TA, Kubie M. (2003) Generation and transmission prospects for solar electricity: UK and global markets. J. Energy Conversion and Management 44 (1): 35-52. [lo] Bal L, Chabot BC. (2001) Les knergies renouvelables. Etat de I'art et perspectives de developpement. R. Acad Sci. Paris, Sciences de la terre et des planetes 333: 827-834. [ l l ] Pearce JM.(2002) Photovoltaics-A path to sustainable htures. Futures, 34 (7): 663-674. [ 121 Schlaich J. (1999) Tension structures for solar electricity generation. J. Engineering Structures 21 (8): 658-668. [ 131 Trieb F, Langniss 0, Klaiss H. (1997) Solar electricity generation--A comparative view of technologies, costs and environmental impact. Solar Energy 59 (1-3): 8999. [14] Pimentel D, Patzek TW. (2005) Ethanol production using corn, switchgrass, and wood; biodiesel production using soybean and sunflower. Natural Resources Research 14 (1): 65-76. [15] Turrini, E. (1991) La Via del Sole, Edizioni Cultura della pace, San Domenico di Fiesole (Fi). [16] Hoogwijk M, Faaij A, Van den Brock R, Berndes G, Gielen D, Turkenburg W. (2003) Exploration of the ranges of the global potential of biomass for energy. Biomass and Bioenergy 29 (4): 225-257. [17] Wolf J, Bindraban PS, Luijten JC, Vleeshouwers LM. (2003) Exploratory study on the land area required for global food supply and the potential global production of bioenergy. Agricultural Systems 76 (3) 84 1-861. [ 181 F A 0 (2004) www.fao.org, Statistical Database. [ 191 International Energy Agency (IEA) (2001) www.iea.org. Statistics on World Energy. [20] Tiezzi E. (2001) Tempi Storici, Tempi Biologici, Venthnni Dopo. Donzelli Ed. Roma.
Solar Power Agriculture: A New Paradigmfor Energy Production
247
[21] Pazek TW. (1997) http:/lpatzek.berkeley.edu/Ell/Photosynthesis.pdf, based on Energy, Plants and Man by David Walker, OxyGraphics, England, Second Edition, 1993. [22] Kling GW. University of Michigan, http://www.globalchange.umich.edu/globalchange1/current/lectures/kling/energyfl ow/energyflow.html (accessed in 2004). [23] Haberl H,Wackernagel M, Wrbka T. (2004) Land use and sustainability indicators. An introduction. Land Use Policy 21 (3): 193-198. [24] Swenson FU3. (2000) Solar energy meets the new global challenge. http://www.ecotopia.com/apollo2/ises2OOO.htmISES Millennium Solar Forum, Mexico City. [25] Shapouri H,Duffield, JA, Graboski S. (1995) Estimating the net energy balance of corn ethanol. U.S. Department of Agriculture Economic Research Service, Office of Energy. Agricultural Economic Report No. 72 1. [26] Knapp KE,Jester TL. (2000) Proceedings of the Solar 2000 Conference, Madison, Wisconsin. [27] White SW, Kulcinsky GL. (2000) Birth to death analysis of the energy payback ratio and C 0 2gas emission rates from coal, fission, wind and DT-fusion electrical power plants. Fusion Engineering and Design 48 (3-4): 473-48 1. [28] Pimentel D, Eisenfeld HS, Olander L, Carroquino M, Corson C, McDade J, Chung Y, Cannon W, Roberts J. (1994) Achieving a secure energy future: environmental and economic issues. Ecological Economics 9 (3): 201-219. [29] Kitajima Y. (1997) www.formal.stanford.edu/jmc/nature/nodel 1.html (accessed January 2003) [30] Goldsmith E.(1972) Blueprint for Survival, London. [31] Dvorak JC. (2001) Sunshine in the forecast. Forbes Global Magazine 05.14. [32] Rural development in the European Union, Fact Sheet 2003, http://europa.eu.inticommlagriculture/rurlpublilindex_en. htm. [33] BP energy report. 2003, www.bp.com. [34] 2003 World Wind Energy Association (WWEA), http://www.wwindea.org/pdf/WWEA~InstallationFigures~World2003 .pdf. [35] Farhar BC. (1996) Energy and the environment: The public view. Renewable Energy Policy Project, College Park, Md, October. [36] Heberling M. (2002) Environment news, The Heartland Institute, editor. [37] Palazzetti M, Pallante M. (1997) L’uso razionale dell’energia, Bollati Boringhieri ed.
This page intentionally left blank
POSSIBILITIY ASSESSMENT OF WIND ENERGY UTILIZATION IN BOSNIA AND HERZEGOVINA F. BEGIC (l), A. KAZAGIC (l), N.H. AFGAN (*) (I)
JP Elektroprivreda BiH, Sarajevo, Bosnia and Herzegovina, Lisbon, Portugal
(2)
Instituto Superior Tecnico,
The limited reserves of fossil fuels and the need for environmentally friendly electricity generation, requires investigation of possibilities for the utilization of renewable resources in Bosnia and Herzegovina. One of the possible renewable sources is wind energy. The research of potential utilization of wind energy in Bosnia and Herzegovina presently is being carried out. As a part of the research, some locations are investigated using knowledge of atmosphere physics and meteorological conditions for given locations. The research is aimed to create basis for possible utilization of wind energy for electricity generation in Bosnia and Herzegovina, through the assessment of meteorological, technical, technological, economical, social, and environmental aspects of wind energy utilization at selected locations. The research comprises the following: Specification of decentralized wind energy systems - potential share and demand; Data collection of the direction, frequency and velocity of wind at the selected macro locations: in Sarajevo area (Mountain Bjelasnica and settlement Butmir) and in Mostar area (Mountain Velez); Selecting the potential micro-locations for installing the wind turbines field; Selection of the type and construction of the wind turbines; Sustainability assessment of electricity generation from wind turbines on the base of precisely defined criteria and sustainability indicators - Multi-criteria Sustainability Assessment; Optimization of the wind energy systems. The investigation is especially interesting for similar - prevailing mountain regions, and could answer whether utilization of wind energy for electricity generation in such regions is feasible. In the paper, investigation on wind characteristics in Sarajevo area is presented, as well as sustainability assessment of wind energy utilization being used within the research.
1. Introduction In the past several years, a continuous growing of the electricity consumption and generation in Bosnia and Herzegovina can be noticed. Among the current process of reconstruction of the electric power system in Bosnia and Herzegovina, common for countries in transition, refurbishment of the existing power plants, including investment in environment protection, is a process that is expected to continue. In the last five years, thanks to intensive efforts in this direction, electricity supply has been improved. However, assessments of electricity consumption in domain of JP Elektroprivreda BiH have shown that the growing of electricity consumption in the next period (till year 2020) should continue. Consequently, extending of energy power system by additional 100 MW of installed capacity is expected to be realised in close future. Due to unsatisfied the overall environmental issue, preferably high carbon dioxide emission, extending of energy system by renewable resources is optioned as well. In that sense, research on possibilities of utilisation of wind energy in Bosnia and Herzegovina, that has currently being carried out, is of fundamental significance [ 1,2,3,4].
2. Potential for Wind Energy Utilisation in Bosnia and Herzegovina On the base of data on wind characteristics from The Hydro and Weather Forecast Institute of Bosnia and Herzegovina, preliminary data base on wind frequencies and velocities for selected regions is created. A part of the results for Sarajevo area is 249
250
F. Begic, A. Kazagic & N.H. Afgan
presented here. Charts in Figs. 1 and 2 show frequencies and velocities of wind in respective directions for locations of mountain Bjelasnica and settlement Butmir, [2,3]. It can be seen that the average wind velocity for Butmir is about 2 m / s , so this macro location is rejected from further evaluation. However, the wind velocity for area of mountain Bjelasnica is potentially very convenient for installing the wind turbines, especially for directions NE-SW and N-S.
Fig. 1. Wind frequencies in respective directions in regions of Bjelasnica and Butmir sites.
Fig. 2. Wind velocities in respective directions for regions of mountain Bjelasnica and settlement Butmir.
The wind velocity at different altitudes can be estimated by relation (1):
Wind Energy Utilization in Bosnia and Hevzegovina
25 1
where: v wind velocity in Bjelasnica site vo wind velocity in Butmir site H altitude of top of Bjelasnica Ho altitude of settlement Butmir a average value of 4 prevailing directions Results of this calculation are given in Fig. 3.
I
I
500
7W
900
11W
13w
1500
17W
1900
21W
Atilude, m
I
Fig. 3. Estimated wind velocities for different altitudes.
On the base of presented data, two micro locations are selected for possible installing the wind turbines in Sarajevo area; Top of Mountain Bjelasnica and Lisicija Glava, see Table 1. Table 1. Selected locations for installing the wind turbines.
Average velocity Altitude Maximal power coefficient Rotor diameter Power Wind turbo-generator power
1. Top of Bjelasnica 12 m i s 2067 m A S . 0,4 40 m 451 kW 600 kW
2. Lisicija Glava 10 m i s 1762 m A.S. 0,4 40 m 451 kW 600 kW
3. Sustainability Assessment of the Wind Energy Utilisation Multicriteria sustainability assessment based on ASPID methodology is used for sustainability evaluation of wind energy utilisation in Bosnia and Herzegovina. The point of the used method is in synthesis of specific criteria which describe various aspects of the objects (in this case energy systems of various sources) on the base of initial attributes or indicators, and which are being evaluated by General index characterising all options under consideration. This basic method-MGI (Method General Index) is adopted by ASPID methodology (Analysis and Synthesis of Index at Information Dejciency) [ 5 , 61.
F. Begic, A. Kazagic h N.H. Afgan
252
The aspects considered in the analysis include consumption of resources, environmental issue, social aspect and economic aspect for all options under consideration. Option of wind turbines is compared against other seven options of energy systems for electricity generation of various sources. In this consideration, it is supposed that 7875 GWh of electricity is to be provided in projected lifetime of 15 years for all options under consideration. In calculation of sustainability indicators for wind turbines, in order to get realistic and reliable results, it is assuemd that materials for the wind turbines construction are produced by the electricity generated in a 100 MW pulvurised coal based power station an option under consideration that has a low efficiency (heat consumption of 10673 kJ/kWh or efficiency of 33%) and strong environment impacts. It is calculated that 438 units of 600 kW wind turbines are to be installed to provide aimed electricity generation of 7875 GWh over period of 15 years. Basic data related to wind energy utilisation from this analysis are given in Table 2. At the same time, data from this table presents input data for calculation of sustainability indicators of wind based electricity generation. Table 2. Data on wind energy utilization based on required electricity production.
Required electricily production
7 875 GWh
Life time
15 years
Annual Electricity production
525 G W a
Annual operation time
2000 Wa
Total Power installed
262 500 kW
Power installedper unit
600 kW
Number of units
438
Used area
1000 m2IGWh
Zauzetost zemljiita
7,9 x
Efficiency
25 %
Assumed energy costs
8 cent/kWh
Investment costs
1050 EURlkW
0 & M costs Total 0 & M costs
1,4 cent/kWh
lo5mz/kW
7 350 000 E m g o d .
3.1. Sustainability Indicators of Wind Based Electricity Generation A sub-stoichiometric quantity of oxygen, or air, is used in the partial oxidation fuel reaction. This small amount of oxidant produces some heat and water that reacts with the non-oxidized fuel to generate hydrogen and carbon dioxide, considering theoretical approach showed in Eq. (3). However, practical systems, based on partial oxidation technology, generate syngas with less hydrogen than other methods such as autothermal or steam-reforming. A large amount of carbon monoxide is yielded. The process temperature is also higher, around 1000°C.
Wind Energy Utilization in Bosnia and Herzegovina
253
3.1.I . Resources Indicators (H) Materials needed for construction of a 600 kW wind turbine are given in Table 3. Table 3. Materials for wind turbine construction.
456 134 12 3 3 <2
Concreei Steel GRP Cuprum Gravel Aluminium Oils PVC Other Total
<1 41 <<1
612
14 22 2
<1 4 <1 <<1 100
Indicators of three dominant materials are given in Table 4. Table 4. Resources indicators for wind turbines
R1.r NC”
RIAI
Carbon steel indicator Cuprum indicator Aluminium indicator
The most important resource indicator is fuel indicator. As pointed out above, fuel consumption is considered for this option due to production of materials for wind turbines construction. Calculated value of this indicator is given in Table 5. 3. I .2. Environmental Indicators (EI)
Although wind energy is legitimately held as a clean energy, electricity production based on wind energy still brings by itself certain environment impacts, chiefly expressed through environment impacts of grave energy - energy used for production of materials for wind turbines construction. As it is assumed that electricity used for materials production is generated in PC based power station, pollutant emissions generated by combustion of corresponding amount of coal are calculated, and devided by electricity production in lifetime. The results are given in Table 5 . 3.1.3. Social Indicators (SI)
Job indicator is defined as number of paid hours devided by production of electricity in lifetime, determining engagement of employees for all options under consideration. 3.1.4. Economic Indicators (EcI)
Economic indicator of energy costs (costs of electricity generation) is defined as sum of O&M costs, investment and expense to local community, divided by electricity
F. Begic, A . Kazagic & N.H. Afgan
254
production in lifetime. Near this, economic indicator of investment and economic indicator of efficiency are defined and included in this analysis. Sustainability indicators of wind based electricity generation in B&H are summarised on the base of a.m. procedure and given in Table 5 . Unexpected high values of Resources Indicators and Environment Indicators can be noticed. It can be explained by the fact that a huge amount of materials is to be produced to manufacture and install such an wide wind turbines field. On the other side, economic indicators are being improved with reducing the investment and increasing of efficiency (improving of wind turbine technology). Favourable social indicators for wind turbines in this case is due to installing number of single units, and relating people engagement in production of wind turbines, and in operation and maintenance over lifetime of 15 years. Table 5. Sustainability indicators for option of wind turbines.
Indikator
Indikator
Indikator NOz
R L i
RIcoz
Hs02
kgIkWh
g/kWh
g/kWh
g/kWh
0,1016
119,3
2,032
0,1172
so2
I
Energy costs Indicator
Investment Indicator
Efficiency Indicator
Me e.
RImv
m f f
Mpb
lkWh
WkWh
31,7
0,0010
"Ox
centEUR/kWh centEUR/kWh
3s
5,8
Job indicator
3.2. Review of the Sustainability Assessment of Wind Based Electricity Generation Sustainability assessment is performed including various options of the energy power systems in Bosnia and Herzegovina: 0 Option 1: Pulverised coal based unit in condensing regime - Reconstruction; 0 Option 2 : Pulverised coal based unit in cogeneration regime - Reconstruction; 0 Option 3 : Fluidized bed combustion based unit; 0 Option 4: Combined cycle gas turbine power station; 0 Option 5: Hydro Power Plant - Reconstruction; 0 Option 6: Power plants based on solar energy (PV systems); Option 7: Wind turbines; 0 Option 8: Biomass power plants. A part of the results of the sustainability assessment respecting advance of utilization of wind energy is presented here. The case when social indicators are simply compared for options under consideration within single criteria assessment is given in Fig. 4.
I
I
Social Indicator 0 0012
~
ff =a
O.O0O8
0.0004 0,0002
I
0
1
2
3
4
5 ODtian
6
7
8
I
Fig. 4. Single criteria assessment according to the social indicator.
I
255
Wind Energy Utilization in Bosnia and Herzegovina
It can be noticed that according to the social indicator option of wind turbines is a preferable option. If an advantage to the social aspect within a multicriteria sustainability assessment is given, with taking into account the other aspects of sustainability, option of wind turbines is still the preferable option, Fig. 5.
Fig. 5. Multicritena sustainability assessment - Case SI > RI = ECII= EII
However, if larger weighting coefficients are given to the economic indicators, multicriteria sustainability assessment shows that wind turbines are in a group of unfavorable options, Fig. 6.
Fig. 6 . Multicriteria sustainability assessment-case E d l > RI = EII = SI.
On the other side, allocating advantage to the environment aspect, or resources aspect, option of wind turbines comes back among preferable options, Fig. 7. Preference of the option of wind turbines at giving advantage to the resources and environment aspects would be more expressed if it is assumed that materials for wind turbines construction is produced on the base of electricity generated by some clean power technology. ’
Object
OW
~
5
Opcija 5 RHE Opcija8PB
7
OpcijalVE
~ 8 -~
-
~
6 OpcijaBSP -
4
Opcija 4 NKGPPK ~
1 Opcija 1- RTPUK 2
Opcija 2 RTPUKO
3
Opcila3NTPUFSKO
)010
l020
-
1030
-
(040
1050
-
[OW
1070
lO00
+
. *
1
Fig. 7. Multicriteria sustainability assessment - Case EII > RI = Ec11 = SI
-
llCnl
256
F. Begic, A . Kazagic & N.H. Afgan
4. Conclusions
Preliminary analysis of wind characteristics on the base of the data from The Hydro and Weather Forecast Institute of Bosnia and Herzegovina has indicated a good potential for utilisation of wind energy on some macro locations in Bosnia and Herzegovina. Furthermore, some micro locations have been determined by analysis within Research on possibilities of utilisation of wind energy in B&H, which is being carried out at the moment. To evaluate sustainability of installing the wind turbines on selected locations, and compare it to the other energy systems of various sources within Energy Power System of B&H, multicriteria sustainability assessment on the base of ASPID methodology is made. The assessment comprises four aspects of sustainability: consumption of resources, environment issues, society effects and economic aspect, under which sustainability indicators are defined and calculated. Specific criteria are synthesized on the base of initial attributes or indicators, which are being evaluated by General Index using ASPID methodology (Analysis and Synthesis of Index at Information Dejciency). The results of the evaluation in the case when social aspect of sustainability is preferable show that option of wind turbines has the highest rate of sustainability among all options under consideration. Furthermore, when environment or fuel indicators are preferable compared to economic indicators, option of wind turbines is placed into group of high-sustainability options. Separately, in the case when economic indicators have advantage related to the other indicators, option of wind turbines is not preferable. In the light of the coming process of extending the Energy Power System of B&H by new power stations, installing the wind turbines on selected micro location is considered. Taking into account the current process of restructuring of Energy Power System of B&H, imaged to the countries in transition, and respecting strong social effects of this process within the energy sector, installing the wind turbines is intruded as one of the favourite options. Multicriteria sustainability assessment has proved it.
Nomenclature v vo H Ho
a
wind velocity in Bjelasnica site wind velocity in Butmir site altitude of top of Bjelasnica altitude of settlement Butmir average value of 4 prevailing directions
References [ 11 Begic F, Kazagic A. (2002) Sustainability development of energy system - basis of society development strategy (in Bosnian), 5th International Conference “Energetska i procesna postrojenja - Dubrovnik 2OO2”, Proceeding, Energetika Marketing, Zagreb, May. [2] Begic F, Kazagic A. (2002) Possibilities for utilization of renewable resources for electricity generation in Bosnia and Herzegovina, UNESCO sponzored conference on Sustainable development of energy, water and environmental systems Dubrovnik 2002, Proceeding, Energetika Marketing, Zagreb, June.
Wind Energy Utilization in Bosnia and Herzegovina
251
[3] Begic F, Kadic N, Kazagic A, Masala L. (2003) The role of renewable resources in sustainable development of energy power system of B&H (in Bosnian), VI International Conference of B&H Comity Cigre, Neum, September-October. [4] Begic F, Kazagic A, Tanovic F. (2003) Managing carbon - Requirement of liberalised electricity market, VI International Conference of B&H Comity Cigre, Neum, September-October. [ 5 ] Afgan NH. (2000) Multicriteria indicators for clean air technologies, UNESCO Chair holder, Instituto Superior Tecnico Lisbon, Estoril, Portugal. [6] Afgan NH, Carvalho MG. (2000) Sustainable assessment method for energy systems, Instituto Superior Tecnico Lisbon, Portugal.
This page intentionally left blank
EARTH TUBE VENTILATION SYSTEM: A PROJECT PRE-FEASIBILITY ANALYSIS TOOL M.A.A. COSTA, M.S.A. OLIVEIRA, N. MARTINS ('I Departamento de Engenharia MecLinica, Universidade de Aveiro, Campus Universitario de Santiago, 3810-193 Aveiro, Portugal,
[email protected] Increasing energy demand and rapid depletion of non-renewable resources have forced mankind to search for alternative energy resources. Over the recent decades, there has been a shift towards energy efficient renewable sources of energy as compared to conventional resources. Earth heat exchangers are advantageous features to reduce energy consumption in residential buildings. In winter they pre-heat ventilation air with minimal operation costs - necessary for low energy architecture -, in summer they help to prevent passive houses with relevant solar gains from overheating by pre-cooling ventilation air. The work here presented, reports on the development of an assessment tool for project pre-feasibility analysis of earth tube ventilation technology. The developed tool is suitable for the evaluation of energy production, life cycle costs and green house gas emissions assessment of a specific system to be built in a certain location.
1. Introduction
The use of earth tube ventilation systems (ETVS), Fig. 1, has gained a significant ground in heating and cooling of buildings [ l ] due to a considerable increase in energy consumption and inherent expenditure of buildings. Earth tube ventilation systems provide suitable heating and cooling energy loads of building based upon seasonally varying inlet temperature and wall temperatures which depend on the ground temperature. The literature is unanimous in reporting that a time lag occurs between the temperature fluctuations at the surface and in the ground. Therefore, at a sufficient depth, the ground temperature is lower than the outside temperature in summer and higher in winter. When ambient air is drawn through buried tubes, the air is cooled in summer and heated in winter, before it is used for ventilation [2-41. The literature abounds with studies involving earth tube ventilation systems. Studies where simplified models reporting a reasonable prediction of tube extracted temperature [5-91 have been presented. Others defend that in order to represent the actual phenomena involving earth tube heat exchangers a complex transient approach based on coupled and simultaneous transfer of heat and mass into the soil and the circulating air, is required [lo-171. A number of studies also report on the influence of specific ground covers on the thermal effectiveness of the soil [18-191. Despite the advances on earth tube ventilation system thermal model research and the numerous literature reports, the implementation of such systems, during building installation design, still remains absent. This is believed to be associated to the inexistence of pre-feasibility analysis tool at the engineering level, and consequent lack of thermal effectiveness assurance, cost analysis and therefore project feasibility. This paper reports on the advances in the development of the tool pre-feasibility analysis, where a user friendly interface associates a reliable energy model with financial indicators. The tool will enable a building installation designer in the process of decision making, since, in a matter of minutes, with minimum data input, can gather information in what concerns: tube characteristic dimensions, number of tubes, according to specific depths, investment and maintenance costs, energy gains and savings and finally project feasibility. 259
260
M.A.A. Costa, M.S.A. Oliveira & N. Martins
Fig. 1. Earth tube ventilation system.
2. ETVS - Pre-feasibility Tool
2.1. Tool Structure
The flow chart of the ETVS-Pre-feasibilitytool is illustrated in Fig. 2. The tool comprises four major modules and three databases defining by themselves a project viability analysis strategy. The modules are sequentially accessed and, at the end, the tool provides financial indicators for the analysed solution. From the examination of the determined financial indicators the user will be able to accept or reject a solution and rank alternative ones. Once a specific solution is selected, parametric studies can be performed in order to analyse its sensitivity to most volatile parameters, such as, future energy price or inflation rate. The starting module is the Energy Analysis Model, which has the primary aim of determining the energy production of a defined design in a specific location. This is the basic information required to determine the ETVS benefits, i.e., life-cycle positive cash flows. Additional benefits may be provided by green house gas (GHG) emission reduction, namely CO2. In fact, a secondary effect of the ETVS may be the reduction of GHG emissions to the extent that will depend on energy savings and replaced energy source, which correspond to a financial benefit that may be considered in the prefeasibility analysis. This benefit is determined in the GHG Module. The ETVS gains (energy savings and GHG reduction) can then be determined. The Cost Analysis Module will be used to estimate the system cost. Two types of costs are considered: the initial investment (materials, equipment and labour) and maintenance costs (spare parts and required energy consumption, e.g., ventilation fans). The information gathered in the previous modules, together with financial parameters such as discount rate and replaced energy price, will be used to determine life cycle financial indicators by the Financial Analysis Module. This will be the final step of the pre-feasibility analysis procedure which will provide quantitative data required to support the end-user decision.
261
Earth Tube Ventilation System
Energy Gain
Tax Savlngs (due to GHG emmisions
Energy price
!-----
Financial Summary
Cash Flom
Financial
parameters
Financial Indicators
Fig. 2. Flow chart of ETVS-pre-feasibility tool
2.2. Energy Model
The energy model was developed bearing in mind a situation of compromise between essential aspects reviewed in the literature regarding actual thermo-hydraulic design of earth tube ventilation systems, with computational time minimization and minimum user expert knowledge. Therefore, in this section it will be described how a simplified model (i.e. a one-dimensional transient approach) was developed, what assumptions where considered, the reasons which support those decisions and how the results compare with a complex three-dimensional transient approach (i.e. numerical simulation carried out with a commercial CFD benchmarked code). The computational three-dimensional domain which serves the latter is illustrated in Fig. 3.
.
.
.
.
.
.
. .
.
.
. . ~
Fig. 3. Computational domain.
M.A.A. Costa, M.S.A. Oliveira & N. Martins
262
2.3. Physical and Mathematical Models The fluid enters a buried tube at a certain depth, with a constant velocity and flows through the domain (i.e. soil is considered to be a continuous media with specific values in what concerns thermal conductivity, specific heat and density), leaving at the other extremity. It is assumed a transient, turbulent flow and a constant physical properties' fluid [O]. The thermal boundary conditions imposed at the surface ground and inlet air temperature were described by a sinusoidal function where seasonal thermal amplitude was taken into account:
where T,, is the average annual temperature, A, the thermal amplitude, At, the time step and P, the period of time under analysis. Ground analysis. The ground is considered to be at a given uniform temperature and its temperature evolution is evaluated using a one-dimensional explicit finitedifference method. Where the inner nodes, and outer soilhbe node, linked with transient thermal conditions are described as follows: Inner nodes
Outer soilhbe node
The timestep considered complies with the following stability criteria:
1
FO(~+B~)S2
aAt where Fo, is the Fourier number given by Fo = -
&I2
hAx and Bi, the Biot number given by Bi = k
Earth Tube Ventilation System
263
Air flow analysis. The analysis refers to the air flow inside the horizontal part of the tube (see Fig. 3). In what concerns the air flow inside the tube, a fully developed region can be 4m assumed for > 10 and a Reynolds number given by Reg = -,[191. EDP Considering the flow through a circular tube and performing an energy balance in a given control volume, (Fig. 4) the transient air temperature at the exit of the tube (Tout) can be calculated (Eq. 5).
76
ED’ Heat flux from air flow: q, = p---mcpTn Ix 4 Heat flux from soilhbe interface: q, = hnDrizAx (T’ - Tn) zD2 . Heat flux leaving the tube: q, = p-mcpTn 4
Ix + Ax
I
Fig 4.Flow through a circular tube
It should also be noted that the above equation is only valid for a fully developed region, where h (heat transfer coefficient) can be obtained through the equation of Dittus-Boelter, [22] where the Nusselt number is expressed through: NUD = 0.023Re;l5
(6)
Prn
where n = 0.4 (heating) and n=0.3 (for cooling), for Pr = 0.7 and fully developed turbulent pipe flow. When considering the tube length corresponding to the entrance region, the existent correlations do not easily apply; therefore a few tests were carried out in order to describe the entrance flow region for earth tube ventilation systems. These tests will be further discussed ahead. The correlation developed can be described as follows: h ( x ) = Aln(x) + B
where A = - b a x - b i n B
and g
(7) =
&ax
+bin
2
The above relationship will thus enable re-writing Eq. 5, increasing the scope of its validity, acting as a correction factor for the tube inlet temperature (Ti,,).
M.A.A. Costa, M.S.A. Oliveira & N. Martins
264
-7ZD -(Aln(x)+(B-A)x)
Tout = T,
+ (qn- T').
exp
rizC,
(8)
3. Results & Discussion The results obtained will be presented in comparison with a three-dimensional transient analysis using advanced CFD tools [20],aiming at both the simplified model validation and the support for the above referred assumptions. In order to be able to answer to issues concerning: the relative influence of the air flow inside the soil and therefore the soil ability to dissipate the stored heat; the influence of the entrance flow region and the validity of the proposed correlation (Eq. 7); the influence of tube length on the heat transfer coefficient and the influence of seasonally thermal amplitudes on the system thermal performance; a transient process was modelled for a period of three years, with a time step of a week. Two different soil types were analysed, a sand based soil and a standard soil characterized by [211. A sinusoidal boundary condition describing both inlet air temperature and surface ground temperature was imposed (Eq. 1). The seasonal thermal amplitude between 30 1K and 279K (Summer and Winter average temperatures for Aveir*Portugal was considered [23]. In order to analyse the soil capacity for heat dissipation and the influence of the air flow on the soil thermal stability, a study is presented in Figs. 5-8 (results referring to the 3D transient analysis). From the results it can clearly be seen, how the presence of the tube affects the soil thermal stability, mainly during Summer (Fig. 5) and Winter (Fig. 7). This study therefore justifies the introduction of a transient heat flux term from the interface soilhbe along the tube air flow (q3)on the simplified model energy balance.
Temperature (Cantour 2) 301 300 299 298 296 295 -294 293 292 291 289 288 281 286 285 284 282 281 280 279
Fig. 5 . Soilhbe system cross-section during Summer.
265
Earth Tube Ventilation System
As far as the influence of the entrance flow region is concerned, it can clearly be depicted from Fig. 9 that the heat transfer coefficient (h) is highly influenced by the tube length. Is it therefore postulated that the air flow thermal analysis could incur inaccuracies if an entrance region was not taken into account. The latter was therefore considered for < 10 , in terms of air tube characteristic dimensions corresponding to approximately
6
flow exit temperature calculation through Eq. 8.
Temperature (Contour 2) 301 300 299 298 - 2295 96
-294
Fig. 6 . Soilitube system cross-section during Autumn.
Fig. 7. Soilitube system cross-section during Winter.
MA.A. Costa, M.S.A. Oliveiru & N. Martins
266
Figs. 10-11 highlight, for a chosen tube buried depth of 2m and for the two soils type studied, the air flow exit temperature acquired with both 1D and 3D models with the same boundary conditions, for the same duration period of 3 years. From these results a few conclusions can be withdrawn: the ID model can compute the air flow temperature value with an error of approximately 1%. It should also be noted that the 3D analysis implied a computational time of over than 3 hours whereas the ID model required only a few minutes. The ID model due to its simplicity is able of becoming a module in a prefeasibility tool, not requiring a great deal of expertise, whereas the 3D model implied specific knowledge and training in computational fluid dynamics physical and mathematical modelling.
(Contour ature Temper 2)
301 300 299 298 296 295 -294 293 292 29 1 289 288 287 286 285 284 282 281 280 279
[Kl
Fig. 8. So ilh b e system cross-section during Spring.
-+-PI17 PI30 PI43
-
PI56
I
0
2
4
6
8
10 12 14
16
18
20
22
24
26
x Iml ~
Fig. 9. Heat transfer coefficient variation with tube length.
Fig. 12 represents the variation of the air temperature difference between the exit and inlet ports (Tout-Ti,) for the ETVS, again for a 3 years period and two different soil types. The influence of the soil characteristics and the climatic conditions is quite
E m h Tube Veatila-tionSy.ytein
267
notorious; both factors infer a heavy dependence on site gocation upon the system t h e ~ n a ~ effiectiveness. Perhaps a better understanding of these factors could be acquired if the model would include a cons~derationof daily thermal amplitudes. The latter can easily be i ~ ~ l e m e non t ea~revised tool version. From the analysis can also be postulated that the greatest thermal effectiveness correspond to Summer (where cooling is required) and Winter (where heating i s concerned). As expected, the system does not appear to be effective on Spring mnd Autumn, due to the small thermal amplitudes observed durhg these periods.
0
26
13
52
33
65
78
91 104 117 130 143 156
Time [Weeks)
Fig. 10. Air flow exit temperature for a sand based soil (tube depth of 2m).
l
i
-
-
0
13 26
-
39
-
52
65
-
78
-
I
91 104 117 130 143 158
Time (Weeks)
i
Fig. 1 I . Air flow exit temperature for a standard soil (tube depth of 2m).
F
I
I
Fig. 12. ETVS thermal effectiveness for two different soil tg.pes.
MA.A. Costa, MS.A. Oliveira & N. Martins
268
Once the time variation of the ventilation air temperature difference between the ETVS inlet port and outlet port is know, it is possible to determine the energy gain during a defined period, typically one year. As different technologies are many time used to produce heating and cooling, specific energy gains have to be determined separately as follows:
Winter and Summer periods are respectively defined as those when (To,, - qn)> ATmin and
(Tn - Tout) > ATmin.
Theoretical value for ATminis zero, however, for practical purposes it may be considered a slightly higher value, e.g., 0.5 “C, to take into account only relevant thermal effectiveness (i.e. neglecting minimal local effects).
4. Green House Gas Emission Module Greenhouse gases that are most relevant to energy project analysis are carbon dioxide (C02), methane (CH,) and nitrous oxides (NO,); these gases are considered in the GHG emission reduction analysis. Global Warming Potentials (GWPs) have been assigned to greenhouse gases to allow for comparisons of their relative heat-trapping effect. GWPs of gases are defined as a unit multiple of that given to carbon dioxide, which is assigned a reference value of 1. The default values used are those defined by the Revised Intergovernmental Panel on Climate Change (IPCC) Guidelines for Greenhouse Gas Inventories [24-251 (1 ton CH4 = 2 1 ton C 0 2 and 1 ton NO2 = 3 10 ton COz). The GHG Module provides an estimation of the GHG emission reduction based on the data entered by the user, tabulated GHG emission factors (EF) and values calculated in other modules, namely, annual energy produced. C02, CH4 and NO, emission factors (EF) represent the mass of GHG emitted per unit of energy generated. Emission factors will vary for different types and qualities of fuels, and for different types and sizes of heating equipment. For each fuel type, units are given in mass of gas emitted per unity of primary heating or cooling energy generated. For the base case heating system, emission factors corresponding to the heating fuel type selected are provided. If the heating fuel type is electricity, emission factors of the base case electricity mix are used. For the base case cooling system, emission factors for the base case electricity mix are used. The default values are given in Table 1. Custom data for GWPs and EF can be used instead.
269
Earth Tube Ventilation System Table 1. Default Emission Factors and Conversion Efficiencies [ 2 6 ] .
Fuel Type
EF - Emission Factor [Kg/GJ]
Fuel conversion efficiency [%]
GHG emission reduction is calculated as equivalent tons of COz avoided per year as follows:
(
(
GHGreducfion = EFbase case ' EETVSgain)heoring + EFbase case ' EETVSgain)cooirng
(1 1)
5. Cost Analysis Module The Cost Analysis Module is used to estimate the system cost. Two major types of costs are considered, initial investment and maintenance costs. The general structure of this module is based on the use of unitary tabulated costs both for labour and materials. The knowledge about unitary costs and case study conditions (given by the user in the energy analysis module) are enough to provide an estimation of the total project costs. Initial investment cost includes the following cost categories: Feasibility analysis; typically includes such items as site investigations, soilhydrology assessment, environmental assessment, a preliminary system design, including loop sizing and layout, detailed cost estimate, and a final report. This is particularly important for projects larger than 100 kW. Development; includes costs for contract negotiations, permits and approvals, project financing, development phase project management and any development related travel costs. Engineering; includes costs for the GSHP system design, tenders and contracting, and construction supervision. Renewable energy equipment; as defined here, includes, when applicable, the system's ventilators, drilling, trenching, pipes, valves, fittings, electrical equipment, controls civil engineering and transportation costs. Contingencies; the allowance for contingency items should be based on the level of accuracy associated with the pre-feasibility estimate of the project costs. Typically, a pre-feasibility level cost analysis should be accurate within 40 % to 50 YO.However, this accuracy will depend on the expertise of the study team, the scale of the project being considered, the level of effort put forward to complete the pre-feasibility study and the availability of accurate information. It is certainly possible that experienced users could estimate costs in the range of 10 % to 20 YOof the total initial project costs (excluding contingency costs). Periodic cost represents recurrent costs that must be incurred at regular intervals to maintain the project in working condition. ETVS may require yearly inspection and
270
M.A.A. Costa, M X A . Oliveira & N Martins
cleaning procedures and as there are not major moving parts only the ventilator may require to be replaced e.g., every 10 years. The output of the Cost Analysis Module is the required investment cost and annual maintenance costs for a system as defined in the Energy Analysis Module. These represent the negative project cash-flows that should be balanced by the positive ones estimated in the two previous modules. The analysis of their combined effect along the project life is made by the Financial Analysis Module.
6. Financial Analysis Module The purpose of the Financial Analysis Module is to determine financial indicators taking into account the project generated cash-flows (positive and negative) along its life-cycle (several years) under defined economic conditions. The project costs (negative cashflows) are estimated in the Cost Analysis Module. The project gains (positive cash-flows) result from the combination of the energy delivered by the system (heating and cooling) and respective market value. Major inputs are the retail price of base case energy which is multiplied by the heatinglcooling energy delivered by the ETVS to determine the avoided cost of heating and cooling energy, the energy cost escalation rate, inflation rate, discount rate, project life and environmental credits and/or subsidies. Resulting outputs are financial indicators such as the Return on Investment (ROI), Internal Rate of Return (IRR), Simple Payback Time, Net Present Value (NPV) and Year-to-positive cash flow. As the project life can be as long as 30-40 years, it is not easy to predict for such a long period the energy cost escalation rate or the inflation rate. However, the resulting financial indicators are strongly dependent of these financial parameters. For the reasons referred, there is a implicit 5'h step in the tool which consists in a sensitivity analysis of the project under analysis to the most uncertain parameters.
7. Conclusions The paper presents an intermediate stage of development of pre-feasibility tool for Earth Tube Ventilation Systems. Pre-feasibility tools for engineering assessment of ETVS assume major importance in what concerns market dissemination. The present study proposes a modular methodology suitable for engineering level pre-feasibility analysis with user reduced input requirement. The simplified physical and mathematical model proposed was developed and tuned with the support of advanced and well established CFD tools. The results obtained with the simplified model present a deviation less than 1% when compared with CFD predictions. The proposed methodology is able to convert estimated energy savings into financial indicators which take into account direct and indirect costs, macro-economic parameters, energy escalation factors and externalities, namely GHG emission taxes. Furthermore, this methodology appears to be quite promising in what concerns its suitability as an engineering project assessment tool.
Earth Tube Ventilation System
27 1
References
[ l ] Antinucci M, Fleury B, Lopez D, Asian J, Maldonado E, Santamouris M, Tombazis A, Yannas S. (1992) Passive and hybrid cooling of buildings-state of art. Int. J. Solar Energy 11: 251-272. [2] Stevens WJ. (2004) Optimal placement depth for air-ground heat transfer systems. Applied Thermal Engineering 24: 149-157. [3] Bi Y, Chen L, Wu C. (2002) Ground heat exchanger temperature distribution analysis and experimental verification. Applied Thermal Engineering 22: 183-189. [4] Mihalakakou G. (2002) On estimating soil surfaces temperature profiles. Energy and Buildings 34: 251-259. [ 5 ] Chunga M, Junga PS, Rangelb RH. (1999) Semi-analytical solution for heat transfer from a buried pipe with convection on the exposed surface. Int. J. Heat and Mass Transfer 42: 3771-3786. [6] Deglin D, Van Caenegem L, Dehon P. (1999) Subsoil heat exchangers for the air conditioning of livestock buildings. J. Agric. Engng. Res. 73: 179-188. [7] Kumara R, Rameshb S, Kaushika SC. (2003) Performance evaluation and energy conservation potential of earth-air-tunnel system coupled with non-air-conditioned building. Building and Environment 38: 807-8 13. [8] De Paepe M, Janssens A. (2002) Thermo-hydraulic design of earth-air heat exchangers. Energy and Buildings 35: 389-397. [9] Pahuda D, Matthey B. (2001) Comparison of the thermal performance of double U-pipe borehole heat exchangers measured in situ. Energy and Buildings 33: 503507. [lo] Bojic M, Trihnovic TN, Papadakis TG, Kyritsis S. (1997) Numerical simulation, technical and economic evaluation fair-to-earth heat exchanger coupled to a building. Energy 22 (12): 1151-1158. [ 111 Pfafferott J. (2003) Evaluation of earth-to-air heat exchangers with a standardised method to calculate energy efficiency. Energy and Buildings 35: 971-983. [12] Hollmuller P, Lachal B. (2001) Cooling and preheating with buried pipe systems: monitoring, simulation and economic aspects. Energy and Buildings 33: 509-5 18. [ 131 Hollmuller P. (2003) Analytical characterisation of amplitude-dampening and phase-shifting in airlsoil heat-exchangers. Int. J.Heat and Mass Transfer 46: 43034317. [14] Gauthier C, Lacroix M, Bernier H. (1997) Numerical simulation of soil heat exchanger-storage systems for greenhouses. Solar Energy 60 (6): 333-346. [15] Serres L, Trombe A, Conilh JH. (1997) Study of coupled energy saving systems sensitivity factor analysis. Building and Environment 32 (2): 137-148. [16] Delalic N, Mulahasanovic Dz, Ganic EN. (2004) Porous media compact heat exchanger unit-experiment and analysis. Experimental Thermal and Fluid Science 28: 185-192. [ 171 Kabashnikov VP, Danilevskii LN, Nekrasov VP, Vityaz IP. (2002) Analytical and numerical investigation of the characteristics of a soil heat exchanger for ventilation systems. Int. J. Heat and Mass Transfer 45: 2407-2418.
272
MA.A. Costa, MS.A. Oliveira & N. Martins
[18] Reesa SW, Adjalib MH, Zhoua Z, Daviesc M, Thomas HR. (2000) Ground heat transfer effects on the thermal performance of earth-contact structures. Renewable and Sustainable Energy Reviews 4: 213-265. [19] Mihalakakou G, Lewis JO, Santamouris M. (1996) The influence of different ground covers on the heating potential of earth-to-air heat exchangers. Renewable Energy 7 (1): 33-46. [20] Ansys-CFX5.6. http:/lwww-waterloo.ansys.com/cfx. [21] Incropera FP, DeWitt DP. (1998) Fundamentals of Momentum, Heat and Mass Transfer, 3" Edition, McGraw-Hill, New York. [22] Welty JR, Wicks CE, Wilson RE, Rorer G . (2001) Fundamentals ofMomentum, Heat and Mass Transfer, 4" Edition, John Wiley & Sons, Inc. USA. [23] Surface Meteorology and Solar Energy; http://eosweb.larc.nasa.gov/cgi-bin/sse/. [24] IPCC Reference Manual page 1.13: http://www.ipcc-nggip.iges.or.jp/public/gvguidelin/chl refl .pdf. [25] IPCC Reference Manual pp. 1.35 and 1.36: http:/Iwww.ipcc-nggip.iges.or.jp/public/gllguidelin/ch1ref2 .pdf. [26] Fenhann J. (1999) Projections of emissions of greenhouse gases, ozone precursors and sulphur dioxide from Danish sources until 2010. The Danish Energy Agency, December.
GREENHOUSE GASES EMISSION REDUCTION IN AN URBAN BUS FLEET: APPLICATION TO A CASE STUDY IN FUNCHAL, MADEIRA ISLAND A.M. SIMOES('), P. LAGES'~),T.L. FAMAS''), C.M. SILVA"), M.N. AGUAS(*) (I)
Mechanical Engineering Department, Instituto Superior Tecnico, Av. Rovisco Pais 1, 1049-001 Lisbon, Portugal, andreia@navier. ist.utl.pt ()'
Iberdrola II,Lda, Avenida da Liberdade, 180-A, 7', 1250-146, Lisbon, Portugal
The Transport Sector is now one of the main consumers of energetic resources. In Portugal, an Energy Audit and Energy Consumption Rationalization Plan was published in the form of law number 58/82, set out in the RGCE-regulations. The primary objective of the RGCE-regulations is to influence companies to consider energy as a cost factor. Auditing a bus operator company implies understanding where energy is consumed and how can it be more efficiently used. In this kind of company the vehicles are the main consumers of primary energy. Trying to improve the energy efficiency and consequently reduce the greenhouse gases emissions of all the fleet implies to analyze those vehicles in normal operational conditions. This is not an easy task, because it means to study all the vehicles operating in all the network lines, for different traffic conditions and different occupancy rates. TRANSTEP is the new developed methodology that intends to simplify this task. This new methodology estimates energy saving gains in operational functioning, from reduced amount of experimental data.
1. Introduction
The Transport Sector is now one of the main consumers of energetic resources, and the energetic consumptions, pollutant emissions and greenhouse gases emissions associated to this sector are increasing, 000. Measures to prevent the trend of this evolution are being implemented all over the world. In Portugal, an Energy Audit and Energy Consumption Rationalization Plan was published in the form of law number 58/82, set out in the RGCE-regulations, prepared by the General Directorate for Energy, DGE. The primary objective of the RGCEregulations is to influence companies to consider energy as a cost factor. The regulations establish goals for companies to progressively reduce their specific energy consumption. In addition, the regulations require companies to account their energy use, to prepare and carry out plans for rationalizing their energy consumption and to achieve the energy savings. Energy savings necessarily implies greenhouse gases emissions reductions, what converges with the establishments of Kyoto protocol. This way, Portuguese transport companies with energy consumptions above 500 toe must be audited and establish an energy rationalization plan for following three years'. Auditing a bus operator company implies understanding where energy is consumed and how can it be more efficiently used. In this kind of company the vehicles are the main consumers of primary energy. For the case study presented - Horarios do Funchal (the urban bus operator of the city of Funchal in Madeira Island2)-fuel consumption is responsible for 95.9% of the primary energy consumption. Trying to improve the energy efficiency and consequently reduce the greenhouse gases emissions of all the fleet implies to analyze those vehicles in normal operational
' Portaria
n." 228/1990, 27th March, approves the regulation for management of energy
consumptions for the transport sector. http://www.horariosdofunchal.pt/.
273
274
A.M. Sirnoes et al.
conditions. This is not an easy task, because it means to study all the vehicles operating in all the network lines, for different traffic conditions and different occupancy rates. TRANSTEP is the new developed methodology that intends to simplify this task. This new methodology estimates energy saving gains in operational functioning, from reduced amount of experimental data. This new methodology was applied to a case study in Funchal, Madeira Island; Horarios do Funchal, S.A., company (further addressed to as HF). This company has 150 bus fleet and a network with 56 lines. HF exhibits special functioning conditions because of Funchal’s hilly topography. With the application of TRANSTEP was possible to characterize the fuel consumptions for the studied fleet and to establish a rationalization plan having in account the estimated impact of fuel saving measures regarding energy and emissions.
2. Methodology To evaluate all factors influencing the fuel consumption a simulation model (TRANSTEP) was developed, in order to estimate fuel consumption and greenhouse emissions for all the fleet in each line. TRANSTEP main input data is schematically presented in Fig. 1 and is discriminated as follows: Fleet characterization: number of vehicles and vehicles characteristics such as: Engine maps (power, consumption and torque as a function of load and rpm); Transmission ratios and efficiency; Gear box management maps; Tires characteristics; Vehicle’s dimensions and weight; Network operational characterization: number of lines and trip frequency, occupancy rate, number of bus stops (urban and non-urban) and vehicle used in each line; Network topography characterizations: distances (urban and non-urban) and slopes of each line; Experimental measurements data: vehicles dynamics, topography data, registered events (e.g. time on a bus stop, stops because of traffic or lights); 0 Instantaneous fuel consumptions for specific events (accelerations, decelerations, cruise velocity and idle) for different slopes and traffic conditions. This data can be obtained experimentally by installing a flux meter in the bus. An alternative solution is to use numerical simulation models that can estimate on a micro scale level the fuel consumption of a predefined bus running in a specific line. In the present study the authors used a numerical model developed at Instituto Superior Tcnico named EcoGest3 0.
EcoGest solves the dynamic laws of vehicles for specific accelerations and decelerations curves of
typical driving modes. EcoGest is capable of calculating along the trip the localization of the vehicle, vehicle velocity, position of the accelerator pedal, gearbox selection, and the engine rotation speed. In addition, using the throttle position and the engine speed, EcoGest is capable of estimating instantaneous fuel consumption as well as instantaneous and average NOx, CO, CO, and HC exhaust emissions.
Greenhouse Gases Emission Reduction
275
Some of this data can be supplied by the bus operators or by the vehicles manufactures. However, experimental measurements in normal operational conditions are always needed leading to a complicated and cumbersome task. Fleet
Experimental
Network
treated data
Experimental data
treated data
Fig. 1. Scheme of interaction of TRANSTEP and its components.
Since it takes too much time to evaluate every single line and bus for different periods, the methodology involves the selection of two or three characteristic lines for a detailed experimental evaluation. This experimental data is further on treated and fed into TRANSTEP. The output results of TRANSTEP are effective fleet fuel consumptions disaggregated in: 0 Fuel consumption for different events; 0 Fuel consumptions for different lines; 0 Annual fuel consumptions to the fleethework; 0 Elasticity of each parameter influencing fuel consumption. With these results it becomes possible to identify measures in order to improve the company’s energy efficiency and, therefore, reduce the fuel consumption of the fleet. These results are also used to estimate greenhouse emissions based on fuel conversion factors into COz emissions (for diesel vehicles green house emissions are basically COz emissions), see Table 1. Table 1 . Diesel conversion factor into CO2 emissions.
I I
kg C 0 2per unit
Fuel Type Diesel (1)
I
2.68
I I
Sources: National Air Emissions Inventory, UK Greenhouse Gas Inventory, Digest of UK Energy Statistics DTI 1998, Greenhouse Gas Inventory Reference Manual IPCC 1996 (http://www.defra.gov.uWenvironmen~envrp/gas/O5 .htm).
3. Experimental Measurements
Experimental measurements should occur after choosing representative lines of the network and representative vehicles of the fleet. These measurements must represent
A.M. Simoes et al.
276
normal operational conditions at different periods of the day and week, in order to allow the c h a r ~ c ~ e ~of~ peak t ~ oand n off peak hours, and week days versus weekend days. During experimental measurements data is collected regarding the ~ o p o ~ ~ of phy each analyzed line, vehicles dynamics (speed versus time and distance), s ~ ~ i ~ c a n t events, occupancy rates, atmospheric conditions and road surface conditions. For the case study presented, the equipment used for the topography and dynamic c h ~ a ~ ~ e r i ~ awas t i othe n following: PP Sumto’s Escape 203 altimeter to register altitude; B) Corssys Datron Sensorsystem, M3, Doppler’s effect Microwave sensor, to register distances and instantaneous velocity and accelerations~ e Laptop, for data logging. For this case study, it was not possible to install flux meters in the buses, therefore our in-house developed numerical model EcoGest was used to estimate the i n s ~ ~ n e o ufuel § consumption. These data was then calibrated with measwed data regarding monthly and m u a l fuel consumption. In this particular application, two measurements methods were adopted, one with the equi pm ~n~ installed directly in the bus and other with the equipment installed in a car that followed closely the bus along the route, thus emulating its dpamics. In the filrst case the equipment was installed in the front of the bus and the electric feeding was provided by an auxiliary battery as presented in Fig. 1.
Fig. 2. Experimental measurement’sequipment installed in the front ofthe bus.
In the second case the equipment was installed on the car door as shown in Fig. 3.
Fig. 3. Experimental measurment’s equipment installed in the lateral ofa car.
4, ~~~~~~~e~~~~ Data
The e x p e r i m e n ~data ~ is then analyzed and treated. ~ i ~ i ~ cfuel a ncto n s ~ p t i oevents ~ are dent^^^^, namely as h c t i o n of:
211
Greenhouse Gases Emission Reduction
Idle at bus stops, traffic stops and light stops, for different periods of the day; Idle at terminus bus stops; Occupancy rates, for different periods of the day; 0 Mean distances at accelerations and decelerations in stops. This data was then used to calculate some of TRANSTEP parameters, such as effective bus stop factor, urban light stop factor, urban and non urban traffic stop factor, idle time at each event and mean distances in acceleration and deceleration at each event. The vehicles dynamics are used to determine instantaneous fuel consumptions using EcoGest. These results are then disaggregated in order to evaluate the evolution of consumption for the studied fleet. The criteria used to disaggregate are: 0 Consumption due to slope; 0 Consumption due to resistance to motion; 0 Consumption due to drag; 0 Consumption due to inertia; Idle consumption; Acceleration consumption; 0 Deceleration consumption; Consumption as a function of slope (classes of slopes were defined based on the lines topography, in order to simplify the characterization of the topography of each line). These disaggregated data are then calibrated against average measured fuel consumption and finally used in TRANSTEP. 0
5. Results
From the analysis of the collected data from the bus operator, bus manufacturers, experimental measurements and EcoGest results it was possible to calculate the evolution of fuel consumption and greenhouse gases emissions with slope4in a normal operation condition and in acceleration and deceleration for different vehicle gross weights. Fig. 4 shows the cumulative distribution of slope for the entire HF network. 100%
-c
90% 80%
C
.P 70% a 60%
-'
g
j
50% 40%
30% 20% 10%
J
0% 0%
I
I
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
s l o p (%)
Fig. 4. Cumulative distribution of slope for the entire HF network.
Classes of slope were created having in account the topography characteristics of the network
studied.
278
A.M. Sirnoes et al.
Based on this distribution classes of slope were defined, they are presented in Table 2. Fig. 5 presents for different weights, the evolution of CO;! emissions, during rolling (C02 emissions during acceleration from stops and idle in stops are not included in rolling), along slope variation. From these results is possible to conclude that COz emissions for negative slopes is not very diverse for each weight. While for positive slopes those emissions are quite higher for the weightier vehicles, particularly in hilly conditions. Figs. 6 and 7 present the evolution of C 0 2 emissions in acceleration and deceleration from stops, in opposition to gross vehicle weight. Table 2. Slope classification. Class of Sloae
Slooes
-5
Lower than - 12%
-4
[-S%;-l2%]
-3
[-5%;-8%]
-2
[-2%;-5%]
-1
r0%;-2%1
1
[0%;2%]
2
[2%;5%]
3
[5%;8%]
4
[8%; 12%]
5
Higher than 12%
In Figs. 5-7 the most significant results are presented.
-5
-3
1
1
3
5
Slope Class
Fig. 5 . Rolling CO1 emissions versus class of slope for different gross vehicle’s weights.
219
Greenhouse Gases Emission Reduction
~
0
~
7
8
9
10
11
12
13
14
15
16
17
Vehicle Gross Weight (ton)
+ Acceleration from light slops
m
& Acceleration from bus stops --Linear
Acceleration from traffic slops (Acceleration from light stops)
-Linear
-Linear(Acceleration
(Acceleration from traffic stops)
from bus stops)
Fig. 6. COz emissions in acceleration from stop versus gross vehicle weight. 4
+
3.5
-.B
3
+
2
-I
2.5
n
0
2
'i 1.6 0" 1
0.5
0
7
a
9
10
11
12
13
14
15
16
17
Vehicle Gross Weight (Ion)
Fig. 7. COz emissions in deceleration until stop versus gross vehicle weight.
These C 0 2 emissions evolutions were then used in TRANSTEP in order to determine disaggregated emissions for all the fleet. The global results are presented in Fig. 8.
68%
Fig. 8. Disaggregated COZemissions, for each event, calculated for all the HF fleet in 2002.
A.M. Simoes et al.
280
After characterizing the actual COz emissions, it was possible to begin the sensitive analysis, in order to propose measures for improving energy efficiency and therefore reduce the emission of greenhouse gases. These measures essentially focus on three levels of possible action the network, the vehicle and the motorist. Therefore, the motorist effect on COz emissions was one of the evaluated factors. In order to do it, the velocity profile measured during the experimental data collection was modified. The accelerations were softened according to a factor5.In Fig. 9, the C02 emissions evolution according to the acceleration factor for two HF network lines, line 21 and line 24, is presented. Line 21 is quite hilly and line 24 has soft slopes, that is why soften accelerations implies higher C02 emissions reduction in line 24. In line 21 the driving behaviour can not be submitted to significant changes. The CO;! emissions reduction related to each acceleration factor are presented in Table 3.
E 0
2 150 1
.E
100
w 50
0 0.5
0.6
0.7
1
0.9
0.8
1.1
1.2
5 1.3
Acceleration Factor
1 + Line 21
8 Line 24 -Poly.
(Line 21) -Poly.
(Line 24)
1
Fig. 9. CO2 emissions versus acceleration factor. Table 3 . COZemissions reduction for different acceleration factors. Acceleration Factor
a) For line 21 was not considered acceleration factors above 1, given that in almost all cases load is already maximum.
Despite the fact softer accelerations represent lower COz emissions, it also implies longer travel time, as presented in Fig. 10, nevertheless, this extended travel time is not significant for line 24, because it only adds about five minutes to normal travel time, what can be easily compensated with less time in terminal bus stops (which also leads to less time on idle in terminal bus stops). Acceleration factors from 0.5 to 1 represent 50% to 0% of acceleration softening.
Greenhouse Gases Emission Reduction
28 1
40
I.
I
20
0,s
0.8
0,7
0.8
1
0.9
1.1
12
1.3
Acceleration Factor
1 + Line 21
Line 24 -Poly.
(Line 24) -Poly.
(Line 21)
1
Fig. 10. Travel time versus acceleration factor.
It was also evaluated the effect of vehicle’s gross weight reduction, minimization of idle time at stops and improvement of traffic conditions. Table 4 shows, for the case study, the energetic achievements and greenhouse gases emissions reduction associated to these measures. Table 4. Annual energetic achievements and greenhouse gases emissions reduction associated to each measure.
For reducing the vehicle’s gross weight it was suggested a plan for the fleet renewal were the vehicles should be 10% lighter6, [5-61. In three years it would be possible to achieve a 3% weight reduction on the fleet, by renewing three vehicles. This way the annual C02 emissions could be reduced in 1.65%. This measure also reduces fleet age which also has advantages in what concerns pollutant emissions. Regarding idle time at terminal stops it was noted that it was responsible for 3% of the annual fuel consumption and COz emissions. For solving this problem it was suggested that during motorist’s training, they should be taught in order to achieve idle of top 60 seconds. This measure can accomplish a 2% reduction of annual C02 emissions. In what concerns Ecodriving it is possible to achieve almost 2% reduction of annual COz emissions. This measure also involves motorist training in order to reduce speedy driving.
‘
At the moment there are Canadian and European research projects which purpose is to reduce vehicles weight in order to improve energy efficiency.
A.M. Simoes et al.
282
Better traffic conditions, implies reducing traffic and light stops. This measure is perhaps the most difficult one when it comes to implementation. Although in the case study presented it is possible to achieve a reduction of 50% of traffic stops just by controlling parking stops and by giving priority to buses in some urban streets. This measure can achieve 1.2% reduction on annual CO2 emissions. All these measures promote lower pollutant and greenhouse gases emissions, lower fuel consumptions and lower operational and maintenance costs.
6. Conclusions and Further Work This new methodology simplifies energy audits and reduces the amount of experimental data needed. The experimental data collected for this case study also allowed the development of equations that describe the fuel consumption and C 0 2 emissions of urban buses in normal operation conditions, for different slopes and different vehicle weights, in order to characterize every possible line and occupancy rate of the vehicle. This way it is possible to use it for any urban bus fleet, taking into consideration similar mean velocities. This is particularly interesting when it is not possible to collect experimental data in normal operation conditions. Although, these equations still present significant errors for a confidence of 95%, which means the sample is still small, therefore there is still need of experimental data. Especially during acceleration from stops, these events have high COz emissions variability. Table 5 summarizes the expectable error associated to rolling7 COz emissions estimation in TRANSTEP. Table 5. Error associated to rolling COz emissions estimation
Some of the authors are currently collecting additional experimental data near urban fleet’s operators in the cities of Oporto and Lisbon in order to improve the equations that represent fuel consumption and C02 emissions behaviour.
Acknowledgements This research was supported by the Portuguese Science and Technology Foundation FCT (Project POCTI/MGS/37601/01, approved by FCT and POCTI, financially supported by the European Community Fund FEDER).
Rolling does not include acceleration and deceleration from and to stops, in that order.
Greenhouse Gases Emission Reduction
283
References - Directorate-General for Energy and Transport European Union - Energy & Transport in Figures 2003, Office for Official Publications of the European Communities (ed), (2003). (URL verified at May 2004
[ 11 European Commission
http://europa.eu.int/commidgs/ener~~transport/figures/pocketboo~2003~en.htm).
[2] Instituto do Ambiente Relatorio do Estado do Ambiente 2001, September (2002). [3] CESUR and DCEA/FCT Estudos de Base - Medidas adicionais para o sector dos transportes. In Programapara Alteraqoes Climaticas, February (2003). [4] Silva CM, Farias TL, Mendes-Lopes JMC. (2002) EcoGest Numerical modelling of the dynamic, fuel consumption and tailpipe emissions of vehicles equipped with spark ignition engines. Proc Urban Transport and the Environment 2002Eighth International Conference on Urban Transport and the Environment in the 21st Century, 523-532. [5] Transportation Development Centre Building a Better Bus Lighter and More Energy Efficient (URL verified at May 2004 http://www.tc.gc.ca/tdc/publication/updates/v 14n1.htm). [6] European Commission - Research - Sustainable Surface Transport Drive an Mg and lose weight - it’s better for the environment (URL verified at May 2004 http://www.europa.eu.int/comm/research/transport/news/magnesium~en.html. ~
This page intentionally left blank
ENVlRONMENTAL IMPACT OF HYDROGEN IN URBAN
TRANSPORTS K.R. CALHAU, G.A. GONCALVES, T.L. FARIAS Instituto Superior Tecnico, DEM-STE, 1049-001 Lisboa, Portugal The Sociedade de Transportes Colectivos do Porto (STCP, the main public transport company in the city of Porto, Portugal) has introduced, under the project CUTE - Clean Urban Transport for Europe and together with other 8 European cities, 3 fuel cell transit buses in their fleet, with an hydrogen refuelling infrastructure supplied by BP. The objective of the CUTE project was to demonstrate and develop a locally emission-free and lownoise transport system, including the accompanying energy infrastructure. The hydrogen used in Porto was produced by electrolysis (using grid electricity) in a Linde factory in Alenquer (north of Lisbon). Three different hydrogen production strategies where evaluated: steam reforming of natural gas, water electrolysis using grid electricity (used in the project) and water electrolysis using renewable electricity. Two other fuels/technologies were also evaluated and compared to the hydrogen results: Diesel and natural gas. Two parameters with relevant impact (both environmental and energy) where evaluated: consumption of non-renewable energy as MJ per km and emission of greenhouse gases (GHG) as equivalent kg of C 0 2 per km.
1. Introduction Under the Kyoto protocol, Portugal was allowed a 27% increase of emissions of GHG (relative to the values of 1990) by 2010. As by 2000, emissions were already 30,4% above 1990 levels. The transport sector is the largest contributor to the global GHG emission, with about 30% of the total national emissions. Within the transport sector, 70% of the emissions result from road transports. Besides GHG emissions (to which pollutants should be added, although this is beyond the context of this paper), the growth of the road transport induces an ever increase of importation of fossil energy sources, creating a problem of security of supply. Renewables must contribute to a solution. To address these issues, the European Commission has developed policies with the objective of: Better integration of transports, using more environmentally benign and less energy demanding modes of transport; Stimulating the development of new technologies; 0 Improving road safety. Within this framework, the project CUTE-Clean Urban Transport for Europe was approved and, with the support of the European Commission, had the objective of introducing hydrogen fuelled vehicles in Europe. This was successfully achieved by the operation of 27 buses in 9 European cities (one of which is Porto). The buses operated for a period of 2 years, the Porto test starting February 2004. 2. Objective
Currently, about 30% of the STCP fleet runs on natural gas, the remaining being conventional Diesel powered. The three hydrogen based buses started operation in 2004 and finished en January 2006. Under the CUTE project, there is the need to appraise the technologies under scrutiny, both conventional and hydrogen. Two parameters are considered: non-renewable energy consumption and GHG emissions for the following five technologies: 285
K.R. Calhau, G.A. Goncalves & T.L. Farias
286
Conventional Diesel from refinery; Conventional compressed natural gas (CNG); Fuel cell with hydrogen produced from grid electricity; Fuel cell with hydrogen produced from renewable electricity; Fuel cell with hydrogen produced from steam reforming of natural gas.
0
3. Methodology The evaluation was conducted in three phases: first the production of the fuel up to the point it reaches the vehicle tank, Well-to-Tank (WTT); second consumption of the different fuels by vehicles, Tank-to-Wheel (TTW) and; third the integration of the two previous evaluations, providing the complete Well-to-Wheel (WTW) impact of each combination. 3.1. Well-to-Tank
For this phase, a life cycle analysis software was used, GAB1 [l], using the information available in its database and country specific information when available. The analysis considers only emissions of gases that contribute to the greenhouse effect, C02, C h and N20. The following values where considered regarding sources of crude (Table 1) and he1 properties (Table 2). Table 1. Energy resource properties [I] [4].
Table 2. Fuel properties [1] [4].
Diesel Compressed Natural Gas Hydrogen
LHV MJkg 43,O
MJ/Nm3
-
33,8
COz content g/MJ
72,8
56,4 0
119,3
For quantification purposes, the emissions of the three GHG are presented in gram of CO2equivalent, based on the IPCC tables of Global warming potential (Table 3). Table 3. Global warming potential [8].
Greenhouse gas
coz CH4
N2O
g C 0 2equivalent per g 1 21 310
Environmental Impact of Hydrogen in Urban Transports
287
3.1.1. Hydrogen - Electrolysis
In this process, hydrogen was produced by electrolysis in the Linde factory in Alenquer (ca 300 km south of Porto), compressed to 200 bar for transport to Porto and finally has a compression to 400 bar for filling the bus tanks. Two options were considered, in the first the electricity used is obtained directly from the Electricity grid, and represents an average of the total electricity production plants in the country. In the second option, the electricity for the electrolysis process is of renewable source, all other processes (compression, transport) are similar. The processes for electricity production were modeled according to the database of the GAB1 software. The efficiencies considered reflect the technology used. Renewable energy sources (wind, hydro, solar) are considered to have an efficiency of 1OO%, as they convert renewable energy directly into electricity, without a thermal cycle; solid waste use is also considered to have an efficiency of 100% due to the fact that it uses waste and not primary resources. The sources of energy for electricity production were obtained from [ 5 ] . Table 4 provides an overview of the electricity production in Portugal. Table 4. Sources of energy for electricity production in Portugal. Power plant
Coal Hydro Natural gas (simple and combined cycle) Fuel oil + Diesel oil Co-generation
rImtGrts Solid waste Wind
I Contribution
I Efficiency
I 34.28%
130% 100% 43% 29% 58% 133% 100% 1100%
25,55% 14,07% 11,71% 10.90% 12.00% 1,17% 10.31%
I
1
I
According to [ 5 ] , losses in the transport grid amount to 1,7% of total electricity production, the losses in the distribution grid are 8,4%. Global efficiency of production, transport and distribution is 42,7%. When only renewable electricity is used, the overall efficiency is 90%. Table 5 . Energy consumption, losses and GHG emission for hydrogen production-grid electricity
According to [4], the efficiency of electrolysis plant is 65%, with an energy consumption of 0,43 kWh/MJ of hydrogen, In the case of the equipment used in the Linde Factory, the final pressure of the hydrogen is 10 bar. This hydrogen suffers a primary compression up to 250 bar for transport (the tanker pressure is 200 bar). It is assumed that the efficiency of the compressor is the same than the one used for natural gas: 68%.
K.R. Calhau. G.A. Goncalves & T L . Farias
288
The compressed hydrogen is transported by road using diesel tankers, the total distance covered is 300 km,the return trip also considered. The Secondary compression raises the pressure from 200 to 400 bar to fill the vehicles. The results for the production of hydrogen are presented in Table 5 for grid electricity and in Table 6 for renewable electricity. Table 6. Energy consumption, losses and GHG emission for hydrogen production-renewable electricity
Production of H2 Primary compression Transport of Hz Compression and filling
Total
0 0,089 0,010
0,007 0,089
0 0,079 0,003 0,007 0,089
0 88,7 3,4 7,9 100
0 14,66 0,20 1,36
16,22
0 90,4 12
8,4 100
3.1.2. Hydrogen - Steam Reforming In this situation, hydrogen is assumed to be produced in the STCP depot by steam reforming, using the same natural gas supply route already installed. This process will be addresses further in section 3.1.4 been mentioned, and therefore only the steam reforming process will be analyzed. The hydrogen pressure at the output of the steam reforming plant is 15 bar, and according to [3], the efficiency of the process is 70%. In this situation, for filling the bus the hydrogen must be compressed from 15 to 400 bar, a higher compression ratio than in the previous case (from 200 to 400 bar). The global results are presented in Table 7. Table 7. Energy consumption, losses and GHG emission for hydrogen production-steam reforming.
GHG emission
1 Total
I
1,589
I 0,589
I
100
I
103,67
3.1.3. Diesel
For the production of Diesel &el it is assumed that all the crude is supplied to the Leixoes refinery (ca 10 km north of Porto) through sea transport, where it is refined. The fuel is then supplied by road tanker to the STCP depot, where the buses are filled. According to Galp Energia [2], the crude imported in 2002 was supplied by South America, Western Africa, CIS, Middle East, North Sea and Mediterranean, with the contributions presented (Table 8).
Environmental Impact of Hydrogen in Urban Transports
289
Table 8. Origin of imoorted crude.
28,5% 19,7% 5.3% 27,2% 13,2% 6,1%
Western Africa South America CIS Middle East North Sea Mediterranean
According to existing studies [3] [4], the efficiency of crude extraction is assumed to be 98 % (this efficiency refers to the amount of energy spent in extracting one unit of energy in the form of crude, in this case, 1,02 MJ of energy are needed to extract 1,OO MJ of crude. The emissions associated with the extraction were calculated according to the same sources, providing a value of 3,245 g C 0 2 equiv per MJ of crude extracted. The transport of the crude includes transport from oil wells to central storage, there to terminal storage, to refinery and storage in refinery. From the terminal storage to the refinery the crude is transported by sea tanker. The distances traveled by the tankers are summarized in Table 9. Table 9. Average distance to the Leixoes terminal (KM_.
Western Africa South America CIS Middle East North Sea Mediterranean
6707 9414 546 1 8127 2537 4615
The refining of the crude takes place in the Leixoes refinery, the values for energy consumption and emissions were taken from literature sources [3]. From the same source the overall efficiency of refining is 89%. The distribution of the Diesel fuel includes transport to terminals through pipeline, storage and distribution by road tanker to the filling stations. The distance by road is assumed to be of 75 km, with the same distance in the return trip. The energy consumption of refilling the bus is of 0,0009 k W M J of electricity. The overall Well-to-Tank analysis is presented in Table 10. Table 10. Energy consumption, losses and GHG emission for diesel fuel production
I
Crude extraction Crude transport Crude refining Diesel fuel distribution Vehicle filling Total
1,154 1,134 1,129 1,005 1,003
I 1,154
I
0,020 0,005 0,124 0,002 0,003 0,154
1
12,9 3,6 80 1,4 231 100
3,25 0,38 9,71 0,14 0,72
I 14,20
I
22,s 23 68,4 1,O 5,o 100
I
K.R. Calhau, G.A. Goncalves & T.L. Farias
290
3.1.4. Compressed Natural Gas
The natural gas used is extracted in Algeria and transported through pipeline to the STCP depot. According to [4], extraction efficiency is 99,4%. Natural gas processing includes removal of contaminants as C 0 2 and heavy hydrocarbons. In the process leaks occur, and this adds to the total emissions of C 0 2 equivalent. The natural gas is transported by pipeline, at a pressure of ca. 70 bar, through Algeria, North Africa up to Tangiers, where it crosses near Gibraltar to Spain. Inside Portugal, the gas is distributed in the main network at a pressure of 70 bar. The gas is considered to reach the final customer at a pressure of 7 bar. The total distance is assumed to be 1250 km.Some natural gas is used as fuel in the intermediate compression stations, the overall efficiency is assumed to be 97,8% according to [3]. The natural gas vehicles of STCP store the natural gas onboard at a pressure of 200 bar, so a compression stage is needed. The final pressure is 250 bar in order to guarantee enough pressure differential to fill the tanks. The work needed for an isentropic compression can be expressed as
where:
k = adiabatic exponent n = number of moles T 1 = temperature at p 1 R = Perfect gas constant = 8,3 14 J/(mol.K) p l = inlet pressure p2 = output pressure (filling)
The isentropic efficiency of the process is considered to be 68 % [4]. The final energy consumption (electricity) for the filling process is 0,008 kWhel/ MJ of CNG. The global result for natural gas processing is presented on Table 11. Table 11. Energy consumption, losses and GHG emission for natural gas production.
Extraction of natural gas Processing of NG Transport of NG CompressiodFilling Total
Energy suuolied (MJMJNG) 1,065 1,059 1,052 1,029
(MJ/MJNG) 0,006 0,007 0,023 0,029
1,065
0,065
Energy losses
GHG emission ~
3.1.5. Comparison Well-to-Tank Table 12 presents the global results for fuel production.
%
9,2 10,8 35,4 44,6 100
(&/MJNG) 0,62 0,61 2,53 5,37 9,13
%
63
6,7 21,l 58,s 100
Environmental Impact of Hydrogen in Urban Transports
29 1
Table 12. Comparative results for fuel production.
3.2. Comparison Well-to-Tank
Emissions for the different vehicles (diesel and natural gas; hydrogen fuelled fuel cell vehicles emit no harmful emissions) were calculated using two tools, Copert [6] for Diesel vehicles and Mobile 6 [7]for CNG vehicles. Consumption, mileage values and vehicle characteristics where obtained directly from STCP. 3.2.1. Hydrogen Vehicles
For the hydrogen vehicles an autonomy of 200km was accepted with the consumption of 40 kg of hydrogen (reference value suggested by DaimlerChrysler, road test confirm this value as a good estimate) [9]. Final energy consumption is 23,9 MJIkm. 3.2.2. Diesel Vehicles For Diesel vehicles, the results for a Euro III vehicle are presented in Table 13. Table 13. Results for a Euro III diesel bus operating in regular service in porto.
Energy consumption
coz Nz0 CH4 COzequivalent
18,4 M J / h 1,34OE+03g/km 3,000E-02 g h 8,600E-02 gikm 1,37OE+03 g/km
3.2.3. CNG Vehicles
For CNG vehicles, the results are presented in Table 14. Table 14. Results for CNG bus operating in regular service in Porto
Energy consumption COZ N2O CH4
COz equivalent
22,9 M J h l,292E+O3 g/km 4,130E-02 gikm 6,561 g/km 1,4433+03 g/km
3.2.4. Vehicle Comparison Table 15 shows the comparison between the different technologies used in Porto.
292
K.R. Culhuu, G.A. Gonculves & T.L. Furius Table 15. Comparison between the different technologies.
Diesel (Euro 111) CNG Hydrogen
Energy consumption I8,4 22,9 23,9 (MJih)
GHG Emission 1,370 1,443 0 (kg C02 equivalentih)
3.3. Well-to-Wheel
Global results for the total life cycle of each combination of fuelhehicle technology are presented in Table 16. Table 16. Comparative Well-to-Wheel results.
Diesel CNG Hydrogen (electrolysis-grid electricity) Hydrogen (electrolysis-renewablc) Hydrogen (steam reformer-NG)
Energy Consumption MJkm 21,2 24,4 9 1,8 2,1
38
(102 emissions kg C02 e q u i v a l e n t h 1,612 1,65 1 1,441 0,388 2,478
In addition to energy consumption and GHG emission, values where calculated for the emission of pollutants (Table 17). It must be noted that the pollutant emissions from hydrogen vehicles occur exclusively in the production of hydrogen, and thus have minimal or virtually no local effect, while most of the emissions of conventional vehicles occur locally. Table 17. Well-to-Wheel emission of pollutants (kg/!un)
4. Conclusions
Since the early 19th century, scientists have recognized hydrogen as a potential source of fuel. Current uses of hydrogen are in industrial processes, rocket fuel, and spacecraft propulsion. With further research and development, this fuel could also serve as an alternative source of energy for several purposes and fueling motor vehicles. However it is an intermediate fuel, as there is no natural resource of hydrogen of sufficient dimension and intensity. It must be produced from other resources. When produced from renewable resources and technologies, such as hydro, solar, and wind energy, hydrogen becomes a renewable fuel.
Environmental Impact of Hydrogen in Urban Transports
293
The main result of this work shows that public transport by buses consuming hydrogen produced by conventional technologies does not as yet present a solution for the reduction of GHG emissions. Conventional technologies are still less polluting and more efficient, when the global process is appraised (life cycle analysis). This stems from the inefficiencies of the conventional processes for hydrogen production. However, hydrogen from renewable energy resources shows a clear benefit, but it must be noted that this solution seems to stumble on economic barriers, as renewable power generation does not stand alone, at least without subsidies. The production of hydrogen from grid electricity is clearly an intermediate solution, as it is currently the most practical way of producing low amounts of hydrogen, but in the long term it is not a sustainable strategy. The disadvantages associated with hydrogen can be greatly reduced if the local effect of the less polluting vehicle is taken into consideration. Whether this is enough to offset the disadvantages is a matter for further studies. The energy consumption of each vehicle reflects the technological development of each technology. Diesel and CNG vehicles (with a slight advantage for Diesel) are currently very efficient; the technology is well dominated but has currently reached a point where large efficiency increments are not expected. On the other side, fuel cell technology has the potential to be much more efficient, possibly double the actual values, and much experience will be gathered during the operation of the vehicles in Port0 for two years. This will help reduce the fuel consumption of hydrogen vehicles to a point where the extra energy used in the production of the fuel could be compensated.
Acknowledgments This project was developed in the framework of the CUTE project, supported by the European Commission. The authors would like to thank the support of STCP, DGTT and DG-TREN.
References [ 11 GaBi 4, Institute for Polymer Testing and Polymer Sciences, March 2003.
[2] Relatorio de Gestao e Contas 2002, Galp Energia. [3] GREET 1.5 - Transportation Fuel-Cycle Model, Argonne National Laboratory, August 1999. [4] GM Well-To-Wheel Analysis of Energy Use and Greenhouse Gas Emissions of Advanced FueWehicle Systems - A European Study, L-B-Systemtechnik GmbH, September 2002. [51 ERSE (Entidade Reguladora dos Serviqos Energeticos) “Caracterizacao do Sector Electric0 - Portugal Continental 200 1”. [6] COPERT I11 - European Environment Agency, November 2000. [7] MOBILE6, U.S. Environmental Protection Agency, January 2002 [S] Intergovernmental Panel on Climate Change, Climate Change 1995: The Science of Climate Change, Cambridge University Press, 1996. [9] http://www.fuel-cell-bus-club.com.
This page intentionally left blank
DEVOLATILISATION OF COAL-BIOMASS BLENDS AND COMBUSTION BEHAVIOUR OF THEIR CHARS P. CIAMBELLI, V. PALMA, P. RUSSO, S. VACCARO, V. VAIANO
Dipartimento di Ingegneria Chimica e Alimentare, Universita di Salerno, Fisciano (SA), Italy A bituminous coal, two biomasses and their blends were devolatilised at high heating rates (103-104Ws). The combustion of the chars was studied by air flow TPO tests while the dependence of the rate of combustion from the temperature was obtained through constant temperature tests in the range 673-823 K. Specific surface area (SSA) and Active Surface Area (ASA) measurements were also performed on fresh and partially burned-off chars. TPO tests evidenced that in some cases the minerals contained in different parent materials may interact each other enhancing or hindering the combustion process. Constant temperature combustion tests allowed the evaluation of the frequency factor and of the apparent activation energy for the char combustion. As carbon conversion increases, coal and blend chars SSA’s markedly increase, the ASA, from CO2 desorption, increases and the profile’s maximum moves towards higher temperatures while the ASA, from CO desorption, decreases and the profile’s maximum shifts at lower temperatures. The number of C-0 surface complexes strongly increases but this does not correspond to a sample reactivity increase.
1. Introduction The use of biomass as fuel in power plant technology attracts increasing interest because of its potentiality to lower CO2 emissions being, essentially, a nearly carbon dioxide neutral fuel. Indeed, investigations showed that the provision (including harvesting, transport and milling) of biomass for power generation produces only about 6 kg C02/GJ against 96.6 kg C02/GJ emitted by hard coal combustion in power generation [l]. However, the complete replacement of fossil fuels with biomasses is prohibitive at least for retrofitting existing plants because of problems occurring both during the preparation (transport, drying, milling, storage) and the combustion (ignition, slagging, burnout, emission problems, low calorific value) of such fuels [ 2 ] .Nonetheless, a promising shortterm option for the use of renewable fuels may be the coal-biomass co-combustion. This process allows the reduction of the consumption of fossil fuels and may be advantageously implemented in the existing coal-fuelled power plants with minor modifications [2]. In addition, it has been suggested that the co-firing of biomasses may help also in reducing NO, and SO, emissions at the exhausts of coal fired boilers [3-51. Keys parameters for the use of coal-biomass blends in existing plants are the devolatilisation behaviour of such blends, the quantitative assessment of the contributions of each component to volatile and char yield, and the role of possible interactions (of the organic and inorganic fractions) during the devolatilisation and/or the combustion stages. However, discrepancies may be found among results obtained by various authors [6].For coal-biomass mixtures, some studies [7, 81 have reported that the pyrolytic behaviour of blends, in any proportion, consisted of the additive behaviour of the two individual samples, while others [9] have reported some interaction between the two fuels during copyrolysis. Many authors have analyzed dynamic data and their apparent kinetics [ 10-131. Substantial differences in the values of reported kinetic parameters can be due to several 295
296
P. Ciambelli et al.
factors related to the experimental methods, operating conditions and data analysis, but also to the chemical composition of the raw materials investigated [6]. A further crucial point is to carry out experiments under operating conditions comparable to those found in industrial furnaces. Only in this way, indeed, experimental data may result suitable for supporting both devolatilisation and char combustion modelling as well for allowing the quantity of biomass in blends to be properly programmed to supply the suitable volatile matter release and to prevent possible undesirable effects due to the interaction between the fuels [14]. It was shown that laboratory-prepared chars, obtained under conditions milder (TG apparatus) than those encountered by coal in utility boilers, resulted more reactive to oxygen than chars from boilers [15]. A study concluded that the differences in reaction rates between residual utility boiler and laboratory chars could be explained by low intrinsic reactivities and that chars were deactivated in the boiler plant [16]. Further experiments using a hightemperature wire-mesh (HTWM) reactor [ 171 were able to show that thermal deactivation alone at realistic combustion particle temperatures of 1600-1800 "C and heating times up to 2 s would give chars with intrinsic reactivities as low as those found in utility boiler residual carbons [ 181. Complementary studies on residual carbons showed that, compared with the laboratory-generated chars, the residual carbons presented higher crystallinity [ 191 and suggested that a thermal deactivation process had occurred. Therefore, detailed char combustion models incorporating deactivation kinetics based on the HTWM reactor data and statistical distributions of particle properties could account for the persistence of unburnt carbon in the later stages of pulverized coal combustion [20]. In contrast, overall char combustion kinetics can satisfactorily model only the first 90% of burnout but seriously overpredict conversion rates in the later stages of combustion. This work aimed at studying the combustion behaviour of coal-biomass blends and at comparing it to those of the correspondent parent materials. To this end, 10 wt% biomass blends of a bituminous coal and cocoa shells and wood chips were prepared. The samples were pounded under 45 pm and devolatilised in a HTWM reactor. Their reactivity to oxygen was investigated by constant temperature tests in a flow reactor and by temperature programmed tests in a microbalance apparatus. The specific surface area and the active surface area of fresh and partially burned off samples were also determined. 2. Experimental Equipments and Procedures
2.1. Materials The materials employed in the experiments are listed in Table 1 where their ultimate and proximate analyses are also reported. They comprise a coal, in the following called coal K, and two biomasses, i.e. cocoa shells and wood chips. Fuel conditioning was accomplished by preliminary sample milling under similar conditions: samples were dried in a stove at 110°C for 12 hours and then pounded in an electric mortar to a size below 45 pm. After pounding, blends of coal K and biomasses at
Coal-Biomass Blends and Combustion of Chars
291
10 wt% of the latter were prepared. Finally, the samples were subjected to fast devolatilisation in the HTWM reactor. Table 1. Ultimate and proximate analysis of the employed materials. ~
Cocoa shells
Wood chips
wt Y'
wt %
carbon
71.43
4,05
71.43
hydrogen
4.47
6.00
4.47
sulphur
0.81
0.23
0.81 0.27
Ultimate analysis
wt %
clorine
0.27
0.01
nitrogen
1.12
2.64
1.12
oxygen
8.13
36.39
8.9 1
wt %
wt %
wt %
52.60
18.26 64.24 10.65
22.58
Proximate analysis fixed carbon volatiles
28.73
humidity
5.68
1.36 54.80
aqh
A number of measurements were carried out with the chars resulting from the devolatilisation in the HTWM reactor. They include specific and active surface area measurements, temperature programmed oxidation tests and isothermal combustion tests. 2.2. Samples Devolatilisation
A high temperature wire mesh (HTWM) reactor was employed in the experiments. It was designed, built and used to produce chars of the various samples to be tested. A scheme of the HTWM reactor and of the ancillary apparatuses is reported in Fig. 1. This device comprises a Pyrex container tightly fixed to a Perspex panel. Through the latter, two holes allowed the inlet and outlet of the washing gas (N2) and two brass rods, connected to the electric feeding device, allowed the transport of electric current to the brass electrodes at which the wire mesh was fixed. The sample was held on the wire mesh (or the sheet), which has the dual function of sample holder and heating element. Several materials were tested as candidate for the wire mesh. Among the materials having chemical and physical characteristics suitable for our purposes, the attention was focussed on graphite, molybdenum and tantalum. A number of physico-chemical properties were taken into account such as electrical conductivity, fusion temperature and their behaviour with respect to oxidising environment. Finally, graphite was chosen mainly because of its higher resistance to oxidising agents. The test procedure included the container cleaning, during which the sample (about 20 mg) was kept at room temperature for 15 min in nitrogen flow (99.999 % purity), then the test started by heating the sample at a given heating rate. The heating rate, the maximum
P. CiambeNi ei a[.
298
~ e ~ p eand r athe~ test ~ durat~onwere obtained by appropriate tuning of the regulation parameters o f the electrical feeding device. In most tests the sample was heated up to 1300 K and kept at this temperature for a time ranging from B to 4 s, In some m s with coal K heating was performed up to 1800 K. Heating rates from 1.1103 Us to B*184Ws were e mp~o~e d. Temperatures were measured by a two-color infrared ~ y r o m ~ t e ~ connected to a ~crocomputerfor feedback temperature contr~l.The particles were heated at the peak temperature and held for a specified time period before rapid, but uncontrolled, cooling.
Fig. 1. Scheme of the HTWM reactor apparatus.
2.3. Temperature Programmed Oxidation ~~~~~~~
A thermograv~me~c analyses (NETZSCH TC 209) was employed to perform TPO ( T e m p e ~ a ~Pro~rapntlled e Oxidation) tests for the rapid determination o f tbe overall sample reactivity. About 20 rng of sample was loaded in the sample holder then the ~emperaturewas raised in air flow at a rate of 10 Wmin from ambient to 900 K. The test yielded the current sample mass and its rate of change as a function o f both time and sample temperature. 2.4 Constant Temperature Oxidation
A Wow microreactor was employed to perform CTO (Constant Temperature Oxidation) tests in order to evaluate tbe kinetic parameters. A scheme of the reactor and of the ancillary apparatuses is reported in Fig. 2. The microreactor comprises a 300 m height, 20 pntll P.D. quartz tubular flow reactor, heated by an electrical furnace. A p r o ~ ~ e r controller (ASCON) controlled the reactor temperature. Cylinder air and nitrogen (99.999 96 pwity) were mixed downstream of mass flow controllers (HI-TECH) for
299
Coal-Biomass Blends and Combustion of Chars
obtaining different oxygen partial pressures in the gas entering the reactor. HARTMANN & BRAUN continuous analysers determined exhaust gas concentrations: URAS 10E (for carbon monoxide and carbon dioxide) and MAGNOS 6G (for oxygen). Signals from the analysers were sent to a personal computer for data processing. Samples to be tested were diluted with 180-350 pm quartz particles at a reacting matter to quartz mass ratio of 0.01 to avoid local rises of temperature during the tests. The diluted mixture was loaded to the reactor filling about 3 cm in length of its central zone. The remaining part of the reactor was filled with 700-2000 pm quartz particles. The test started raising the temperature to the desired value in nitrogen flow. Then the gas feed to the reactor was turned to the air stream starting the oxidation. Carbon burn-off was followed by monitoring, through the continuous analysers URAS 10E, the carbon mono and dioxide produced during gasification. Carbon mass balance was verified within 5% tolerance range for all tests. Kinetic experiments were carried out under constant operating conditions.
IGAS OUT/ HO ,
AIR
I
- 1
r
. . . . . ..
rnfc CONTROL
Fig. 2. Scheme of the flow reactor apparatus.
In all the runs parameters were kept at values such to allow the reactor working under differential conditions. The feeding gas flow rate was 500 cm3/min (STP). Operating pressure was 101 Wa. The temperature was investigated in the ranges 673-823 K. The initial mass of the sample was in the range 5-15 mg. 2.5. Temperature Programmed Desorption
The flow microreactor, shown in Fig. 2, was also employed for performing TPD (Temperature Programmed Desorption) tests in order to determine the amount and the rate of desorption of carbon surface complexes that allows the evaluation of the active surface area (ASA) of fresh and partially burned off samples. In this case the tests were carried out by rising up the temperature from 293 K to 1373 K at a rate of 20 Wmin in a
300
P. Ciambelli et a1
500 cm3/min (STP) nitrogen flow. The concentration of CO and CO, evolved from the carbon surface were measured on line with the mentioned NDIR analysers. Typical TPD runs were performed on fresh samples or on partially burned off samples. After a given oxidation time at fixed temperature and oxygen concentration, the reaction was frozen by lowering the temperature down to 473 K. When neither carbon monoxide nor carbon dioxide were detected by the analysers, the oxidising stream was replaced by the nitrogen stream and the temperature was raised up at the rate of 20 Wmin. TPD tests yielded the temperature profile of desorption of carbon-oxygen complexes from the char surface. The initial mass of char was about 200 mg. During the oxidation, oxygen molar fraction in the gas fed to the reactor was 0.1, but in any case the reactor was operated under differential conditions with respect to the oxygen. The oxidation rates as well as the conversion degree of carbon (X) in each oxidation step were calculated from the values of carbon oxides concentrations in the outlet gases. Carbon mass balance was verified in the range 95- 100%.
2.6. Specific Surface Area Measurements A Sorptomatic 1990 (ThermoQuest) was employed for performing specific surface area (SSA) measurements of samples through NZ adsorption tests and data elaboration by B.E.T. method. The experimental procedure included the sample conditioning at 363 K in He flow for 2 h, then the sample cooling at 77 K, and finally the gas flow was switched to NZ.SSA measurements were performed on both fresh and charred samples and, in the case of the latter, freshly prepared or after partial burn off.
3. Results and Discussion 3.1. Fast Pyrolysis Treatments The main results of sample devolatilisation in the HTWM reactor are reported in Table 2. In particular, for each sample the mass of volatiles released in the HTWM reactor, the char mass and the mass of residual ash yielded in the subsequent TPO experiment are listed as percentage with respect to the initial sample mass (on dry basis). In the same table the amounts of ash, volatile matter and fixed carbon obtained for each unblended component according to the ASTM standards are also reported. Comparing the percentage values obtained for each quantity with that achieved by standard analysis, it appears evident that the ash amounts are practically the same while the relative amounts of volatiles are markedly higher when measured through the HTWM reactor (Table 2). Correspondingly, the amount of char yielded results significantly lower than the relative quantity of fixed carbon.
3.2. Tests Performed with Raw Samples Preliminary tests of reactivity and combustion behaviour of the single fuels and of their blends were carried out on the uncharred samples. Such tests included TPO analysis
Coal-Biomass Blends and Combustion of Chavs
301
Table 2. Volatile yields, residual ash after TGA and estimated char yield. Volatiles loss at the HTWM reactor, %
I
Ash
Estimated char yield*, %
I
11
I
82
coal K
15.10
37
48
13.57
wood chips
2.95
90
7
2.16
14.37
44
cocoa shells-coal K ~
wood chips-coal K
.
I
I
42
Fixed carbon **, %
**, %
7.45
cocoa shells
Volatile matter
**, %
7.68
I I
71.89
20.43 55.71
1
1
21.55
I ~
41
45
*Char yield was estimated as difference between volatile yield and ash residue after TGA of chars **determined according to the ASTM standards
in the thermogravimetric apparatus and CTO tests in the flow microreactor. SSA measurements of the samples were also performed. The main results of tests carried out on uncharred samples yielded the following indications: a) the biomasses present double peak DTG profiles that reflect the presence of large amounts of volatiles (first peak). The coal, instead, present a single peak DTG profiles but, in this case a shoulder at the lower temperature end (about 600 K) of the peak is representative of the volatiles; b) cocoa shells and wood chips have combustion profiles very similar each other (DTG profiles have a main peak at 600 K and a secondary peak between 650 K and 850 K) and ash amount significantly lower than coal K; c) the measured values of the SSA of fresh blends made with coal K and biomasses differ significantly from the weighted average of those of the correspondent parent coal and biomass (Table 3). Table 3. Measured SSA (m2/s)of various samples of single materials, and measured and calculated SSA of blends before and after charring. Samnle
1I
1
wood chios
I
\\coal K (short charring times) II llcoal K
I I
I
I
I k K - c o c o a shells
1
Calculated
I
0.1 1
cocoa shells
coal K-cocoa shells (short charring times)
Measured
0.78 5.00 5.00
3.86
1
Calculated
I I I
I
4.50
I
4.58
4.50
3.08 0.21
I
20.39
I
30,30
1.06
0.57
I I
II
I
-1 I
1
I1I
13.34
I I I
1.98
coal K -wood chips
Measured
I
0.26
20.25 0.62
/I 1I)
P. Ciambelli et al.
302
3.3. TPO Tests Chars of the single fuels and of their blends, as obtained in the HTWM reactor, were subjected to tests of reactivity and combustion behaviour by TPO analysis in the thermogravimetric apparatus. The results are reported in Table 2 and Figs. 3 and 4. In particular, the residual weight after combustion is shown in Table 2. Fig. 3 reports the comparison between the combustion profiles of wood char, coal K char and their blend char while Fig. 4 shows the corresponding profiles for cocoa shells and coal K.
100 80
60 40
--+--
20
--+- wood char ---+--blend char
coal char
1
0 -1 ;. -2
-3
3
Pn
-4
-5 400
600
800 T, K
1000
1200
Fig. 3. Comparison between TPO profiles of chars of coal K, wood chips and their blend
TPO tests, not reported here, were also carried out with the uncharred samples. The main considerations that can be drawn from the comparison between TG and DTG profiles of fresh and charred samples are: a) the combustion profile of coal K char differs from that of the parent because the peak temperature of the char (810 K, Figs. 3 and 4) is shifted 60 K towards higher temperatures and there is the lack of a shoulder (shown by the fresh sample) at about 600 K, attributable to the presence of volatiles. b) The wood chips char shows one DTG peak at temperature (750 K; Fig. 3) correspondent to that of the higher temperature peak shown by the fresh sample, which also presents a lower temperature peak at 610 K. It is as though the loss of volatiles left the resulting char as reactive as the less reactive part of the fresh sample;
Coal-Biomass Blends and Combustion of Cham
303
c) The cocoa shells char shows one DTG peak at temperature intermediate (700 K; Fig. 4) with respect to those of the two peak shown by the fresh sample. It is as though the loss of volatiles made the resulting char more reactive than the less reactive part of the fresh sample.
coal char
+cocoa cha -+--
blend char
O
F
‘ 0
. Pn
-1 .L
E
-2
-3
-4 -5
600
400 ~
800 T, K ~
1000 ~
1200 ~
~~
Fig. 4. Comparison between TPO profiles of chars of coal K, coca shells and their blend.
With reference to the charred samples, Fig. 3 shows that the coal K-wood blend char results less reactive than both chars of the blend components. Given the much higher volatile yield of the biomass (Table 2) and the weight ratio of the two components in the blend (10 wt%), most of the blend char derives from coal. Therefore, it seems that the presence of the biomass inhibits coal char combustion. In contrast, the char of the cocoacoal K blend behaves as though the two components were burned independently (Fig. 4).
3.4. SSA Measurements Table 3 reports the results of SSA measurements performed with the fresh samples and with the samples after devolatilisation. The char’s SSA are much larger than that of the correspondent parent materials especially in the case of cocoa shells for which SSA increases of two order of magnitude (Table 3) while for the other single samples and blends the increase is only of one order of magnitude. However, the SSA’s of the blend’s
P. Ciambelli et al.
304
char are, normally, the weighted average of those of the correspondent parent coal and biomass chars (Table 3). With the uncharred samples the measured SSA’s of the blends are quite lower than that calculated on the basis of a weighted average contribution of the single components in the mixtures. In contrast, the opposite was observed with the chars of the same blends (Table 3) either when subjected to relatively short charring times (1 s) or when held in the HTWM reactor for longer times (4s). The evolution of SSA during sample burn off is shown in Fig. 5 where the measured SSA is reported as a function of carbon conversion (X) in the case of cocoa shells char and of the cocoa shells- coal K blend char. It appears that the specific surface area of cocoa shells char increases of a factor 3 as the carbon conversion changes from 0 to 60% with an almost constant slope. The specific surface area of the blend char, instead, increases much more (about 8 times) as X changes from 0 to 70%: This increase is very strong at the beginning of the combustion process (0<X<1OYO)then raises less markedly Up to x=70%. 250
I
I
II
I
11
I
II
I I1
200 -
-
dcocoa shells char
100
50 0
20
40 X
60
0 80
Fig. 5. Specific surface area of cocoa shells char and of a char of the blend coal K-cocoa shells as a function of the sample conversion (d.a.f.b.).
3.5. CTO Tests All the char samples, as obtained in the HTWM reactor, were tested in the flow microreactor under isothermal conditions at four different temperatures: 673, 723, 773 and 823 K. Typical results of such combustion tests, in the case of cocoa shells char, are reported in Fig. 6 as the rate of change of the carbon conversion degree dWdt (where X = (mo-m)/mo, m and mo being the current and the initial mass of carbon in the sample, respectively) as a function of X. During CTO tests the separate contributions of CO and C 0 2 formation to the carbon overall combustion were measured and the average values measured during each test are reported in Table 4 for all materials. In any case, the CO/CO2 ratios remain
305
Coal-Biomass Blends und Combustion of Chars
~ r a c ~ ~ constant c a ~ ~ yalong the whole carbon conversion and9 as the test ~ e ~ p changes from 673 to 823 K, slightly increases as expected 121-233. However, in the case of the blends chars the values are higher than those exhibited from the two blend md, in addition, first decrease and then increase (Table 4). This c o m ~ ~ n e nseparately ts behaviom is quite strange since, as shown in the literature [all, the ratio should increase with the temperature. However, it is known that some oxides and salts may play as catalysts f ~ the r oxidation o f c~bonaceousmaterials [23-251. F ~ h e ~ o rite has , been shown that the presence of catalysts during the carbonaceous material ~ x ~ d a t i opromotes n the formation of surface (C-Qa) complexes giving rise to CQ2 so influencing the value of the a ~ o of ~CO~andsCQ2evolved firom the carbonaceous surface 1231. 4
3 -._. t E
4 2
3 W
1 0
0
0.2
0.8
0.6
0.4
1
X
Fig. 6 . Comparison between CTO profiles o f chars of cocwd shells at: four different temperahues.
'Table 4. Average CO/C02ratios during the CTO tests at four different temperatures o f chars produced from various samples.
I I I coal K char
material
I 673
I
I
T. K
0.35
1 I
723
I
773
I
823
I
0.4
I
0.45
I
0.6
I
cocoa shells char
wood chips char _ I
coal K-cocoa shells blend char
Therefore, d w h g the blend combustion the interaction o f the i n Q ~ g matter ~ ~ c o f the blends c ~ m p o n ~may n ~ sgive rise to the formation, if not present, or to the ~ a ~ s f ~ ~ ~ ~ e ~present, j o u of s ~ substances y eventually acting as catalyst for combustion.
P. Ciambelli et al.
306
Constant temperature combustion tests allowed the determination of the kinetic parameters of the various char samples. All the evaluated kinetic parameters are summarised in Table 5. Here the values of the frequency factor (KO) and of the apparent activation energy (Ea), calculated at three values of X, are reported. The analysis of such results shows that: i) all the chars show an apparent activation energy that increases with carbon conversion; this could be ascribed to a possible non homogeneity of the chars that would lead to the faster combustion of the more reactive part and consequently to the slower combustion of the remaining less reactive part, or to a sort of annealing of the material because of the permanence at relatively high temperature; ii) cocoa and wood chars are very similar materials because their kinetic parameters are comparable; iii) coal char shows an apparent activation energy significantly higher than that of the biomass chars; iv) surprisingly, the apparent activation energy of the coal K-wood chips blend char is markedly higher than that of the coal K char although most (97-98%) of the blend char is actually coal char; v) this characteristic is not shown by the coal K-cocoa shells blend char, which also shows the higher constancy of the Ea with X; vi) cocoa and wood chars, obtained at lo3 Ws and lo4 Us,show comparable apparent activation energy being almost constant with the carbon conversion (X) and with the heating rate of charring. Table 5. Kinetic parameters of chars produced from carious samples. K0[min-1], Ea, [Kcal/mol].
X=0.3 G
I
X=0.5
Ea
Kr
I
X=0.7
Ea
I
Kn
I
Ea
Coal char Cocoa char Cocoa char lo4 Ms Wood char Wood char lo4 W s Coal K-cocoa blend char Coal K-wood blend char
It is worth noting that during the first tests for producing chars with the HTWM reactor very short sample residence times were used. That is 1 s when heating rates were lo3 Ws and 0.1 s when it was lo4 Ws. These times appeared enough for the biomasses charring but were insufficient for the coal K and coal K-derived blends. Indeed, such samples were characterised by low SSA (i.e. lower than that of the correspondent raw samples: see italic characters in Table 3) and activation energies and frequency factors strongly increasing with the carbon conversion, which give clear evidence for a bad charring process. Therefore, residence times of 4 seconds (1 s for heating and 3 for keeping the sample at the highest temperature) were employed when the sample contained coal K. Altogether, results suggest that a role is likely played by the interaction of the minerals in the blend otherwise it is quite difficult to justify the higher reactivity (lower Ea) of the blend’s char with respect to that of the chars of pure samples.
Coal-Biomass Blends and Combustion of Cham
307
3.6. TPD Tests The evolution of the ASA (active surface area) of the samples with carbon conversion was obtained by measuring the CO and COz desorption rate during the TPD tests of samples partially burned off at various carbon conversions. The raw data of desorption from a given sample, characterised by a given carbon conversion, were then normalised with respect to the SSA of the same sample. In this way for each carbon conversion of the sample we obtained the amount of CO and CO2 desorbed per unit sample surface. The results are shown as profiles of CO and COz desorption per unit sample surface in function of the desorption temperature in Figs. 7 and 8, respectively. 25
20
a
4
r 2
15
10 5 0
400
600
800
1000
1200
Fig. 7. Desorption of CO complexes per unit Specific surface area of char of the blend coal K-cocoa shells at different partial bum off as a function of desorption temperature. I
16
4
12
r€
8
(J?
2
4 0
400
600
800 T, K
1000
1200
Fig. 8. Desorption of COz complexes per unit SSA of char of the blend coal K-cocoa shells at different partial bum off as a hnction of desorption temperature.
308
P. Ciambelli et al.
Fig. 7 shows that when the carbon conversion changes from 0 to 26%, the CO desorption profile’s maximum increases slightly and moves from 1100 K to 900 K. Further increasing X gives rise to a decrease of the profile’s maximum that remains, however, at the same temperature (about 900 K). As a result the quantity of CO desorbed per unit sample surface area decreases but its desorption is easier since occurs at lower temperatures. At 1100 K where there was the maximum of the GO desorption profile at X= 0% there is the presence of a shoulder. Fig. 8 shows that as the carbon conversion changes from 0 to 26%, the C02 desorption profile maximum strongly increases and moves from 950 K to 900 K. Further increasing X gives rise to an increase of the profile’s maximum that shifts, however, again at about 950 K. As a result the quantity of GO2 desorbed per unit sample surface area increases.
Fig. 9. Comparison between TPO tests of coal K-cocoa shells blend chars after partial burning at X= 0%, 26%, 55% and 13%.
3.7. TPO of Partially Converted Samples The results of ASA measurements above suggest that the sample reactivity should increase with sample conversion as a consequence of the huge increase of SSA (Fig. 5) and of the substantial invariance, or increase, of the number of carbon-oxygen complexes per unit carbon surface able to give rise to CO and C 0 2 production (Figs. 7 and 8). However, this does not occur: This is evident from the results in Fig. 9 where the results of TPO tests, carried out at the same operating conditions with the four samples partially
Coal-Biomass Blends and Combustion of Chars
309
converted at different X, are shown. The weight loss and DTG profiles of the four samples are, indeed, practically the same but for what concerns the amount of ash, which, as expected, increases as the sample previous conversion increases. There is, therefore, an apparent insensitivity of the sample reactivity to changes in SSA and ASA occurring with carbon conversion. This aspect deserves further study. Acknowledgments
This work was partly funded by the European Commission in the framework of the Non Nuclear Energy Program JOULE IV (Contract No: ENK5- 1999-00004). References
[I] Ruediger E, Kircherer A, Greul U, Spliethoff H, Hein KRG. (1996) Two concept for a combined combustion of biomass with hard coal. Proc 21th Int Tech Conf Coal Utilit Fuel Syst: 557-568. [2] Biagini E, Lippi F, Petarca L, Tognotti L. (2002) Devolatilisation rate of biomasses and coal-biomass blend: An experimetal investigation. Fuel 81 (8): 1041-1050. [3] Pedersen LS, Nielsen HP, Kiil S, Hansen LA, Dam-Iohansen K, Kildsig F, Christensen J, Jespersen P. (1996) FuI1-scaIe co-firing of straw and coal. Fuel 75: 1584-1590. [4] Robinson AL, Iunker H, Buckley SG, Sclippa G, Baxter LL. (1998) Interactions between coal and biomass when cofiring. Proc 27th Int Symp Comb: 1351-1359. [5] Spliethoff H, Scheurer W, Hein KRG. (2000) Effect of co-combustion of sewage sludge and biomass on emissions and heavy metals behaviour. Process Safety Envivon Prot 78: 33-39. [6] Vamvukaa D, Kakaras E, Kastanakia E, Grammelis P. (2003) Pyrolysis characteristics and kinetics of biomass residuals mixtures with lignite. Fuel 82 (1 5-17): 1949-1960. [7] Klose W, Stuke V. (1993) Comparison of the pyrolysis of different types of biomass and coals. Fuel Process Techno1 36: 283-289. [8] Pan YG, Velo E, Puigjaner L. (1996) Pyrolysis of blends of biomass with poor coals. Fuel 75 (4): 412-418. [9] Sjostrom K, Chen G, Yu Q, Brage C, Rosen C. (1999) Promoted reactivity of char in co-gasification of biomass and coal: synergies in the thermochemical process. Fuel 78 (10): 1189-1194. [lo] Reina J, Velo E, Puigjaner L. (1998) Thermogravimetric study of the pyrolysis of waste wood. Thevrnochirn Acta 320 (1-2): 161-167. [ l l ] Gronli M, Antal Jr. MJ, Varhegyi G. (1999) A round-robin study of cellulose pyrolysis kinetics by thermogravimetry. Ind Eng Chem Res 38 (6): 2238-2244. [ 121 Orfao JJM, Antunes FJA, Figueiredo JL. (1999) Pyrolysis kinetics of lignocellulosic materials-three independent reactions model. Fuel 78 (3): 349-358.
310
P. Ciambelli et al.
[13] Perez MG, Chaala A, Yng J, Roay C. (2001) Co-pyrolysis of sugarcane bagasse with petroleum residue. Part I: thermogravimetric analysis. Fuel 80 (9): 12451258. [14] Russell NV, Gibbins JR, Man CK, Williamson J. (2000) Coal char thermal deactivation under pulverized fuel combustion conditions. Energy & Fuels 14: 883-888. [IS] Hurt RH, Davis KA, Yang NYC, Headley TJ, Gibbins JR. (1993) The combustion reactivity and physicochemical properties of residual carbon in fly ash from pulverized coal combustion. Proc 7th Int Conf Coal Sci: 241-244. [ 161 Hurt RH, Gibbins JR. (1995) Residual carbon from pulverized coal fired boilers: 1. Size distribution and combustion reactivity. Fuel 74 (4): 471-480. [ 171 Gibbins JR, Lockwood FC, Man CK, Williamson J, Hesselmann G, Downer BM, Skorupska NM. (1995) Coal selection for NOx reduction in pulverised fuel combustion. Proc 2nd Int Conf Comb Emiss Control. [18] Beeley TJ, Gibbins JR, Hurt RH, Man CK, Pendlebury K, Williamson JJ. (1994) Coal char reactivity characterisation for burn-out prediction in utility boilers. Am. ChemSoc., Div. Fuel Chem., Prepr. Pap. 39: 564. [19] Hurt RH, Davis KA, Yang NYC, Headley TJ, Mitchell GD. (1995) Residual carbon from pulverized-coal-fired boilers: 2. Morphology and physicochemical properties. Fuel 74 (9): 1297-1306. [20] Hurt RH, Lunden M, Brehob EG, Maloney DJ. (1996) Statistical kinetics for pulverized coal combustion. Proc 26th Int Symp Comb: 3 169-3177. [21] Du Z, Sarofim AF, Longwell JP, Tognotti L. (1991) The CO/C02 ratio in the products of the carbon-oxygen reaction. In J Lahaye & P Ehrburger (eds), Fundamental Issues in Control of Carbon GasGcation Reactivity, 192: 91-106. Kluwer Academic Publishers, Amsterdam. [22] Du Z, Sarofim AF, Longwell JP. (1991) Kinetic measurements and modeling of carbon oxidation. Energy & Fuels 5 : 2 14-221. [23] Ciambelli P, Palma V, Vaccaro S. (1993) Low temperature carbon particulate oxidation on a CuNIK supported catalyst. Catal Today 17: 71-78. [24] Ciambelli P, Di Pietro A, Palma V, Vaccaro S. (1993) Characterization of carbon black catalytic oxidation by TPD. Thermochim Acta 227: 19-26. [25] Walker Jr PL, Rusinko F, Austin LG. (1959) In DD Eley, PW Selwood & PB Weisz (eds), Advances in Catalysis 11. Academic Press, New York.
DUAL FUEL COMBUSTION AS A WAY OF INCREASING SUSTAINABILITY OF POWER PRODUCTION D.R. SCHNEIDER, Z. BOGDAN Department of Energy, Power Engineering and Environment, Faculty of Mechanical Engineering and Naval Architecture, University of Zagreb, Zagreb, Croatia In recent years a possibility of simultaneous combustion of different kinds of fuels has gained an increased attention regarding ecological aspects. By using combination of heavy fuel oil and natural gas the pollutant emissions could conform to the ecological norms in force. A second issue of using dual fuels in power plants is of economical importance regarding the prices and availability of supply of different fuels. The model that describes dual fuel combustion is based on the probability density function formulation (PDF). The main difference from the “classical” PDF model suited for the system comprising one fuel and one oxidizer is in addition of the second fuel. An analysis was conducted for a steam generator of the 210 MW Power Plant Sisak (Croatia). Five different burner configurations were investigated for each of the three steam generator operating modes. In the basic configuration all four burners were fed with the heavy fuel oil, which corresponded to the highest emissions of CO,, SOz, NO, and soot. By replacing the heavy fuel oil on some or all burners with the natural gas the pollutant emissions decreased.
1. Introduction It is well known that carbon dioxide contributes considerably to the greenhouse effect which causes global warming. Since combustion of fossil fuels is the primary source of greenhouse gases (GHG), Croatia like many other countries, is now trying to reduce GHG emissions by using cleaner fuels. Croatia has signed and ratified the United Nations Framework Convention on Climate Change (UNFCCC) [ 11 as an Annex I country. It has pledged in the Annex of the Kyoto Protocol [ 2 ] to reduce its GHG emissions by 5% from the pre-transition level by the time of the first budget period 2008-12. Compared to the energy demand collapse in some transitional countries like Russia and Ukraine, Croatia has passed through a relatively limited long term reduction of GHG emissions since 1990 because of higher efficiency of its pre-transition economy. It is expected that in case of business as usual scenario it will breach the Kyoto target in 2006 since the demand for energy will be high, especially as the income continues to rise, particularly in domestic use for heating, transportation and electricity. Therefore Croatia will have to envisage an energy strategy for clean energy use in power generation sector including a fuel switching from heavy fuel oil to natural gas. In recent years a possibility of simultaneous combustion of different kinds of fuels has gained an increased attention regarding ecological aspects. Combustion of fuels of lower quality like fuel oils and coal rich in sulphur and asphaltene, has as a consequence an increased emission of pollutants. On the other hand gaseous fuels like natural gas, methane or hydrogen are more ecologically friendly. Apart from the demands for GHG reduction due to the Kyoto agreement, the Republic of Croatia as a candidate for EU membership will soon face even stringer regulations considering the pollutants emissions 311
312
D.R. Schneider & Z. Bogdun
[3]. Using the combination of the above mentioned fuels the pollutant emissions could conform to the ecological norms in force. A second issue of using dual fuels in power plants is of economical and strategic importance. Nowadays, with the onset of the liberalization of fuel and electricity markets, it is often necessary to calculate with prices and availability of supply of different fuels. Not only increasing price of oil but also possibility of oil supply crises requires diversification of energy sources. Feasibility of combustion of two different types of fuel, alone or combined in certain proportion, allows the power plant management to set the best plant operation strategy. The Power Plant Sisak is one of the main power generation plants in Croatia. It comprises of two condensing turbine blocks of 210 MW each and it was commissioned in 1976. During its lifetime the main fuel was heavy fuel oil of low quality and natural gas when available. Last year it was decided that the old burners and auxiliary equipment would be replaced by the new M&S low-NOx burners that are capable of burning both heavy fuel oil and natural gas. The possibility of simultaneous combustion of two kinds of fuels, introduced into furnace in any ratio, will certainly change designed operating conditions in the furnace. In this paper, possibility of safe and efficient burning of such fuels has been investigated using numerical simulations, with the special emphasis on the pollutant emissions. 2. Modeling
A modeling of simultaneous combustion of two different types of fuel was always challenging task. Many of the present combustion models do not allow such an option. For that reason the CFD code FluentTM[4] was chosen due to capability of one of its combustion models to enable such simulation. The combustion model is based on the probability density function formulation (PDF). The main difference from the “classical” PDF model suited for the one fuel - one oxidizer system is in the addition of the second fuel. In that case the system is described by three mixture fractions, for the primary fuel,&,, for the secondary fuel andfoxfor the oxidizer:
The two mixture fractions fhel and A,, cannot be varied independently, as the relation below shows:
where psec is the normalized secondary mixture fraction and it is assumed that is statistically independent from the primary fuel mixture fraction. Unlike psecis not a conserved scalar.
xe,,
Dual Fuel Combustion as a Way of Power Production
313
The effects of the turbulent fluctuations of mean species concentrations, density and temperature are incorporated by using statistical approach of PDF [ 5 ] . The PDF combustion model involves the solution of the transport equations for the mean mixture fraction
7 and its variance f
l2
for primary and secondary fuel, respectively. The mean
(time-averaged) mixture fraction equations are given as:
(4)
ax,
where S, represents the source term due to transfer of mass into the gas phase from liquid fuel droplets. The variable pse, is calculated from Eq. (2). The mixture fraction variances are used in the closure model describing turbulence-chemistryinteractions:
where f ' = f
-
7 and C,,
C, and q are the model constants.
Fluctuations of mixture fraction f are represented by a beta-function. The assumption of equal diffusivities is acceptable for turbulent flows where turbulent convection is much higher than molecular diffusion. The reacting system is modeled by the equilibrium chemistry formulation. This formulation is based on the assumption that chemistry is fast enough to preserve chemical equilibrium at the molecular level. Such assumption is valid for main reactions and species in the combustion process but it fails in the case of slow kinetics like NO, and soot formation. The global 3D mathematical model of the furnace comprises coupled gas flow and liquid spray physics in the non-premixed turbulent flame. Transport of mass, momentum, energy and chemical species (only NO, and soot) in the gas phase is obtained by solving standard Favre-averaged Eulerian conservation equations. Turbulent flow is modeled by the realizable k-E model. The pressure field is calculated from the continuity equation using the PIS0 algorithm. The heavy fuel oil spray is described as a discrete second phase using the "particle-in-cell model". It is assumed that the particles are dispersed in
D.R. Schneider Cf. Z. Bogdan
314
continuo^ (gas) phase. Trajectories, mass, mom en^ and heat balance equations for the particles, i.e. fuel droplets, are solved in the Lagrangian Game of reference. The turbulent of droplets is predicted using the stochastic andom om walk" (or "eddy d~spe~sion lifetime") model. Tlhe radiation heat transfer is described with the discrete ordinates radiation model, which could account for particulate effects between gas and droplets. The a ~ s o ~ ~ coeEcient ion of the mixture is calculated as the w e ~ g h t e d ~ s ~ - o f - ~ e y gases. The soot formation is modeled on the basis of the simple single-step model of Khan and Greeves. The fomat~onof NO, is attributed to three reaction mechan~sms: themal (extended Zeldovich), prompt and fuel NO, mechanism. 3. ~~~~~~~~y
An analysis was conducted for a R m i n type steam generator P-56-1 built in the TPP Sis& Two steam ~enerato~s produce steam for a turbine comprising a 210 M W power block. The h a c e (Fig. 1) of the boiler has dimensions of 8.4~8.4~25.1m and enables thermal input of 300 MW. On each side-wall of the chamber, at height of approx. 5 rn, there are two burners. Newly installed low-NOx burners NAB 68430 (Fig. 2 ) consist of combust~onair swirl generating register, steam atomizer and gas lances,
flue gas exit
I -
Fig. 1. Discretization ofthe fitmace.
The airflow is divided into two stream: the primary stream that is swirled by passing through the flame holder and the secondary stream, which is swirled by the tangentially arranged blades. The oil preheated to 120 "C at 12.5 bar is atomized into fme misf by steam at 18 bar in so called 79-nozzles (eighth openings) forming a hollow cone in
Dual Fuel Combustion as a Way of Power Production
315
front of the atomizer. Natural gas at 1.2 bar is fed through eight gas lances. Each lance, having 21 nozzles, is rotated relatively to the previous one, so that the outgoing gas flow from the burner would gain rotational component. Tables 1 and 2 give the composition of natural gas and heavy fuel oil, respectively.
Fig. 2. Schematic diagram of the burner. Table 1. Heavy fuel oil composition (elemental analysis). mass%
vol%
C
85.02
37.74 61.64
H
11.65
N
0.17
0.06
0
0.17
0.06
S
3.00
0.50
Table 2. Natural gas composition.
96.59
4. Results
Five different burner configurations were investigated for each of the three steam generator operating modes (50%, 75% and 100% of load). All results are given for the
D.R. Schneider & Z. Bogdan
316
furnace exit (Fig. 1). Data are expressed in mg/m3 except for COz that is given in percentage of mole fractions. C 0 2 and SO2 emissions (Figs. 3, 7, 11 and 4, 8, 12, respectively) show expected trends. By replacing the fuel (heavy fuel oil) on one of the burners with natural gas the C 0 2 and SO2 emissions decreased. Further increase in natural gas contribution (two, three and finally all four burners fired by gas) continued similar trend. The main reason for this decrease is that natural gas contains less carbon than fuel oil and none of sulphur. One will question such a high SO2 concentration in the case of all burners fired by heavy fuel oil but, it is the consequence of the composition of the chosen fuel oil that contains 3% of sulphur (Table 1). Nowadays, legislation will force the TPP Sisak to use fuel oil with lower sulphur content. Nitrogen oxide (NO,) emissions are also reduced by further increase in natural gas contribution (Figs. 5, 9, 13) due to presence of nitrogen in the fuel oil. It is known that NO, formed by fuel NO, mechanism measures almost 80% of total NO, created in the oil flame. Calculated NO, emissions are higher than the actually measured in operating conditions. Therefore, data obtained by the NO, model used in the simulation should be viewed qualitatively i.e. trends should be considered and not the absolute values. Results of soot emission show descending trend with the higher ratio of natural gas in total amount of the burned fuel (Figs. 6, 10, 14). That is in accordance with the measurements in operating conditions where it could be established that the soot and particles emissions are higher in the oil flames than in the gas flames. Relatively high soot emission in the case where all burners are fired by fuel oil could be most likely explained by the fact that the soot model does not include complex processes of pyrolysis, polymerization, nucleation etc., which are present in soot formation. The soot model used in the simulation is the approximate one and it includes only empirical expressions for soot formation and oxidation. Therefore, similarly as in the case of NO, emissions, the soot results should be taken qualitatively and not quantitatively. 50% of Furnace Load
12 10 8
6
4
2
I
0 4gas
3gas+loil 2gas+2oil Igas+3oil
Fig. 3. C 0 2 emission (50% of load).
4oil
I
00+3'0
90-3'1 90-3'Z 90-3'E 90-3'P
90-35 90-3'9
0 002
OOP 009
008
0001
I
0
ooz OOP 009
008 000 L
ooz L
L I E317
D.R. Schneider & Z. Bogdan
318
75% of Furnace Load
12
10 8 6
4 2
I
n 4gas
3gas+loil 2gas+Zoil Igas+3oil
4oil
Fig. 7. COz emission (75% of load).
Ioon
.3nn 600
400
zoo 0
I
4gas
3gas+loil 2gas+2oil Igas+3oil
Fig. 8. SO2 emission (75% of load)
NO [rng/m3] 1800 1600 I400 1200 Inno 800
600
400 200
0
Fig. 9. NO emission (75% of load).
4oil
Dual Fuel Combustion as a Way of Power Production
Z5E-04 2,OE-04 1SE-04
1,OE-W
5,OE-05 O,OE+OO
Fig. 10. Soot emission (75% of load).
100% of Furnace Load co, [“h]
4gas
3gas+loil 2gas+2oil Igas+3oil
4oil
Fig. 11. CO2 emission (100% of load).
1000
800
600 400 200
I
0
4gas
3gas+loil Zgas+Zoil Igas+3oil
Fig. 12. SO2 emission (100% of load).
4oil
319
D.R. Schneider & Z. Bogdan
320
N O [mg/mj 2000 T1800 1600 1400 1200 1000 800 600 400 200 0
Fig. 13.NO emission (100% of load). soot [mglm? 7,E436
r
-
~
______
6,E-06 5,E-06 4,E-06 3,E-06 2,E-06 I,E-06
O,E+OO
Fig. 14. Soot emission (100% of load).
This paper focuses on the ecological aspect of using dual fuels. But it must not be forgotten that such combustion changes thermal and aerodynamic conditions in the furnace. Namely, regions of safe and efficient boiler operation must be established by considering heat fluxes transferred to the furnace walls and temperature/concentration fields in the furnace. So, the quality of combustion is investigated by analyzing local formation of soot and carbon monoxide as a product of incomplete combustion. That comprehensive analysis is not completely presented in this paper, but only some parts are given as an illustration, as shown on Fig. 15. Contours shown on this figure represent the wall heat flux (W/m2) for 100% of load. Strong asymmetly could be observed for the cases where two fuels are burned simultaneously. It is well known that gas flames have lower emissivity than oil flames. Therefore less heat is transferred to the furnace walls by the gas flames and gas temperature maximum (which is higher than in the oil flames) is shifted upwards, resulting in higher flue gas exit temperatures. Lower temperatures
Dual Fuel Combus~ionus u W q ofPower Production
32 1
released by oil flame can cause s t e m temperature, at the outlet of the end superheater, not to reach its nominal value. The problem could be solved by applying reci~cu~at~on of flue gases in the case of oil combustion. Recirculated Rue gases, which are taken &om the convective section o f the boiler (amomt and tem~er~ture o f which depends on boiler load and burner c o n ~ ~ a t i oare n ~introduced , through the bottom of the furnace (cold h e l ) which consequently lowers the maximal tempe~turein the furnace and shifts the furnace s flue ~ e m a x~ towards ~ ~ the ~~ h a c e~exit. Hence, ~ the ~ i f~f ~ r e ~ icne~ e gas exit temperatures for the oil, gas and their omb bin at ion§ become much smaller than without Rue gas recirculation.
$$$$$$ .......... . Fig. 15. Neat flux at the back wdf (100%of load).
D.R. Schneider & Z.Bogdan
322
5. Conclusions
In this paper the possibility of using dual fuels in power plants has been examined. The paper focused on the ecological aspect of using dual fuels, but the thermal and aerodynamic conditions in the furnace were also investigated in order to establish regions of safe and efficient boiler operation. The analysis was conducted for the steam generator of the 210 MW Power Plant Sisak (Croatia) in which old burners were replaced by the new low-NOx burners capable of burning both heavy fuel oil and natural gas simultaneously in any proportion. Five different burner configurations were investigated for each of the three steam generator operating modes (50%, 75% and 100% of load). In the basic configuration all four burners were fired by the heavy fuel oil, which gave the highest emissions of COz, SOZ, NO, and soot concentration. By replacing the fuel oil on some of the burners with the natural gas the pollutant emissions decreased. Therefore, dual fuel combustion could be seen as a way of increasing sustainability of power production. Not less important issue of using dual fuels is a possibility to set the best power plant operation strategy considering prices and availability of fuel supply. Acknowledgments
The authors would like to thank to the Croatian Electricity Utility (HEP) for supplying all necessary information regarding modeling of the Sisak TPP. Nomenclature
C
f k
P S U X
&
0
P
P
model constants mixture fraction turbulent kinetic energy normalized mixture fraction source term of species i, kg/m3.s velocity, d s spatial coordinate, m dissipation rate of turbulent kinetic energy Schmidt-Prandtl number density, kg/m3 dynamic viscosity, Pa.s
Subscripts fuel primary fuel; heavy fuel oil ox oxidizer sec secondary fuel; natural gas
Dual Fuel Combustion as a Way of Power Production
323
References [ 11 United Nations Framework Convention on Climate Change. Published on: http:llwww .unfccc.de/. [2] Kyoto Protocol to the United Nations Framework Convention on Climate Change. Published on: http:llwww.unfccc.de/. [3] EC directives on stationary source emissions. Published on:
http://europa.eu.intcomm/environment/air/stationary. htm. [4] (1998) FLUENT 5 User’s Guide. Fluent Inc., Lebanon, NH. [5] Peters N. (2000) Turbulent Combustion. Cambridge University Press.
This page intentionally left blank
HYDROGEN PRODUCTION BY ALLOTHERMAL REFORMING OF ETHANOL FOR FUEL CELL APPLICATIONS. PROTOTYPE DEVELOPMENT A.J. MARIN NETO('J), E.P. SILVA(',~),J.C. CAMARGO(',~),N.P. NEVES m('a2), C.S. PINTO Hydrogen Laboratory, Campinas State Universi& Campinas, Brazil ('I Brazilian Reference Center for Hydrogen Energy (CENEH), Campinas, Brazil (Ii
This paper describes preliminary results obtained with the allothermal ethanol reforming system for synthesis gas (syn-gas) production and hydrogen upgrading and purification for he1 cell applications. The system was designed to supply hydrogen to a 500 W PEM (Proton Exchange Membrane) fuel cell, with an electrical efficiency of 45%, which requires approximately 0.45 m3.h-' of hydrogen, with a maximum carbon monoxide concentration of 20 pmol.mol-' (ppm). The study was performed changing the operation temperature and analyzing the resulting syn-gas through gas chromatography for a specific catalyst. This catalyst was tested up to 700"C, 1 bar and fixed stoichiometric steam to carbon ratio. The syn-gas, before carbon monoxide shift reactor implementation, was submitted to a two-bed-three-segments purification step composed of chemical and physical molecular sieves for hydrogen purification. The carbon monoxide shift reactor (water gas shift reactor) is under development to improve the efficiency-to-hydrogen and maximize the life of the purification bed. The final results also include a discussion about possible reactions involved in ethanol steam-reforming for such catalyst.
1. Motivation for Ethanol Appliances
The use of ethanol for electricity generation is an attractive opportunity for sugar-cane agricultural-based countries mainly considering its use with fuel cell systems. The advantages of sugar-cane ethanol are superior to those of the corn ethanol due to the higher ratio output energy (ethanol heat capacity) to input energy (consumed during cultivation and fuel processing). For sugar-cane ethanol, this ratio is about 7, whereas for corn, it is 1.3. Another advantage for ethanol utilization is its higher energy density per volume when compared to hydrogen. This suggests an interesting option for hydrogen storage to be used in fuel cell electric vehicles. This solution is directly applicable for countries where the fuel distribution logistics and refueling infra-structure is consolidated, as it is in Brazil. Considering the volume of fuel storage, the energy gain with the hydrated ethanol (ethanol fuel for internal combustion engines - ICE) storage is about 11.5 times as much as that of hydrogen at 200 bar. The storage of ethanol on pure or hybrid fuel cell vehicles also increases the system safety since it is less likely to leak compared to hydrogen, the fuel is diluted and it shows less toxicity compared to other hydrocarbon fuels. Concerning the environmental aspects, the use of bio-ethanol (alcohol from biomass) brings advantages due to its by-products sustainability. Regarding to the overall reaction (Reaction l), the carbon dioxide produced is re-absorbed by the sugar cane in the subsequent crop, since it is part of the photosynthesis process and the carbon cycle. In fact, this renewable hydrogen source can decisively contribute towards the hydrogen economy insertion in the developing countries, which are the major sugar-cane suppliers in the world, and to increase the trade opportunities for them. 325
326
A.J. Marin Net0 et al.
2. Ethanol Processing
There are essentially three different processes to reform hydrocarbon fuels: 1. Steam-reforming; 2. Autothermal reforming; 3. Partial Oxidation. Applying these for ethanol, the follow reactions are reached:
The Eqs. (1-3) correspond to each global process and only lead to a first approximation. The experimental reactions for steam-reformingcase will be presented later.
2.1. Steam-Reforming The steam-reforming consists in the catalyzed reaction of the proposed fuel with water, situation remarked by large energy consumption. It is important to define all parameters at this time. Selectivity is the capacity of the catalyst to promote formation of a specific product at a well-defined thermodynamic condition. It can be defined as the molar fraction of one product at one specific condition. Another important parameter is the catalyst activity. It could be defined as the catalyst capability of converting the fuel. The catalyst is generally the central focus of most studies in order to promote the reaction at lower temperatures (usually between 350-700°C), and to minimize systems’ dimensions and costs related to construction materials [ 11. Frequently, the desired product is hydrogen [2-131, but catalysts could be developed for higher selectivity to yield other compounds. 2.2. Autothermal Reforming In autothermal reforming two different reactions are coupled: the exothermic reaction, represented by the fuel oxidation, and the endothermic reaction, represented by the steamreforming. Generally, the oxidation reaction, which is exothermic, releases much more energy than the steam-reforming reaction, which is endothermic, absorbs. This implies that less fuel is needed to be oxidized than reformatted. The small amount of fuel which is oxidized generates heat to the fuel that will be reformatted. The temperature is similar to that of steam-reformingprocess.
Hydrogen Production by Allothermal Reforming of Ethanol
327
2.3. Partial Oxidation
A sub-stoichiometric quantity of oxygen, or air, is used in the partial oxidation fuel reaction. This small amount of oxidant produces some heat and water that reacts with the non-oxidized fuel to generate hydrogen and carbon dioxide, considering theoretical approach showed in Eq. (3). However, practical systems, based on partial oxidation technology, generate syngas with less hydrogen than other methods such as autothermal or steam-reforming. A large amount of carbon monoxide is yielded. The process temperature is also higher, around 1000°C. 2.4. Fuel Cell System Requirements
Syn-gases are composed usually by hydrogen (H2),carbon dioxide (C02), methane (CH4) and carbon monoxide (CO). For ethanol processing many other compounds are obtained. The possible products could be theoretically predicted based on standard heats of formation, providing that the catalysts are not active for by-products and intermediary compounds. Literature [2-131 reports the existence of 12 products. Among them are the four previously mentioned, plus ethene (C2H4), dietil-ether ( C ~ H ~ O C ~ Hethane S ) , (C~HS), acetaldehyde (CH3CHO), acetic acid (CH3COOH), and carbon (C). Non-reacted ethanol and water could be also found. The products in liquid phase at ambient temperature and atmospheric or low pressure could be easily removed by condensation, which could be achieved by transferring the thermal energy of the syn-gas to heat any desired part of the fuel processor. Gaseous products require a purification step. The purification could be started in the fuel processing reactor, in the case of membrane reactors, or in conventional purification beds, in the case of a system based on a fluidized bed reactor. For PEM (Proton Exchange Membrane) fuel cell systems, which is the present case, compounds such as CH4 and C 0 2promote mainly the dilution of the hydrogen-fuel inside the fuel cell. This dilution promotes a reversible decay in electrical performance, predicted by Nerst equation. Carbon monoxide interacts irreversibly with the PEM electro-catalysts, due the low temperature of this type of fuel cell (usually between 6O-8O0C), promoting an irreversible loss in electrical performance [ 141. This is the problem number one in a PEM fuel cell system since the commercial stacks do not admit CO concentrations above 20 pmol.mo1'. One solution for the CO high concentration in the syn-gas is the implementation of high andor low temperature shift reactor [15-171 in the system before the purification step. This component generally ensures better gas quality. These reactors promote, catalytically, a single exchange reaction between CO and water to produce additional H2 and C02.
328
A .J. Marin Net0 et al.
3. Experimental
This work reports the development of a test bench facility for ethanol processing, H2 upgrading, purification and demonstration of fuel cell integration feasibility. This system is composed by one allothermal steam-reforming reactor heated by an electronically controlled electrical furnace, a low temperature shift reactor for CO depletion, and purification step. Ethanol and water mixtures are pumped by a manually controlled electrical pump. Many sampling points are located in the overall system in order to make possible to analyze the efficiency of all components in any part of the system. The apparatus scheme is presented in Fig. 1.
Reservoir with Level ShiFtReactor
I
PEMFC t Air
Reformlng R
PUMlCatlD"
II l IIl
++--Cooling
0
Ethanol +Water
-Syn-Gas - - Hydrogen -Cooling Water
H
Valves Temperature Controller ~
Fig. 1. Test bench for ethanol processing system evaluation.
The purification step begins with a one-pass water cooled heat exchanger, followed by a trap to retain the liquid phase (produced or non-processed) after ethanol and CO processing. The partially dried syn-gas goes to the 2-bed 3-segment treatment for chemical water drying, C 0 2 and CH4 separation. Finally, CO separation occurs exclusively in the second purification bed. There are several sampling points in the system to collect gas samples to be analyzed. Determining the gas composition is useful to verify theoretical approaches for catalyst and purification elements and also to define the lifetime of the purification beds in order to implement an automated gas purifier based on PSA (Pressure Swing Adsorption) and TSA (Temperature Swing Adsorption) hybrid cycle. The integration of the prototype to a PEM fuel cell is only intended to demonstrate the technical feasibility. Since the system is based on an electrical furnace and considering the energy losses involved, the assertion that this assembly corresponds to an electrical generator based on ethanol fuel using a thermo-electrochemical route, is not
Hydrogen Production by Allothermal Reforming of Ethanol
329
true. But this system makes it possible to find many necessary parameters to develop an autothermal ethanol processor, whose project is under way. This new processor is illustrated in Fig. 2
Fig. 2. Autothermal ethanol processor.
The first evaluation of the assembly showed in Fig. 1 should be seen as the performance of the system rather than the catalyst performance itself. The CuNi / y - A1203catalyst was previously studied with respect to its structure and composition for different preparations and thermal treatments [4-61. In this work, the catalyst was evaluated at higher temperatures (400-700°C) and atmospheric pressure. The results are presented ahead.
3.1. Catalyst Preparation The catalyst was prepared by the impregnation method using hexa-hydrated nickel nitrate (PA-ECIBRA) and tri-hydrated cupper I1 nitrate (PA-ECIBRA), as precursors of the “active elements”, and distilled water for aqueous solution preparation. Pellets of yalumina (OXITENO, 200 m2.g-’, 0.04 m3.g-’,and pellets of 1 cm length) were used to support the metals. The impregnated support (6% Cu and 6% Ni, wlw) was dried in air at 100°C during 12 h and calcinated at 500°C during 3 h. The catalyst was not reduced before testing and the data with respect to catalyst performance was acquired after 2 h of ethanol processing. A stoichiometric mixture of ethanol and water (1 :3) was chosen as the first approximation for process evaluation.
330
A .J. Murin Net0 et ul.
3.2. Catalyst and System Performance in Ethanol Steam-Reforming Efficiency for hydrogen production can be defined as the product between the selectivity for its substance, at a specific thermal condition, and the fuel conversion (activity of the catalyst for processing the fuel). This maximum value occurred at approximately 600"C, and was better analyzed. The catalyst behavior is illustrated in Fig. 3.
I
590
'
I
600
'
I
610
'
I
'
620
I
630
'
I
'
MO
Temperature ("C)
Fig. 3. NiCu / y - A1203 performance on steam-reforming of ethanol.
The distribution of products coming from the steam-reforming of ethanol was obtained by analyzing the collected samples by gas chromatography. For high concentration syn-gas, the analysis was carried out with a TCD (Thermal Conductivity Detector), a 3.6 m Porapak N column (80-100 mesh) followed by another 0.90 m molecular sieve 5A column (60-80 mesh). Argon was the carrier gas. The tests were carried out with the reactor operating under temperature fluctuations not superior to 4°C. The mixture was supplied at 40 m1.h'' and the fuel conversion was always superior to 93%, for temperatures between 590 and 640°C. The H2 originated from this mixture flow would provide 57 W (electrical) if used in a PEM fuel cell with an electrical efficiency of 50%. The optimal temperature for H2 production was 613°C which corresponds to 5.38 mol of H2 yielded per mol of ethanol processed. This means a H2 production efficiency of approximately 89.7%. A preliminary analysis of the results of ethanol steam-reforming product distribution at 6 13°C leads to the occurrence of ethanol decomposition followed by methane steam-reforming and the promotion of a single exchange reaction between CO and water. This mechanism can be represented as follows:
Hydrogen Production by Allothermal Reforming of Ethanol
331
C~HSOH + CH4 + CO + H2
(4)
CH4 + H20 + 3H2 + CO
(5)
CO + H20 + H2 + CO2
(6)
At 598"C, very close to optimum, the promotion of methane steam-reforming is highly unfavorable what implies in the decrease of HZ concentration. This situation could be also characterized by a small CO methanation: 2CO + 2H2 + CH4 + COZ
(7)
The syn-gas was conducted to the purification step. The result after 16 minutes of operation is presented in Fig. 4.
650
--
E9
-
E,
Y
640 A
m
87
*0°630 150-
.-
0 ._ c
E c
8
co
100-
C
u,
i
U
620
$ z
R
v
610
600 0
2
4
6
8
10
12
14
16
Time (minutes)
Fig. 4. Performance of purification without water gas shift reaction.
The purification system is composed of two columns. The first contains silica-gel (Synth, spheres of 1-4 mm, PA) as the dehydrating agent for chemical water removal, and activated carbon (Synth, dispersed particles of 1-2 mm) for CH4 and COz removal, preferentially. The second column is mainly dedicated to CO removal and is composed of 5A zeolites (Baylith-Bayer, spheres of 1-4 mm). The quantity of each element on the columns is not equal, but proportional to the importance of each contaminant.
332
A.J. Marin Net0 et al.
Accordingly to the data illustrated in Fig. 4, the H2 purity was approximately 99.98% with a maximum 25 pmol.mol-’ CO concentration which was measured up to 16 minutes. The results show large potential for practical applications, even without a water gas shift reactor (WGSR) for CO depletion. The CO shift reactor is being developed based on Cu-ZnO / y - A1203 dispersed ceramic catalyst. This catalyst will be obtained from nitrate species and must contain 7% w/w of each active metallic element. At least 85% of activity in water gas shift reaction is expected, with no depletion of H2 concentration in the syn-gas. This characteristic would provide, at least, an increase of 85% in the CO-purification life-time. The new system, showed in Fig. 2, collapses the purification unit for one single column. It could be duplicated to perform a “push-pull” operation in practical systems. 4. Conclusions
The present work reports the development and initial results of an ethanol steamreforming processor and the ancillary systems to produce pure hydrogen for PEM fuel cell applications. The initial results were obtained for NiCu / y-A120s steam-reforming catalyst. The best selectivity for hydrogen was around 68% at 613°C and catalyst activity was always superior to 93%, which resulted in 5.38 mol of H2 per mol of ethanol processed. The catalyst longevity analysis was not reported yet. The hydrogen purity was about 99.98%, for an interval of 16 minutes of operation with carbon monoxide concentration remaining below 25 pnol.mol-’. The purification was carried out by physical and chemical adsorption processes by means of molecular sieves. The results achieved in this work will be used to develop an autothermal ethanol reforming system, initially designed to provide 5 kW when coupled to a PEM fuel cell. The advantages and opportunities of using bio-ethanol from sugar-cane for electricity generation in select applications were also discussed.
Acknowledgments The authors thank to Fundagao de Amparo a Pesquisa do Estado de Sao Paulo-FAPESP, to CENPES-PETROBRAS and FINEP for the financial support of the current projects. The authors also thank Conselho Nacional de Pesquisa e Desenvolvimento-CNPq and the Brazilian Ministry of Mines and Energy for the financial support. The acknowledgments are extended to Ana Carolina Sanches Zeferino and Miriam Camila Garcia Lima for the precious collaboration in the syn-gas chromatographic analysis.
References [l] Duane MB, Ariff GD, James BD, Lettow JS, Thomas CE, Kuhn RC. (2002) Cost and performance comparison of stationary hydrogen fueling appliances: Task 2 Report. DirectedTechnologies,Inc..
Hydrogen Production by Allothermal Reforming of Ethanol
333
[2] Hubner P. (2001) Research on the reforming of ethanol. Proceedings o f 3 r d L A M E T Workshop, Brasil. [3] Breen JF', Burch R, Coleman HM. (2002) Metal-catalysed steam reforming of ethanol in the production of hydrogen for fuel cell applications. Applied Catalysis B 39: 65-74. [4] Marino F, Boveri M, Baronetti G, Laborde M. (2001) Hydrogen production from steam reforming of bioethanol using CulNilWy-AI203 catalysts. Effect of Ni. International Journal of Hydrogan Energy (IJHE) 26: 665-668. [5] Marino F, Baronetti G, Jobbagy M, Laborde M. (2003) Cu-Ni-Kly-Al203 supported catalysts for ethanol steam reforming. Formation of hydrotalcite-type compounds as a result of a metal-support interaction. Applied Catalysis A 238: 41-54. [6] Mariiio F, Cerella EG, Duhalde S, Jobbagy M, Laborde M. (1998) Hydrogen from steam reforming of ethanol. Characterization and performance of cupper-nickel supported catalysts. IJHE 23: 1095-1101. [7] Galvita W. et al. (2001) Synthesis gas production by steam reforming of ethanol. Applied Catalysis A 220: 123-127. [8] Segal SR. et al. (2003) Catalytic decomposition of alcohols, including ethanol, for in situ H2 generation in a fuel stream using a layered double hydroxide-derived catalysts. Applied Catalysis A 8563: 1- 13. [9] Diagne C, Idriss H, Kiennemann A. (2002) Hydrogen production by ethanol reforming over Rh I Ce02-Zr02 catalysts. Catalysis Communications 3: 565-571. [lo] Cavallaro S, Freni S. (1996) Ethanol steam reforming in a molten carbonate fuel cell. A preliminary kinetic investigation.IJHE 21: 465-469. [ l l ] Freni S, Maggio G, Cavallaro S. (1996) Ethanol steam reforming in a molten carbonate fuel cell: a thermodynamic approach. Journal of Power Sources (JPS) 62: 61-13. [ 121 Freni S, Cavallaro S, Mondello N, Spadaro L, Frusteri F. (2002) Steam reforming of ethanol on NiMgO catalysts: H2 production for MCFC. Journal of Power Sources (JPS) 108: 53-57. [13] Batista MS. et al. (2003) Characterization of the activity and stability of supported cobalt catalysts for the steam reforming of ethanol. Journal of Power Sources (JPS) 124: 99-103. [14] Zhigang Q, Chunzhi H, Kaufman A. (2002) Effect of CO in the anode fuel on the performance of PEM fuel cell cathode. Journal of Power Sources (JPS) 111: 239247. [I51 Tanaka Y . et al. (2003) CO removal from reformed fuel over CdZnOIA1203 catalysts prepared by impregnation and coprecipitation methods. Applied Catalysis A: General 238: 11-18. [16] Tanaka Y . et al. (2003) Water gas shift reaction over Cu-mixed oxides for CO removal from the reformed fuels. Applied Catalysis A: General 242: 287-295. [ 171 Utaka T, Sekizawa K, Eguchi K. (2003) CO removal by oxygen-assisted water gas shift reaction over supported Cu catalysts. Applied Catalysis A: General 194-195: 21-26.
This page intentionally left blank
HYDROGEN PRODUCTION FROM STEAM OR AUTOTHERMAL REFORMING OF ETHANOL FOR LOW- AND HIGH-TEMPERATURE FUEL CELL V. GALVITA Max-Planck-Institutefuer Dynarnik Kornplexer Tehnischer Systerne, Sandtorstrasse 1, 39106, Magdeburg, Germany In the present work the reaction of ethanol steam or autothermal reforming for hydrogen rich gas streams production, over bimetallic catalysts Cu (5wt. %&Pd(lwt.% )/C and Cu(5wt. %)-Pd(lwt.%)/CeOz-Zr02,and over Pd(l wt.%)/C was investigated. In particular, the dependence of the catalytic activity and selectivity on reaction temperature, H20/C2H50Hmolar ratio and contact time was studied. In order to evaluate the catalytic stability long-term experiments were also performed. It was found that the bimetallic catalyst being essentially more active then the catalyst containing Pd only. Thus, the use of the bifunctional, bimetallic catalyst as the catalyst for the first layer of the two-layer reactor for ethanol decomposition proved to be quite promising for the steam reforming of ethanol to syngas. In the two-layer fixed-bed catalytic reactor ethanol is first converted to a mixture of methane, carbon oxides and hydrogen over a Pd-based catalyst and then this mixture is converted to syngas over a Ni-based catalyst for methane steam or autothermal reforming. It has been shown that the use of the two-layer fixed-bed reactor prevents coke formation and provides the syngas yield closed to equilibrium. For autothermal reforming of methane the Ni (10 wt.%) catalyst supported over CeOz-Zr02CaO/Alz03was used. This catalyst showed high activity and stability for autothermal reforming of methane.
1. Introduction
The development of the energy production at the 21St of century must be focuses for highly efficient, environmentally friendly energy devices. Fuel cell technology holds the promise to produce electricity at vehicles and stationary application from a wide range of fuels with high efficiency. On the basic of the electrolyte employed, there are five types of fuel cells, PEMFC, AFC, PAFC, MCFC, and SOFC. Each of them has advantages and disadvantages relative to each other. Hydrogen-rich gas for feeding PEMFC may contain only trace amounts of CO (>lOppm) to avoid poisoning of the electrocatalyst of the FC. Internal reforming and electrochemical conversion of CO (CO is produced after reforming of hydrocarbon fuel) is only possible with high-temperature SOFC and MCFC and this essentially distinguishes them from low-temperature FC. Currently, the main fuel for all of these types of FC is hydrogen. Chemical storage of hydrogen in liquid fuels is considered to be one of the most advantageous way for supplying hydrogen to fuel cells. A variety of liquid fuels, such as methanol, ethanol, and hydrocarbon are suitable for this purpose. Among liquid fuels, ethanol is a promising source of hydrogen while it is an easily accessible and non-polluting raw material which can be readily produced from renewable sources by fermenting the sugars founding grains, such as corn and wheat, as well as potato wastes, cheese whey, corn fiber, rise straw, sawdust, urban wastes, and yard clippings. Hydrogen production from non-fossil, renewable energy sources such as ethanol to became technically, economically an ecologically important for sustainable development in the 2 1St. century. 335
V. Galvita
336
Thermodynamic aspects of ethanol steam reforming have attention in the literature [ 11. The overall reaction of hydrogen production from ethanol is strongly endotermic and corresponds to the formation of 6 mol of H2 per mol of ethanol: C2H50H + 3H20 = 2C02 + 6H2
(1)
However, other undesirable products such as CO, CH4, CHlCHO, and C2H4 are also usually formed during reaction. In addition, the formation of coke on the surface of the catalyst is also not uncommon. Coke formation may occur by Boudouard reaction: 2co+
coz + c
(4
Another possible rote for the carbon formation is through ethylene. At, present, two processes for ethanol SR (step (i)) are available, namely a onestep process on supported metal catalysts [2, 31 and two-step process in a two-layer reactor [4] (Fig. 1). The layer in such reactor are made of different type of catalysts which working at different operating temperature. Within the first layer ethanol is converted into a mixture of CH4, CO, and H2 [4]. Over the second catalyst layer, the products of the ethanol decomposition are converted by autothermal reforming to a gas mixture enrich with hydrogen Fig. 1.
CZHSOH ~~0 Catalyst Layer 1 H~ Ethanol decomposition ,
..
HTFuelCeU
Catalyst Layer 2 Autothermal Reforming
4 I(02) Air
coz H2
..
b
i HT Fuel Cell
Fig. 1. Scheme of a two-layer catalytic reactor for ethanol conversion into a hydrogen-rich gas mixture.
2. Experimental 2.1. Catalysts Preparation The catalyst Pd (1 wt.%) supported on Sibunit catalyst (PdC) prepared as described in [5] was used in the present study. The catalyst consisted of sphere particles of size 1.2-1.5 mm, pore volume and the BET surface area were 0.74 cm3/g and 400 m2/g, respectively. The CeOz-ZrOz was prepared via urea hydrolysis. The ceria and zirconia oxide samples was prepared from Ce(N03)3 and ZrO(NO&. The starting metal salt was dissolved in distilled water to the desired concentration 0.1 M. Then, the mixed metal salt solution will be added to a 0.4 M urea solution with the salt to urea solution ratio of 2: 1,
Hydrogen Production from Ethanol for Fuel Cell
337
and the mixture will be kept 48 h at 100 OC. The sample will be then allowed to cool to room temperature prior to being centrifuged to separate a gel product from the solution. The gel product will be washed with ethanol and dried for 12 h in an oven at 105 "C. Then, the product was calcinated at 400 "C for 4 h. The catalyst consisted of sphere particles of size 1-1.2 mm, pore volume and the BET surface area were 0.44 cm3/g and 120 m2/g,respectively. Catalysts containing 5wt.% Cu and 1 wt.% Pd will be prepared by incipient wet impregnation of Ce02-Zr02(or Pd/C for Cu-Pd/C catalyst) with an aqueous solution of Cu nitrate and Pd nitrate. The samples obtained were dried for 12 h in an oven at 105OC and then heated at 400 "C in air for Ce02-Zr02and in He flow for Pd/C. The CeOZ-ZrO2-CaO/Al2O3was prepared by the impregnation method. Commercial alumina granules were immersed in the solution containing metal components. The samples obtained were dned for 12 h in an oven at 105OC and then heated 6h.at 1000°C in air. Ni loading was carrying out by the incipient wet impregnation of Ce0z-Zr02-CaO/A1z03with an aqueous solution of Ni nitrate.
2.2. Catalytic Activity Measurements Experiments of ethanol decomposition were performed under atmospheric pressure, temperature of 200-450°C, WHSV = 1600-10000 cm3/h.g and inlet watedethanol molar ratio of 8.1-1.04. In a typical experiment a desired flux of helium and water-ethanol mixture to a plug flow reactor, which contained 25 mg of the catalyst diluted with an equal amount and size of quartz particles. The composition of the reactor effluent was analyzed by means of Agilent Mass Spectrometer (for HZ,CH4,C0,C02 EtOH and Ac). For autothermal reforming, a gas mixture composed of 10 vol. % of CH4, 10 vol.% COz, 20 vol. % H2, 4.25 vol.% O2 and 55.75 vol.% H20. The catalytic performance was measured with increasing reaction temperature 85OOC.
3. Results and Discussion 3.1. Effect of Reaction Temperature Homogeneous (empty reactor) steam reforming of ethanol was primarily investigated. It was observed that for temperature values ranging from 200 to 7OO0C the main reaction product was acetaldehyde, carbon oxides and methane Fig. 2, which appear at temperature above 40OoC. It is see that as temperature increases up to about 7OO0C the concentration of acetaldehyde passes through maximum, while a carbon oxides concentration increased with increasing temperature. For the same temperature interval the methane concentration was almost constant (Fig. 2). The ethanol conversion increases with increasing temperature and attains 16% at 70OoC. Furthermore, a considerable carbon imbalance was observed in the temperature range 400-700°C and the visual inspection showed, after a 2 h operation the wall of reactor was covered with carbon.
V. Galvita
338
In ref. [6] we have already discussed the effect of temperature on ethanol conversion and products distribution upon ethanol decomposition over P d C catalyst in steam presence. It was shown that when water-ethanol mixture (H20/CzH50Hmolar ratio
EtOH = 22.5 vol.% H,O = 67.5 vol.% He = 10 vol.% EtOHI H,O = 113
F 1.6 cm31s
o l Temperature, 'C
Fig. 2. Homogeneous (empty reactor) reforming of in the presence of steam and ethanol.
8.1-1.04) was fed into the reactor at WHSV 1600-2200 cm3/g.h, the reaction products were H2, CO, C02 and CH4, and 100% ethanol conversion was attained at 330-360°C. Analysis of the temperature dependencies of the outlet H2, CO, C02 and CHq concentrations proves that within the whole temperature interval the concentrations obeyed the following equations:
This means that CH4,CO and H2 were the products of ethanol decomposition by reaction:
CzH-jOH= C&
+ CO + H2
(31
while COz was the product of the water gas-shift reaction:
It is known [7-lo], however, that ethanol decomposition to C&, CO and H2 over metallic palladium containing catalysts proceeds by reactions:
CzH-jOH-+ CH3CHO + H2
(5)
CH3CHO + CH4 + CO
(6)
Hydrogen Production from Ethanolfor Fuel Cell
339
with the limiting stage being the ethanol dehydration, whereas acetaldehyde decomposition proceeding at high rate (Fig. 3).
0 O
Slow \(,
step
2 H 3I- 4 H + H2
Pd
Fast
step
___ __ __ Pd
Fig. 3. Mechanism of ethanol decomposition at Pd-catalysts
The knowledge of the ethanol decomposition mechanism allows to offer the usage of the bimetallic, bifunctional catalyst for an intensification of process in a two layer fixed bed reactor and for price reduction of the catalyst in the first layer. Ethanol-to-acetaldehyde dehydration will proceed on the first component of such catalyst, with acetaldehyde decomposition to carbon oxide and methane proceeding on the second one (Fig. 4). As reaction (3) effectively and with high selectivity proceeds on I-B group metals [4], copper might be used as the first component of the first layer. As the second component of the first layer it is necessary to retain Pd, which as against Co and Ni, does not give rise to the surface carbon deposition [4, 11, 121.
Fig. 4. Mechanism of ethanol decomposition at bimetallic-bifunctional Cu-Pd-catalysts
Fig. S shows comparative temperature dependences of ethanol conversion during its decomposition over bimetallic catalysts 5wt. %Cu-lwt.% Pd/C and 5wt. %Cu-lwt.% Pd/CeOz-ZrOz, and over lwt.% PdK. One can see that total conversion of ethanol is reached at 330' C over the bimetallic catalysts, whereas for the P d C catalyst at this temperature the conversion does not exceed 80 %. The reaction products for those catalysts were H2, CO, C02, CH4 and CH,CHO, except for Cu-Pd- catalysts at 33OoC, were CH3CHO was absent. The data of Fig. Sa prove the bimetallic catalysts being essentially more active and selective then the catalyst containing Pd only. Moreover, bimetallic catalysts increase the water gas shift reaction. Fig. Sb shows that CO
V. Galvita
340
selectivity for bimetallic catalysts decreased with increasing temperature. Selectivity PdIC at experimental condition was stable.
210
240
270
300
330
Temperature, OC Fig. 5a. Effect of temperature on ethanol conversion upon ethanol decomposition over bimetallic catalysts 5wt.% Cu - lwt.% Pd, and over lwt.% P d C . Experimental conditions: WHSV = 5000 cm3/h,g,inlet composition 8 vol.% CzHsOH + 92 vol.% HzO.
60
b
-
Iwt.% Pd 5w.%CulCeO,-ZrO,
I I 225
\\
I
I
I
I
250
275
300
325
b
, 33
Temperature, ' C
Fig. 5b. Effect of temperature on CO selectivity (b) upon ethanol decomposition over bimetallic catalysts 5wt.% Cu-lwt.% Pd, and over lwt.% PdC. Experimental conditions: WHSV = 5000 cm'hg, inlet composition 8 vol.% CzHSOH+ 92 vol.% HzO.
To estimate the feasibility to use Pd-Cu/Ce02-Zr02catalyst in ethanol decomposition in steam presence, it seemed reasonable to determine its stability under the reaction conditions. Under experimental conditions, (temperature 275"C, ethanol conversion was ca. 99% (100% ethanol conversion was attained at 280°C)); only H2, CH4, CO, C 0 2 and
Hydrogen Production from Ethanol,for Fuel Cell
341
C2H50H were detected at the reactor outlet. The outlet concentrations were stable for 50 h. There is no carbon imbalance (within f 5 % accuracy) was observed during this experimental run. Thus, the use of the bifunctional, bimetallic catalyst as the catalyst for the first layer of the two-layer reactor for ethanol decomposition proved to be quite promising for the steam reforming of ethanol to syngas. 3.2. Autothermal Reforming of Methane over Supported Ni Catalysts The effect of temperature and amount of Ni in catalysts In the autothermal reforming of methane is shown in Fig. 6. Methane conversion increases with increasing reaction temperature. Catalytic activity increase with increasing of amount of Ni. The thermodynamic equilibrium of methane reforming were attained only at temperature high 75OoC and for catalysts containing 5 and 10 wt % of Ni. Thus, for autothermal reforming of methane the Ni (10 wt.%) catalyst supported over Ce02-ZrO2-CaO/Al203showed high activity and stability. We can use the Ni- Ce02Zr02-CaO/A1203catalyst as the catalyst for the second layer of the two-layer reactor for methane autothermal reforming. 100
s
95
C
.-0
$ C
CH,"= 10 vol.%
co,,"= 10 vol.%
90
H,," = 20 vol.%
8
H,O,"= 20 vol.%
85
O,,"= 4.25vol.%
m
5 80
2 wt.% Ni
625
650
675
700
725
750
775
800
Temperature, OC Fig. 6. Temperature dependence of methane conversion for autothermal reforming of methane over Ni- Ce02Zr02-CaO/A120, catalysts.
4. Conclusions
The concept of the two-layer fixed-bed reactor seems to be quite promising for the steam or autothermal reforming of ethanol to syngas. According to this concept, ethanol is first converted to the H2, CO, C02 and CH4 gas mixture over a bimetallic catalyst (for example, Cu-Pd/Ce02-Zr02) and then the mixture produced is converted to syngas over another Ni-contain catalyst by autothermal reforming. It has been shown that the use of
V. Galvita
342
the two-layer fixed-bed reactor prevents coke formation and provides the syngas yield closed to equilibrium. It was found that the bimetallic catalyst being essentially more active then the catalyst containing Pd only. For autothermal reforming of methane (produced after ethanol decomposition in the first layer), the Ni (10 wt.%) catalyst supported over Ce02-Zr02-CaO/A1203showed high activity and stability for autothermal reforming of methane. References [I] Ioannides T. (2001) Thermodynamic analysis of ethanol processors for fuel cell
application. J. Power Sources 92 (1-2): 17-25. [2] Aupretre F, Descorme C, Duprez D. (2002) Bio-ethanol catalytic steam reforming
over supported metal catalysts. Catal. Communication 3 (6): 263-267. [3] Fatsikostas A, Kondarides D, Verykios X. (2002) Production of hydrogen for fuel cells by reformation of biomass-derived ethanol. Catal. Today 75 (1-4): 145-155. [4] Galvita V, Semin G, Belyaev V, Semikolenov V, Tsiakaras P, Sobyanin VA. (2001) Synthesis gas production by steam reforming of ethanol. Appl. Catal. A. General 220 (1-2): 123-127. [5] Semikolenov V. (1997) Zhurnal Prikladnoi Khimii 70: 785 (in Russ.). [6] Galvita V, Belyaev V, Semikolenov V, Tsiakaras P, Frumin A, Sobyanin V. (2002) Ethanol decomposition over pd-based catalysts in the presence of steam. React. Kinet. Catal. Lett. 76 (2): 343-35 1. [7] Isawa N, Yamamoto 0 , Tamura R, Nishikubo M, Takezawa N. (1999) Catal. Lett. 62 (2-4): 179-184. [8] Takezawa N, Isawa N. (1997) Steam reforming and dehydrogenation of methanol: Difference in the catalytic functions of copper and group VIII metals. Cutal. Today 36 (1): 45-56. [9] Davis JL, Barteau MA. (1987) Decarbonylation and decomposition pathways of alcohols on Pd (1 11). Surf: Sci. 187 (1-2): 386-387. [ 101 Shekhar R, Barteau MA. (1995) Structure sensivity of alcohol reactions on (1 10) and (1 11) palladium surfaces. Catal. Lett. 31 (2-3): 221-237. [ 111 Freni S, Mondello N, Cavallaro S, Cacciola G, Parmon VN, Sobyanin VA. (2000) Hydrogen production by steam reforming of ethanol: a two step process. React., Kinet. Catal. Lett. 71 (1): 143-152. [I21 Cavallaro S, Mondello N, Freni S. (2001) Hydrogen produced from ethanol for internal reforming molten carbonate fuel cell. J. Power Sources 102 (1-2): 198204.
HYDROGEN FUEL CELL URBAN BUSES OPERATING IN THE CITY OF PORT0 G.A. GONCALVES('), T.L. FANAS('), R. TEIXEIRA(2),A. SILVA(3) (1) IST-Instituto Superior Tecnico, DEM-STE, 1049-001 Lisboa, Portugal (2) STCP-Sociedade de Transportes Colectivos do Porto, Av. Fernao de Magalhaes 1862-13, 4350-158 Porto, Portugal (3) Portugal Comercio de Combustiveis e Lubrificantes S.A., Lagoas Park -Edijcio 3 2740-244 Porto Salvo The aim of the CUTE project was to develop and demonstrate an emission-free and low-noise transport system, including the accompanying energy infrastructure, which has great potential for reducing the global greenhouse effect towards the Kyoto protocol, improving the quality of the atmosphere and life in densely populated areas and conserving fossil resources. For the CUTE project a total of 27 buses were built by EvoBus. These vehicles are based on the Mercedes-Benz Citaro model city bus and run on hydrogen using fuel cells for power generation. One of the participating cities was Porto, where the local bus company (STCP - Sociedade de Transportes Colectivos do Porto) is operating three buses. The hydrogen infrastructure is supplied by BP. The hydrogen is produced centrally by electrolysis from grid electricity. In the project, IST is conducting the study of the environmental impact (based on a well to wheel approach) of the operation of the buses and the comparison with conventional technologies (Diesel, compressed natural gas and electricity). This will include measurements of the actual consumptions of the vehicles for the different routes, identification of the road profiles, driving behavior, load and ambient conditions.
1. Objectives The aim of the CUTE [5] project is to develop and demonstrate an emission-free and lownoise transport system, including the accompanying energy infrastructure, which has great potential for reducing the global greenhouse effect towards the Kyoto protocol, improving the quality of the atmosphere and life in densely populated areas and conserving fossil resources. For this purpose, 27 hydrogen fuelled fuel-cell buses were tested for 2 years in 9 European cities between 2003 and 2005. This project has financial support from the European Community and is currently the largest project worldwide in he1 cell technology demonstration in transports. 2. Project Partners
The participants of the project included the bus operators and local authorities for the 9 cities, BP and Shell as suppliers of hydrogen and the infrastructure and several research organizations. A more detailed list is presented on Table 1.
3. CUTE, ECTOS and STEP: 3 Projects Running in Parallel The ECTOS project aimed to implement and demonstrate the state-of-the-art hydrogen technology by running part of the public transport system with fuel cell busses within Reykjavik, Iceland. The hydrogen was produced from hydro and geothermal power, enabling a C02-free production chain. In the ECTOS project, three fuels cell buses similar to the ones used in CUTE were tested, and the project ran parallel in Reykjavik with the CUTE project. The STEP project is part of the Western Australian Government's commitment to working towards sustainable and 'zero emission' transportation in Western Australia. 343
G.A. Goncalves et al.
344
Table 1. Project participants.
s t a t h a f t SF Storstockholms Lokaltraffik (SL) Stuttgarter StraBenbahn AG (SSB) Sydkraft AB Transports Metropolitans de Barcelona S.A. (TMB) University of Iceland ECTOS University of Stuttgart (USTUTT) ~
NO SE DE SE
ES IS
DE
Energy supplier Public transport authority Public transport authority / company Hydrogen facility manufacturer Public transport authority / company Research Research
Three hydrogen fuel cell buses similar to the ones used in the CUTE project were trialed on normal Perth service routes for two years, from early 2004 to 2006. 4. The Vehicle
The vehicle being used in the project (Fig. 1) is based on the standard Diesel powered Citaro bus commercialized by EvoBus (a subsidiary of the DaimlerChrysler group). The engine and transmission are located in the same position and the vehicle offers the same
Hydrogen Fuel Cell Urban Buses Operating in the City of Porto
345
passenger space. In order to increase the r e ~ ~ a b iofl ithe ~ vehicle, the changes have been reduced to a m ~ n ~keeping ~ u ~as ,much of the conventional bus parts as possible.
Fig. 1 . 1 l e CUTE project bus.
The entire -fuel cell system i s located on the roof of the bus (Fig. 2). This required some changes in the suspension and chassis in order to cope with the extra weight. in a high location, This has been decided based on several considerations: Safety in terns of accidents, as the rooftop ofthe bus i s rarely damaged in the case of a collision; ]Inthe case o f leakage, the hydrogen is ~ ~ e d ~vented a t toe the ~ atmosphere; ~ The location provides easy access for maintenance. X 1. ~~~r~~~~Sforage
The hydrogen is stored on board in compressed form at a pressure of 350 bar. 'The storage ~ n each ~ with a volume of module uses 9 cylinders (carbon fiber wrapped a l ~ m liner), 205 Wes. The total storage capacity (at the nominal pressure of 350bar) is 44kg of hydrogen, suflficient to fulfill the typical daily range requirements for city buses.
Fig. 2. Location of the main elements.
346
G.A. Gomalves et a1
4.2. Fuel Cell The fuel cell is of the PEM type, and the two stacks are capable of providing a total gross power of 250 kW. The DC current produced is converted into AC in an inverter before being fed to the main electric engine (Fig. 3). In order to ensure the simplicity of the system, all other electrical components are supplied from the main engine.
4.3. Engine and Power Train The electrical traction engine is placed in the same position as in a conventional bus (Fig. 4). It is designed to a maximum power of 205kW, comparable to the power of conventional diesel and natural gas city buses. The rear axle and automatic transmission are similar to the ones used in the conventional Citaro model, the changes are essentially in the gearbox gear ratios, which have been adapted for the specific characteristics of the engine. The presence of the gearbox ensures an easier adaptation of the drivers.
I I
Cwnprerror
Cooling
Expander
Pump
Hydrogen
wafer
mmf E
Inverter
Electricat Traction
Tank
Engine
Fig. 3 . Operation of the fuel cell system (Source DaimlerChrysler AG).
Fig. 4. Engine and gearbox location.
Hydrogen Fuel Cell Urban Buses Operating in the City of Porto
347
5. Hydrogen Supply Infrastructure One of the most important factors when developing an alternative fuel transportation system is the capacity to supply the fuel, especially if the nature of the fuel prevents it from being distributed through conventional refueling stations. This is the case for hydrogen, which requires a specific fuelling infrastructure. Some of the bus companies that participate in the project already have recognized experience with gaseous fuels, owning a significant number of natural gas buses. In the case of hydrogen, another problem is the need to manufacture the fuel. This can be done in several ways and from a wide variety of primary energy sources, and was one of the goals of the CUTE project to analyze different combinations of production and distribution of hydrogen.
5.1. Hydrogen Supply Routes One of the main drivers to the introduction of hydrogen is the possibility to have a local as well as global C 0 2 free system. For that purpose, a renewable energy source must be used, and that is the case when using hydro or geothermal power. However, cost and availability issues bring the possibility of using conventional fossil fuels as primary energy sources. At the present stage, steam reforming of natural gas is the most probable route to produce hydrogen on a large scale (needed if it is to be introduced in widespread use as transportation fuel). For the CUTE project, given its demonstration status, it was been decided to cover a broad range of alternatives, not only in what refers to primary energy source but also if the production is centralized or on-site. The routes taken arc summarized in Table 2 . Different filling stations were used in each city, and the hydrogen is either stored at a high pressure (-450bar) prior to refueling or is compressed to a suitable pressure during filling. In any case, the filling time is not expected to exceed 10 minutes. The use of different filling station technologies allowed drawing conclusions to what is the best way to provide hydrogen at such high pressures, regarding not only the energy consumed for compressing the hydrogen, but also the availability of the filling station. Table 2. Hydrogen supply routes used in the CUTE project.
Cities within the project Amsterdam, the Netherlands Barcelona, Spain Hamburg, Germany
I Primary energy source
I Hydrogen Production
IRenewable (solar)
IOn site (electrolysis)
I Renewable (meen electricitvl
I Renewable (wind)
I On site (electrolvsisl
I On site (electrolysis)
London, United Kingdom
Crude oil
Centralized (transport in liquid form)
City of Luxembourg
Natural gas
Centralized (transport in gaseous form)
Madrid, Spain
Crude oil + Natural gas
Porto, Portugal Reykjavik, Iceland - ECTOS
I IVariable (grid mix)
Centralized (transport in gaseous form)
I + On site (Steam ieformei)
Centralized (transport in gaseous form)
Renewable (Geothermal and hydro energy)
On site (electrolysis)
Stockholm, Sweden
Renewable (green electricity)
On site (electrolysis)
Stuttgart, Germany
Natural gas
On site (Steam reformer)
G.A. Goncalves ed QE.
448
5.2. Hydrogen Infrastructure in Porio
The hydrogen ~ f r a s ~ inc Port0 ~ e (Fig. 5 ) was supplied by BP. The hydrogen is produced in the Linde factory located in Alenquer ~ o u g an h electrolysis process using grid elechicity. It is then an sported by road in tankers (in gaseous form at 200 bar) to the falling station. The trailers are then connected to the station, sewing as s t a ~ ~ o n a ~ storage. As the storage presswe in the buses has a m ~ i m u mof 350 bar, in order to transfer the hydrogen to the buses an additional compression stage is required, raising the filling presswe to 400 bar (the system must also cope with tlae reduction in the p r e s s ~ eof the t a k e r as the hydrogen is transferred). The station has parking space and connecting equipment for two trailers.
Fig. 5. Hydrogen tanker, i ~ t e ~ ~ d storage i a t e and filling station with dummy test tank.
6. city $pecific ~ o n ~ ~ ~ o n ~
Adding to the different. hydrogen production routes, each city presents specific operation e s . factors are s ~ ~ conditions in trafGc, topography and average t e ~ p e r a ~ ~These in Table 3. 'The number of passenger transported indudes all the modes of transport operated by the transport company. Table 3. City specific characteristics.
Cities Within the project
Barcelona, Spain
I 1.300.000
I
7. Bus ~~~~~~~o~ in Porto
The buses were operated by STCP. Maintenance infrastrucwes were created that meet the safety regulations for operation with highly volatile and flammable gaseous &el
~
Hydrogen Fuel Cell Urban Buses Operating in the City of Porto
349
(Fig. 6). Equipment installed includes H2 detectors, gas extraction equipment and fire detectors.
Fig. 6 . Hydrogen bus depot location at the Francos station.
The route for the operation of the buses was selected for the high gradients and good visibility to the population. It’s a circular line (Fig. 7) and the hydrogen buses were operated as complement to the Diesel and natural gas vehicles. Therefore all the vehicles were operated under the same conditions. 8. Environmental Impact of the Hydrogen Fuel Cell Bus (Well-To-Wheel Analysis)
When using hydrogen as fuel, the local emissions of a fuel cell bus consist only of water. However, to evaluate the global environmental impact of this fuel, a complete analysis must be made covering all the steps in the production and use of the fuel, from the primary energy source to the end use - a well-to-wheel analysis. This task is being performed by IST together with the University of Stuttgart and consists of two phases.
Fig. 7. Route layout.
Within the first phase, a well-to tank analysis was be performed by the University of Stuttgart. This analysis covers the life of the fuel from the primary energy source (renewable, crude oil or natural gas) to the moment it is filled to the on-board reservoir.
350
G.A. Goncalves et al.
Using a Life Cycle assessment tool (GAB1 Software [ 11, develop at the University of Stuttgart), the environmental impact of the production of hydrogen was compared with the impact of conventional fuels, as can be seen in Fig. 8. WELL-TO-TANK
011
Fig. 8. Well-to-Tank analysis.
In the second phase, the Tank-to-Wheel analysis was performed by IST in order to quantify the impact of the operation of the fuel cell bus and compare it to conventional Diesel and Natural Gas buses (see Fig. 9). TANK-TO-WHEEL \
/
VEHICLE AND FUEL
OPERATION PARAMETERS
\
I
TAILPIPE EMISSIONS
TOPOGRAPHY
I
CLIMATE
1
gas bus
\ I Hydrogen bus
~
Fig. 9. Tank-to-Wheel analysis.
Hydrogen Fuel Cell Urban Buses Operating in the City of Porto
351
For this purpose three tools were used: 0 Mobile6 [2], a tool developed by the EPA (Environment Protection Agency, United States) to calculate the emissions of road vehicles; Copert [3], which fulfils a similar role and represents the European market; 0 EcoGest [4] is a tool developed at IST that uses as input the characteristics of the vehicle, driving behavior and topography and, based on those inputs, calculates on a second by second basis the power needed and the resulting emissions and fuel consumption. These three tools were used in order to ensure reliable results from the simulations: Mobile and Copert are proven tools that calculate moderately accurate results, while EcoGest allows for much more precise calculations. In addition to the simulations, the vehicles will be monitored during their regular operation. This monitorization will be extended to the conventional Diesel and natural gas vehicles. In this case the parameters monitored will be dynamic performance and fuel consumption. The combination of the life-cycle analysis of the production and distribution of the fuels, together with the values from the operation of the different technologies produced a well-to-wheel analysis for the fuel cell bus as well as conventional Diesel and natural gas buses, and thus allowing for the completion of the remaining IST tasks within the CUTE project: Evaluation of the Environmental Impact of the bus operation; 0 Compare Hydrogen with conventional Diesel/Natural Gas technologies; 0 Contribution to the objectives of the Kyoto Protocol; Monitoring of the Energy and Environmental performance of the vehicles. Acknowledgments
The CUTE project was co-financed by the European Union. CUTE is one of the research projects supported by the European Commission, DG Energy and Transport (NNE5/113/2000). One of the partners, STCP, has support from DGTT (Direccao Geral dos Transportes Terrestres). References [ 11 IKP (1992-2002): GaBi Software-System and database for Life Cycle Engineering; www.gabisoftware.com [2] US EPA User's Guide to Mobile6.1 and Mobile6.2: Mobile Source Emission Factor Model, Office of Transportation and Air Quality, United States Environmental Protection Agency, 2002. [3] Ntziachristos L, Samaras Z. (2000) COPERT 111: Computer program to calculate emissions from road transport - Methodology and emission factors (Version 2.1), European Environment Agency. [4] Silva CM, Farias TL, Mendes-Lopes JMC (2002) EcoGest - Numerical modeling of the dynamic, fuel consumption and tailpipe emissions of vehicles equipped with spark ignition engines. Proceedings of Urban Transport and the Environment 2002 - Eighth International Conference on Urban Transport and The Environment in the 2 1st Century, 13-15 March, Seville, Spain. [5] http://www.fuel-cell-bus-club.com.
This page intentionally left blank
ENERGY IN SLOVENIA AND CROATIA: CHALLENGES AND POSSIBILITIES FOR SUSTAINABILITY A. ZIDANSEK“32’,R. BLINC‘22”,1. SLAUS (3), D. NAJDOVSKI‘4’ “)Jozef Stefan International Postgraduate School, Ljubljana, Slovenia ‘2)J.Stefan Institute, Ljubljana, Slovenia “R. Boskovic Institute, Zagreb, Croatia
‘‘jX3data, Ljubljana, Slovenia Three components of a strategy to reduce the negative environmental impact of energy use in Slovenia and Croatia are proposed. This strategy is based on energy production and energy consumption data, which are are presented and analyzed for Slovenia and Croatia. The first component is maximising the use of renewable energy sources. The second strategy component comprises scenarios for changing consumption patterns to meet the sustainability requirements. The third one is emphasising fundamental curiosity driven research that throughout the history has proven to be the most effective solution for various problems including those of energy production. These strategies are being analyzed for Slovenia and Croatia. Besides the technology fix we discuss also social implications of a change in consumption patterns. These three components of the strategy aim not only to meet the requirements of the Kyoto protocol and EU standards, but to make significant progress toward achieving genuine sustainable development according to the original Brundtland Commission definition.
1. Introduction Development is interconnected with energy. Sustainable development requires that energy is being produced and consumed in a way that does not produce a negative impact on the environment, and that the quality of life and development of future generations are not jeopardized. Each country has to establish its own best strategy for achieving sustainable development. Slovenia has just become a member of the EU, and Croatia is bordering with two members of EU and is aspiring to the full membership in EU. Data on energy production and energy consumption for Slovenia and Croatia are presented and analysed and then we proceed to outline the strategy for reducing the negative environmental impact of energy use. 2. Analysis of the Current Energy Production and Consumption in Slovenia and Croatia
Primary energy production, energy consumption, energy consumption per capita and energy consumption per unit GDP for Slovenia and Croatia in comparison with Austria and Macedonia are shown in Figs. 1-4. At present renewable energy sources in Slovenia amount to only 7% of the end energy consumption, if small hydroelectric power plants are also counted as renewable energy. This involves mainly biomass, small hydroelectric power plants, geothermal and solar energy, while large hydroelectric power plants are here not considered as renewable energy sources. In Croatia the renewable energy includes mainly biomass-fuel wood [l-31.
353
354
A. Zidansek et al.
A
A
-0-
Croatia
I
I
I
I
I
1992
1994
1996
1998
2000
2002
Year Fig. 1. Primary energy production for Slovenia and Croatia in comparison with Austria and Macedonia in EJ (lOIR J) [l].
1
I
I
I
I
I
1
a
-v-
v--r-v--v-v-v
-.--\v1992
Macedonia
I
I
I
I
1994
1996
1998
2000
2002
Year Fig. 2. Total primary energy consumption for Slovenia and Croatia in comparison with Austria and Macedonia in EJ ( l o l a J) [l].
355
Energy in Slovenia and Croatia
180
$, 170 Q
8
160 150
L
140
c
0 .c
130
;
a 120 110
8
100
h
p 90 (u
5
80
2 m
70
.E
60
?I
1992
1994
1996
1998
2000
2002
Year Fig. 3. Energy consumption per capita for Slovenia and Croatia in comparison with Austria and Macedonia in GJ (lo9 J) 111.
1992
1994
1996
1998
2000
2002
Year Fig. 4. Energy consumption per unit GDP for Slovenia and Croatia in comparison with Austria and Macedonia in H/US $ GDP [l].
356
A . Zidansek et al.
Table 1 summarizes selected energy data for Slovenia and Croatia. Slovenia and Croatia are equal owners of the Nuclear plant Krsko (NEK), which is located in Slovenia. Table 1. Energy data for Slovenia and Croatia in 2001 [1-3].
Total Electricity Production
Slovenia
Croatia
13.7
12.1
13.8
14.3
35.2%
33.6%
36.8% 1.1 53 1040
3.0 60 1080
in TWh Total Electricity Consumption in TWh ~
Electricity Production by Source
Net Electricity Import Net Oil Import Net Natural Gas Import
Nuclear in TWh in thousand bbl/day in million cubic meters
The fraction of renewable energy sources, excluding hydro-energy, in the electrical energy both in Slovenia and Croatia is even lower as the following discussion shows. About 100 000 square meters of solar collectors are currently installed in Slovenia. The current peak power of the installed photovoltaic systems is about 100 kW. A large part of the photovoltaic systems has been installed in the rural area of the Triglav national park. The average daily solar irradiation in Koper at the Adriatic coast is 5800 WWm2 in summer and 1000 WNm2 in winter, whereas in central Slovenia (Ljubljana) the corresponding irradiation is 5300 WWm2 in summer and 600 WNm2 in winter. The cost of solar energy is however relatively high. The estimated geothermal energy potential of Slovenia is about 50-70 MW. The biomass potential is estimated to 40 MW. Croatia is the owner of half of the nuclear power plant in KrSko, so that a fraction of its energy comes from this source. About 150 000 square meters of photovoltaic systems are currently installed in private houses and hotel roofs. The total installed capacity of solar energy in Croatia was 6 MW in the year 2000, while the total potential of the solar energy is 100 PJ. The solar energy panels in Croatia are mainly installed along the coast and in the islands. The installed biomass capacities amount to 7 MW.
3. Components of a Strategy to Reduce the Negative Environmental Impact of Energy Use We propose three components of a strategy to reduce the negative environmental impact of energy use. The first one is maximising the use of renewable energy sources. The second strategy component comprises scenarios for changing consumption patterns to meet the sustainability requirements. The third one is emphasising fundamental curiosity driven research which throughout the history has proven to be the most effective solution for various problems including those of energy production.
Energy in Slovenia and Croatia
351
A comparison of C 0 2 emissions per capita for Croatia and Slovenia (Fig. 5) demonstrates that the emissions in Croatia are increasing in recent years, while emissions in Slovenia are already at a very high level which is comparable to developed countries. The increase in Slovenian emissions between 1994 and 1999 was mainly due to the fact that prices of gas were significantly lower than in the neighbouring countries.
i-
/
/
-m--0-
Slovenia Croatia
-v- Macedonia
1992
1994
0
1996
1998
2000
2002
Year
Fig. 5 . COz emissions in metric tons carbon equivalent per capita for Slovenia and Croatia in comparison with Austria and Macedonia [l].
3.1. Increasing Renewable Energy To reach lower emission of greenhouse gasses in accordance with the Kyoto protocol and European Union standards Slovenia has to increase renewable energy sources (without hydroenergy) to reach 12% till 2010, and Croatia plans to reach 12.2% of the end energy consumption by 2030. So far no commercial wind energy plants are operating neither in Croatia nor in Slovenia, but several pilot projects are planned and some of them are in the testing phase. Slovenia plans that photovoltaic systems with 3 MW peak power should is installed in the next years. A geothermal pilot power plant of 5 MW is planned to be installed till 2010. Until 2010 the installed biomass power should be 8 - 10 MW. The biogas power plant potential at animal farms and waste treatment facilities in cities on the other hand is estimated to additional 10-30 MW. The wind energy sources have a relatively large potential and the installed power till 2010 is estimated to 40-80 MW
[1,21. The total wind energy potential on Croatian islands is 209 MW, while it is 163 MW at the Adriatic coast. The total installed capacity of geothermal energy sources in
358
A . Zidansek et al.
Croatia in 2000 was 36.7 MW, whereas the total geothermal potential is estimated at 839 MW [ l , 31. 3.2. Sustainable Consumption Patterns Presently, an energy consumption pattern is determined mainly by economic considerations, as well as by personal and local convenience. This has resulted in a large fraction of obsolete energy components producing large negative environmental impact. Currently, energy consumption is characterized by large inefficiency. Analyses and practices in several countries have shown that it is possible to maintain and even increase GDPicapita without increasing energy consumption and even decreasing it. This strategy component is based on modem technologies which are energy frugal, notably on information and communication technologies. It is necessary to simultaneously analyse for each country social, economic and environmental impact introduced by these modem technologies. Implementation of sustainable consumption patterns is slow, because production cost for renewable energy is much bigger than for the energy from fossil fuels or nuclear energy. This cost estimate becomes completely different, if external costs are also taken into account in addition to direct costs of energy production. Here also the price of the damage caused to the environment is taken into account. In Slovenia this damage due to climate changes alone exceeds 100 million Euros annually. In Slovenia until 2010 greenhouse gas emissions will rise by about 10% according to the EEA estimates instead of decreasing by 8% versus the 1986 baseline as agreed by the Kyoto protocol. According to most climate models the total world greenhouse gas emissions should be reduced well beyond Kyoto requirements to about one half of the current emissions in order to stabilize the climate and precent the negative effects caused by the threat of climate changes. In Slovenia C 0 2 emissions alone represent 16.2 million tons anually. These emissions cause external damage estimated to 18 euros per tone by EEA. If we assume that one half of these emissions should be reduced in order to stabilize the global climate, this one half of the emissions amounts to the total of 140 million euros external costs. If Kyoto protocal was accepted, Slovenia should already buy the emission rights for more than 3 million tons COz anually, which amounts to about 60 million euros anually. If this amount is used to support renewable energy sources and in this way take into account the external costs of energy use, the renewable energy sources would become competitive. This can of course not be done at a single country level, but a global agreement should be reached. If we want to achieve sustainable consumption patterns, a global green tax reform is therefore needed. The cost of environmental damage produced by production of energy must be included into the consumer price.
Energy in Slovenia and Croatia
359
3.3. Promising R&D Directions Fundamental curiosity driven research throughout the history has shown that major breakthroughs almost always have large social implications. For the energy related field the well known example is Faraday’s discovery of electromagnetic induction. It is this type of research that has to be encouraged and stimulated. However, often it is said that small countries should try to concentrate on applied research, but we strongly believe that this is not enough since: research resulting in breakthroughs is enriched by differences in culture and environment; it is impossible to find the best solution for a country by just applying outside recipes without strength which is assured only by fundamental research; applications come from fundamental research, which then is the prerequisite for any successful application. It is not the aim of this paper to go into details of what kind of fundamental curiosity driven research should be done in these two countries. It suffices to say that high quality fundamental research must be supported at an adequate level. While fundamental curiosity driven research can’t be planned, mission oriented fundamental and applied research should be planned. In the following some promising R&D directions are presented.
3.3.1. Hydrogen Research
A promising energy source is hydrogen. In principle hydrogen is a completely clean energy source and its only emission at the end user is clean water. However, the problem is that the most efficient way to produce hydrogen is electrolysis, which demands electricity usually obtained from the fossil fuels. A possible solution is to obtain the electricity from a renewable source. In this case hydrogen acts basically as substance to transmit energy. Starr suggests using the nuclear energy as a source for the Continental Super Grid, where the energy is transmitted to the end user via hydrogen, which feeds fuel cells [4]. A better solution would be the use of biologically derived hydrogen. While the research in this field is still in the initial phase, there are some promising results with bacteria [ 5 ] and green algae [6]. The research in nanotube with a possible application for hydrogen storage is beeing developed by the laboratories at J. Stefan Institute. Fuel cell cars have recently been introduced as a possibility to use hydrogen as a clean energy source. The advantage of these cars is that they create no polution in the city, although there is still polution created in producing hydrogen. 3.3.2. Zirconium Ceramicsfor Nuclear Waste Storage
Slovenia and Croatia own a joint nuclear power plant which produces significantly less COz emissions per MWh than fossil fuels. It does not require burning fossil fuels, but uses uranium fuel instead. Therefore there is only a small amount of indirect emissions, connected with the uranium fuel cylcle, and even this only in case that fossil fuels are used in preparation, transport and treatment of the nuclear fuel. Nuclear energy is
360
A. Zidansek et al.
however connected with other risks. One of the most important issues is the necessity to solve the problem of nuclear waste. A possibility to secure sustainable treatment of nuclear waste is the use of zircon ceramics to embed spent reactor fuel. Several billion years old natural zircon crystallites clearly show alpha particle tracks due to uranium impurities, which proves the efficiency of zircon based materials to assure long term containment of uranium [7]. Both J. Stefan Institute and R. BoSkoviC Institute are planning R&D projects in this direction.
3.3.3. Piezo and Solar Panels as Noise Barriers along the Highways One of the possibility to increase the use of photovoltaic power sources is to utilize solar cell panels as anti-noise barriers along the highways. Since anti-noise barriers amount to about 20% of the cost of building the highway roads, the cost of solar energy obtained in this way would be much lower than that from other commercial solar sources [8]. Two projects are currently in pilot phase at two different sections of the highway under construction in Slovenia. The first project is based on the use of highway noise barriers equipped with photovoltaics and piezomaterials for energy production. Technical documentation for the test field at the KrSka vas - Obre2je highway in the length of 100 m has been prepared. In this case the noise barrier will be equipped with photovoltaics and piezomaterials. This project is aimed to demonstrate how much energy can be obtained from the noise barrier while the noise protection is improved. The noise barrier device will serve as a self organized energy device where on one hand the energy is obtained from photovoltaics and piezomaterials, and on the other hand the device also needs energy for absorption of high frequency acoustic noise emissions. The goal will be to satisfy these energy needs with only part of the produced energy, so that the energy balance of the device is positive. Another similar project is designed for a windy area of the Razdrto-Vipava highway in the initial length of 500 m with a possible extension to 8 km.Here the noise barrier will also serve as a wind barrier. Also in this case the energy balance will be studied. Due to high wind loads the wind represents an additional source of energy, which will improve the energy balance of the device. 4. Conclusions
Suggested energy strategies for Slovenia and Croatia are summarised as follows: 1. Increase GDPkapita maintaining energy consumption per capita constant or at most keeping GDP/capita growth twice as large as the energy consumptiodcapita growth. 2. Increase the fraction of solar, geothermal, biomass and small hydro energy in 2010 to be 50% larger than it is in 2004. 3. Maintain and develop nuclear science and nuclear technology skills so that strategies to reduce nuclear waste problems can be implemented. 4. Change the pattern of energy consumption by emphasising renewable energy sources and reducing fossil fuel.
Energy in Slovenia and Croatia
361
5.
Increase R&D support and R&D activity to provide opportunities for major breakthroughs and to assure incremental improvements in energy technologies, in particular renewable technologies. The negative environmental impact of energy use in Slovenia and Croatia has been analyzed. The analysis leads to a pragmatic policy to keep all energy options open allowing each one to seek its role in a competitive and economical energy mix. Discussed changes in the consumption patterns represent an important option for reducing negative impacts and are therefore a vital component of a successful strategy for sustainable development. The fastest way to change the consumption patterns is the implementation of green tax reform, where external costs of the environmental damage are also included into the consumer price. This would however significantly increase the price of fossil fuel energy, and is thus only acceptable, if a global agreement is reached. A proper combination of green tax reform and research can thus lead to new energy consumption patterns which are both environmentally sustainable and economically viable. Here environmental sustainability aims to reduce the fossil fuel use well beyond the proposed Kyoto targets in order to stabilize the global climate, and increased support for energy research is expected to decrease the price of clean energy so that it becomes competitive with the fossil fuels. If only a small fraction of external environmental costs were spent on research, clean and renewable energy technologies would become cheaper, new energy sources like fusion and energy producing bacteria closer, and radioactive waste storage safer. References [I] Central Intelligence Agency (2003) The World Factbook. Potomac Books, Washington, D.C. [2] Ministry of the Environment, Spatial Planning and Energy of the Republic of Slovenia (2002) Porocilo o stanju okolja v Sloveniji 2002. Ljubljana. [ 3 ] Vuk B, Marusic D. (2002) Energija u Hntatskoj 2001 - Energy in Croatia 2001, Ministry of Economy of the Republic of Croatia, Zagreb. [4] Starr C. (2001) National Energy Planning for the Century. ANS Winter Meeting, November 13, Reno, Nevada, USA. [5] Van Ooteghem SA, Beer SK, Yue PC. (2002) Hydrogen production by the thermophilic bacterium Thermotoga neapolitana. Applied Biochemistvy and Biotechnology 98: 177-189. [6] Melis A, Happe T. (2001) Hydrogen production. Green algae as a source of energy. Plant Physiology 127: 740-748. [7] Scott JF, private communication, 2003. [8] Nordmann T, Frolich A, Durr M, Goetzberger A. (2000) First experience with a bifacial PV noise barrier. 16th European Photovoltaic Solar Energy Conference and Exhibition, Glasgow, United Kingdom.
This page intentionally left blank
SUSTAINABILITY ASSESSMENT AS A BASIS FOR THE DECISION MAKING IN SELECTION OF ENERGY SYSTEM F. BEGIC"), N.H. AFGAN(*) (I)
J.P.Elektroprivreda Bosnia and Herzegovina, Sarajevo, Bosnia and Herzegovina (2) Instituto Superior Tecnico, Lisbon, Portugal
In this paper the multicriteria sustainability assessment of various options of the energy power system of the JP Elektroprivreda of Bosnia and Herzegovina is performed. The rehabilitation of a 110 MW Thermal Power Unit, is considered in comparison with some other comparable options, such as Thermal power unit with coal fuelled boiler with combustion in fluidized bed, Combined cycle gas turbine plants, Hydro Power Plant, Power plants based on solar energy (PV systems), Wind turbines and Bio-mass power plants. Used methodology is based on a system of stochastic models of uncertainty, enabling to realize the assessment from various supporting systems, as well as enabling to obtain various normalization indexes by using Non-numeric (ordinal), Non-exact (interval) and Non-complete information (NNN-information) obtained from sources of various reliability and probability. Trough the analysis of multicriteria assessment of various options the decision-makers are offered to evaluate different options and make selection of the optimal option which meet the muticriteria evaluation.
1. Introduction
Most of decision models used in evaluation of complex systems today are based on single criteria analysis. However, in our modern age, decision-making based of single criteria analysis becomes unacceptable, if we set priorities that include costs, requirements, possibilities and risks. It is often necessary to consider several aspects at the same time. Energy power system is a complex system which evaluation depends on a number of economic, environmental, social and technological parameters. One of the perspective methods for evaluation of energy system quality is multicriteria sustainability assessment based on the analysis and synthesis of indicators expressing different aspects of the system. Application of this method in the cases of information deficiency (ASPID methodology) enables evaluation of various energy systems. In this work, multicriteria sustainability assessment of energy systems of various sources is applied to the Energy Power System of Bosnia and Herzegovine in order to investigate this complex system.
2. Energy Power System Under Consideration In the last five years, as the result of intensive efforts of power plants rehabilitation and refurbishment of entire electric power system, electricity supply in Bosnia Herzegovina has been improved. In year 2000, the electricity generation reached 63% of electricity generation in pre-war 1990 year, Fig. 1. The assessments of electricity consumption in domain of JP Elektroprivreda BiH, estimated by various consulting companies, have shown that the growing of electricity consumption in the next period (till year 2020) should continue. Today, the combined capacity of all power plants in Bosnia and Herzegovina is 3940 MW, of which 1983 MW is in hydropower plants (not including the small hydropower plants) and 1957 MW is in thermal power plants. The capacity of power plants belonging to JP Elektroprivreda BiH - a largest power company in Bosnia, is 1839 MW, or 47% of total capacity in Bosnia and Herzegovina. From this power generation, 482 MW is in hydropower plants, and 1357 MW in thermal power plants. An additional 3 63
F. Begic & N.H. Afgan
364
1990
1991
1992
1993
1Y94
1995
1996
1597
1998
1999
2000
Fig. 1. Trend of power generation and consumption within a realistic and complex energy power system under consideration.
11 MW is available to JP Elektroprivreda BiH from small hydropower plants, bringing total capacity to 1850 MW. Thermal power plants are taking the mean load in the electricity generation in Bosnia and Herzegovina. Nowadays, the electricity generation in thermal power plants is app.73% of overall electricity generation. Present growing of electricity consumption demand requires installation a new power capacities very soon. In this light, the multicriteria sustainability assessment based on ASPID methodology is used for the evaluation of sustainability of different options of energy power systems and various sources.
3. Sustainability Assessment of Energy Systems of Various Sources 3.1. Option Under Consideration Eight options for providing the additional power system options within Energy Power System of Bosnia and Herzegovina are considered. Renewable sources and fossil fuel clean technologies are included into consideration. The following options are considered: Option 1: Reconstruction of pulverized coal fired unit in condensing regime; Option 2: Reconstruction of coal fired unit in cogeneration regime; Option 3: Fluidized bed combustion unit-New power station; Option 4: Combined cycle gas turbine power station-New power station; 0 Option 5: Reconstruction of hydro power plant; 0 Option 6: Power plants on solar energy (PV systems)-New power station; Option 7: Wind turbines power-New power units; Option 8: Biomass power plants-New power station. Requirement for all options of energy systems under consideration is production of 7875 GWh of electricity over lifetime of 15 years. This gives average annually electricity production of 525 GWh. Some of energy systems under consideration can produce simultaneously electricity and heat energy (cogeneration), taking into account that this cogeneration plants also must produced 7875 GWh of electricity in projected lifetime.
Sustainability Assessment for the Decision Making
365
Within option 1-Reconstruction of pulverized coal fired unit in condensing regime, the 110 MW pulverized coal fired power plant with slag type boiler furnace is considered, operating in condensing regime-without steam extraction from the steam turbine. Investment at this option is aimed to improve boiler efficiency and upgrade the power. By reconstruction of the boiler, the boiler efficiency is increased from 85% up to 87%; the slag tap furnace is extended by 2 m and new burning system (low NOx burners) is introduced. Additionally, upgrading of the plant is performed through reconstruction of the steam turbine; output power is increased from 110 MW up to 118,55 MW. All these measures affect environment issues, reducing emissions of CO2, NO,, and SO2 per produced kWh. In the option 2-Reconstruction of coal-based unit in cogeneration regime-the 110 MW pulverized coal fired power plant with slag tap boiler furnace operating in cogeneration regime is considered. Additional investment compared to the option 1 is related to the steam extraction control on the steam turbine. Additional 80 MW of heat power installed is provided by steam turbine heat extraction. Option 3 presents a new power station-1 10 MW fluidized bed combustion unit. The power plant will operate in cogeneration regime, producing additional 80 MWh/h of heat. The modern design of the fluidized bed furnace with internal circulation enables burning of low valuable coals, at heat consumption of 8050 kJ/kWh (efficiency of 45%). Option 4 of the considered energy options is combined cycle gas turbine power plant. Installed power of this plant is 102 MW, and it is assumed that all input energy is used for electricity generation. The fuel is natural gas with heating value of 45500 kJ/kg. Estimated heat consumption of the plant is 6545 kJkWh. In the case of option 5-Reconstruction of hydro power plant of 6 x 25 MW power installed, upgrading of the plant by additional 30 MW of installed power is obtained, as well as increasing of efficiency by 4%. This is provided through increasing of design head by 6 m, increasing of installed flow by 10 YO,and introducing new modernized Francis turbines. Option 6 presents solar power plant (PV systems) with total installed capacity of 210 MW. Covered surface of this solar field is 266 000 m2, or 1.26 m2/kW. Option 7 presents utilization of wind energy. According to the available data of wind characteristics on selected micro locations, it is calculated that 438 single units of 600 kW power each is to be installed to provide average annual production of 525 GWh. This gives average annual operation of 2000 h. Within option 8, installing of new biomass power plant is considered. This new plant has 80 MW of installed power, and overall efficiency of 20%. Emision of C02 for this option is reduced by the amount of carbondioxide which is being absorbed by the plants used as fuel in this power station.
3.2. Sustainability Indicators The sustainability assessment is based on the definition and calculation of sustainability indicators. In this work, sustainability indicators reflecting 4 criteria of sustainability are defined and given in Table 1. Indicators from Table 1 are calculated for all options under consideration; on the base of that single criterion assessment can be performed.
366
F. Begic & N.H. Afgan Table 1. Sustainability indicators.
I Tvue of Indicator
I Name o f Indicator
I
unit
I
knlkWh kglkWh kglkWh kglkWh EUlUkWh EUR/kWh l1kWh h/kWh
Fuel indicator
Resource Indicator (RI)
Cupmm indicator
I Environment Indicator (EI)
Economic Indicator (Ecl)
I I
Social Indicator (SI)
Insulation indicator C02 indicator SO2 indicator NOx indicator Energy costs indicator Investment indicator Efficiency indicator Job Indicator Diversity indicator
I I
I
I
I
I
3.3. Results of Single Criteria Analysis The results of single criteria assessment within the analysis are presented. Comparing the considered options according to the environment indicators, e.g. see Fig. 2-indicator of sulfur dioxide emission, it can be noticed that renewable resources, hydropower plant, and clean fossil fuel technologies as Combined cycle gas turbine station and fluidised-bed combustion, are in advantage in reference to the reconstmcted conventional coal-fueled stations.
20 .
15
2
10
.
~
. .
..
..
1 rn . .~ .
5
.
.
.. . .
...
.. ._. . .
0
1
2
3
4
5
6
7
6
0pt ions
Fig. 2. Comparision of SO2 indicators within single criteria analysis
The opposite to this, the comparison of economic indicators, e.g. indicator of investment-see Fig. 3, options of coal-fired power plant is preferable in comparison to the renewable resources. These are obvious examples of unreliability of single criteria analysis, when selection of the most convenient energy system depends exclusively on selected criteria. Consequently, subjectivity of decision-making process is strongly expressed.
Sustainability Assessment,fov the Decision Making
__
8
367
I______
7 r
,
-
................
0 . -
1
2
3
4 5 ODtions
6
7
8
Fig. 3. Comparison of investment indicators within single criteria analysis.
3.4. Multicriteria Sustainability Assessment
The multicriteria sustainability assessment based on ASPID methodology (Analysis and Synthesis of Index at Information Deficiency) [ 5 , 61 of the energy power system under consideration is performed. Used methodology is based on the system of stochastic model of uncertainty, enabling to obtain the assessment from of various supporting systems by computer, as well as enabling to obtain various normalization indexes by using Nonnumeric (ordinal), Non-exact (interval) and Non-complete information (NNNinformation), obtained from sources of various reliability and probability. The multicriteria analysis is based on the determination of sustainability indicators, described in previous chapter, and specific criteria, adopted by weighting factors, and which are being agglomerated into General sustainability indices. General indices are formed through the following procedure: 1. Formation of vectors x = (x, .....)x, of all input attributes (characteristics) which are necessary for full evaluation of quality of the options under considerations; in the work, the attributes are expressed by 4 group of indicators: Resource Indicator, Environment Indicators, Social Indicators and Economic Indicators. 2. Formation of vectors of specific criteria q = ( q l ,........, q m), by which input attributes (indicators) x,, ...,xm are to be evaluated. 3. Introducing of weighting factors, by which multicriteria sustainability assessment of the options under consideration is expressed by means of additive aggregate function, or sintetized function given by relation (1): 4. Selection of vectors w = (y.....w,), wi2 0, wI + ....w m = 1 , i.e. weighting factors. In praxis, vectors w = (w,, .... w,) often can not be exactly determined due to information deficiency. In such a case, method of randomization is used, which enables the values of weighting factors for each considered case to be determined. 5. As final result of this procedure, a list of priority of options under consideration at defined criteria is obtained. The multicriteria assessment taking into account all criteria at the same time, where the different criteria are adopted by respective weighting factors, gives a realistic and reliable sustainability rating of the options under consideration for a lifetime.
368
F. Begic & N.H. Afgan
Fig. 4. Weighting factors for case EcIz > EI2 = SI = RI within multicriteria assessment
Considering the case when the sustainability assessment criteria is based on the economic indicator with domination of investment indicator is preferable in reference to the other indicators, the Fig. 4, it can be noticed that option with the highest rating of the sustainability index is option 5-Reconstruction of hydro power plant. Option 2Reconstruction of coal-based unit in cogeneration regime-is ranked on third place in this case, although this option according to the single criteria assessment is preferable option.
Fig. 5. General indices for the case EcI2 > EI2 = SI = RI within multicriteria assessment.
Additionally, in the case when economic indicator with the domination of the indicator of investment cost, have an advantage in reference to the other indicators, and when the remained indicators are not equal, see Fig. 6, the option 2 falls down on 4thplace in the list of priority, under influence of weighting factors of the other indicators, Fig. 7.
Fig. 6. Weighting factors for case E c l ~ > RI > EIz > SI within multicriteria assessment.
Fig. 7. General indices for the case EcI2 > RI > EI2 > SI within multicnteria assessment.
Sustainability Assessment for the Decision Making
369
4. Conclusions
Most of the models for decision-making for the adoption of power system option are based on single criterion assessment. Presently the sustainability assessment of power system is needed, the models are being advanced and sustainability multicriteria assessment of power system is adapted. However, lack of information doesn’t enable us to obtain a clear assessment of quality of power system option. One of prospective methods to estimate sustainability of the energy system is multicritaeria assessment of sustainability based on the analysis and synthesis of indexes under deficiency of information or ASPID methodology. This methodology is based on the stochastic model of uncertainty, and the assessment from of various supporting systems, to obtain various normalization indexes by using Non-numeric (ordinal), and Non-complete information (NNN-information) obtained from sources of various reliability and probability. Multicriteria sustainability assessment of various energy systems is based on definition of following indicators: resource indicators, environment indicators, social indicators and economic indicators, including also the weight factors. In this paper, the multicriteria assessment of selected options of the energy power system of the Public Enterprise Elektroprivreda of Bosnia and Herzegovina is performed, based on the real and actual example of increasing the energy power system by new power capacity. Through the analysis of multicriteria assessment of selected options, the decisionmakers could be enabled to form opinion related to the selection of an optimum option based on the sustainability assessment. References [I] Begic F. (2003) Method of sustainability assessment of the energy power systems of various sources in Bosnia and Herzegovina. (in Bosnian), PhD theses, University of Sarajevo, Mechanical Engineering Faculty Sarajevo, October. [2] Afgan NH. (2000) Multicriteria indicators for clean air technologies. UNESCO Chair holder, Instituto Superior Tecnico, Lisbon, Portugal. [3] Afgan NH, Carvalho MG. (2000) Sustainable Assessment Method for Energy Systems, Instituto Superior Tecnico, Lisbon, Portugal, Kluwer Publisher. [4] Begic F, Kazagic A. (2002) Sustainability Development of Energy System - Basis of Society Development Strategy (in Bosnian), 5th International Conference “Energetska i procesna postrojenja - Dubrovnik 2002”, Proceedings, Energetika Marketing, Zagreb, May. [5] Hovanov N, Fedotov Yu, Zakharov V. (1999) The making of index numbers under uncertainty. Environmental Indices: Systems Analysis Approach, EOLSS Publishers Co., Oxford, pp. 83-99. [6] Hovanov N. (1996) Analysis and Synthesis of Parameters under Information DeJiciency. St. Petersburg State University Press. (Monograph, in Russian).
