Strategic Closed-Loop Supply Chain Management (Lecture Notes in Economics and Mathematical Systems)

Lecture Notes in Economics and Mathematical Systems 586 Founding Editors: M. Beckmann H.P. Künzi Managing Editors: Pro...

Author: Baptiste Lebreton

24 downloads 549 Views 2MB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

Lecture Notes in Economics and Mathematical Systems

586

Founding Editors: M. Beckmann H.P. Künzi Managing Editors: Prof. Dr. G. Fandel Fachbereich Wirtschaftswissenschaften Fernuniversität Hagen Feithstr. 140/AVZ II, 58084 Hagen, Germany Prof. Dr. W. Trockel Institut für Mathematische Wirtschaftsforschung (IMW) Universität Bielefeld Universitätsstr. 25, 33615 Bielefeld, Germany Editorial Board: A. Basile, A. Drexl, H. Dawid, K. Inderfurth, W. Kürsten, U.Schittko

Baptiste Lebreton

Strategic Closed-Loop Supply Chain Management

With 27 Figures and 23 Tables

123

Baptiste Lebreton (PhD) INSEAD Technology and Operations Management 19, bd de Constance 77300 Fontainebleau France [email protected]

Library of Congress Control Number: 2006935267

ISBN-10 3-540-38907-5 Springer Berlin Heidelberg New York ISBN-13 978-3-540-38907-1 Springer Berlin Heidelberg New York This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. Springer is part of Springer Science+Business Media springer.com © Springer-Verlag Berlin Heidelberg 2007 The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Production: LE-TEX Jelonek, Schmidt & V¨ ockler GbR, Leipzig Cover-design: Erich Kirchner, Heidelberg SPIN 11844174

88/3100YL - 5 4 3 2 1 0

Printed on acid-free paper

To Andrea, for her unbelievable love and patience

Preface

The present PhD thesis is the result of my work as a research assistant at the University of Augsburg, Germany, between 2001 and 2005. Even if a PhD is an individual work, I would like to take the opportunity to thank all the people that have, directly or indirectly, contributed to the completion of this monograph. First of all, I am very grateful to my PhD mentors, Professor Axel Tuma and Professor Bernhard Fleischmann for giving me the opportunity to complete my Masters degree in Germany and for allowing me to stay in Augsburg to write a PhD thesis. Professor Fleischmann’s enthusiasm in solving optimization problems and Professor Tuma’s way of managing a research team have been a great inspiration. I would also like to express my gratitude to Professor Luk Van Wassenhove for his support during the ﬁnal phase of my thesis as well as for the exciting projects I currently work on as a postdoc under his supervision. Research projects with German, French and Belgian companies have highly contributed in the last three years to focus on current problems faced by companies and to keep in touch with real life. Developing applicable models and concepts for their purposes has been challenging but helped me signiﬁcantly to improve this monograph. In this context, I would like to thank Roger Bloemen from Solutia Inc. and Olaf Schottst¨ adt from Knorr-Bremse AG for their trust and openmindedness towards scientiﬁc approaches. Furthermore, I am very indebted to my colleagues and friends (the one seldom excluding the other) that I met during the long sojourn in Augsburg and who made the return to France much more diﬃcult than I expected. Since the list would be very long and I don’t want to forget anybody, the folks that have been invited for a cup of coﬀee at

VIII

Preface

my oﬃce or for a glass of French wine at home should feel concerned by these acknowledgements... I would like to thank my family and especially my parents, Bertrand and Monique, for their unconditional support throughout the years. Raising my children as good as they did with me will deﬁnitively be a challenge... Finally, I am very grateful to my wife Andrea for her love and patience with a husband who makes his job his hobby. Foﬁnha, eu dedico-lhe este livro!

Fontainebleau, June 2006

Baptiste Lebreton

Contents

Part I Setting Up Closed-Loop Supply Chains 1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Sustainable Supply Chain Management . . . . . . . . . . . . . . . 1.2 Outline of This Monograph . . . . . . . . . . . . . . . . . . . . . . . . . .

3 3 8

2

Strategic Aspects of Asset Recovery . . . . . . . . . . . . . . . . . 2.1 Corporate Strategy and Competitive Advantage . . . . . . . . 2.2 Closed-Loop Supply Chains and Competitive Strategies . 2.2.1 Cost Leadership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Diﬀerentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3 Focused Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13 13 17 18 23 27

3

Strategic Impact of Closed-Loop Supply Chains . . . . . . 3.1 Literature Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 A Generic Closed-Loop Strategic Model . . . . . . . . . . . . . . . 3.3 Closed-Loop Supply Chains: Managerial Insights . . . . . . . 3.3.1 The Impact of Green Fees on Asset Recovery . . . . . 3.3.2 Managing the Cannibalization Eﬀect . . . . . . . . . . . . 3.3.3 The Role of Intra-Organizational Incentives Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

31 31 33 39 39 43

Competition in Closed-Loop Supply Chains . . . . . . . . . . 4.1 External Competition as a Signal of Proﬁtability . . . . . . . 4.1.1 Evidence from Current Practice . . . . . . . . . . . . . . . . 4.1.2 OEMs’ Competitive Leverages . . . . . . . . . . . . . . . . . . 4.2 How to Deal with Independent Refurbishers: A Literature Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Analyzing Best Responses Strategies for Manufacturers .

51 52 52 54

4

48

56 58

X

Contents

4.3.1 Competitive Asset Recovery Strategies . . . . . . . . . . 59 4.3.2 Entry Preempting Asset Recovery Strategies . . . . . 63 5

Strategic Network Planning in Closed-Loop Supply Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Strategic Closed-Loop Network Planning: A Review . . . . 5.2 A Generic Strategic Network Planning Model . . . . . . . . . . 5.2.1 The Key Factors of Remanufacturing . . . . . . . . . . . . 5.2.2 Optimization Model . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.3 An Inter-Generational Compatibility Extension . . . 5.3 Extensions to the Generic Strategic Planning Model . . . . 5.3.1 Recovery Path Determination . . . . . . . . . . . . . . . . . . 5.3.2 Location of Recovery Centers . . . . . . . . . . . . . . . . . .

67 71 76 77 81 85 87 87 91

Part II Closed-Loop Supply Chains: Case Studies 6

Tire Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 6.2 Model Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 6.2.1 Demand Segmentation . . . . . . . . . . . . . . . . . . . . . . . . 106 6.2.2 Return Flow Timing and Quantities . . . . . . . . . . . . 108 6.2.3 Reintegration Potential . . . . . . . . . . . . . . . . . . . . . . . . 109 6.3 Optimization Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 6.3.1 Scenario Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 6.3.2 Closed-Loop Supply Chains and Functional Goods 115

7

Computer Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 7.1 The Environmental Challenge of Computer Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 7.2 Model Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 7.2.1 Demand Segmentation . . . . . . . . . . . . . . . . . . . . . . . . 122 7.2.2 Return Flow Timing and Quantities . . . . . . . . . . . . 123 7.2.3 Reintegration Potential . . . . . . . . . . . . . . . . . . . . . . . . 124 7.3 Optimization Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 7.3.1 Impact Assessment of the European Product Stewardship Laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 7.3.2 Impact Assessment of Computer Refurbishing . . . . 129

8

Conclusion and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Contents

XI

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

Part I

Setting Up Closed-Loop Supply Chains

1 Introduction

Museums preserve our past, recycling preserves our future. Theodor W. Adorno

1.1 Sustainable Supply Chain Management The fact that manufacturers have to rethink their supply chain in order to ensure the future of their business has been recognized at least since 1972 and the publication of a Club of Rome’s report entitled The limits to growth (Meadows et al., 1972). The availability of non-renewable resources such as metals or oil is critical for Original Equipment Manufacturers (OEMs) which, generally speaking, need these resources to produce goods. Since OEMs base their business on product selling, the long-term availability of these resources is required for the proﬁt creation to continue. By now, the proﬁt-maximization objective stands in contradiction with the objective of resource conservation. This can be illustrated with the following example: Given a manufacturer with a turnover of X units, a margin per unit of σ − κ1 and a resulting proﬁt of π (eq. 1.1). To sell his products, the OEM requires R non-renewable resources at a consumption rate α (eq. 1.2). Since α is positive, it can be stated that proﬁt maximization implies maximization of the resource consumption (eq. 1.3). max π = (σ − κ) · X where α · X = R max π ⇒ max R 1

σ = retail price, κ = production costs

(1.1) (1.2) (1.3)

4

1 Introduction

However, in order to improve the sustainability2 of their business, OEMs have to reduce the value of their α coeﬃcient. This can be achieved either by reducing the resource consumption (throughput) within the supply chain or by reintegrating already consumed resources into the supply chain. Hence, the consumption rate α can be broken down into the production throughput rate γ and the resource reintegration rate µ (eq. 1.4). α = γ · (1 − µ) α → 0 ⇔ γ → 0, µ → 1

(1.4) (1.5)

Schmidt-Bleek (1998) and von Weizs¨acker et al. (1995) provide examples improving the material intensity per service unit (with a strong emphasize on γ → 0). Nevertheless, the productivity jumps described are diﬃcult to realize without a complete paradigm change towards service selling instead of product selling. Stahel (1986), Giarini and Stahel (1989) and Kostecki (1998) demonstrate in a similar fashion the shortcomings of the product selling concepts illustrated in equation 1.3. Improvements of the γ rate, as depicted in Porter and van der Linde (1995) as well as Romm (1999), are noticeable but not suﬃcient to completely suppress the consumption of additional non-renewable inputs R. As Kopicki et al. (1993) or Thierry et al. (1995) show, an increase of the reintegration rate µ can be achieved at multiple levels also called recovery paths: Product level (reuse, repair), component level (remanufacturing, cannibalization) or material level (recycling). Depending on the recovery path, the reverse ﬂows are processed through ﬁve generic activities: Acquisition, selection, disassembly, cannibalization and mechanical processing (see ﬁg. 1.1). The acquisition process consists of getting the product from the market to the point of recovery. This involves two core activities which are the collection and the procurement process. The procurement process has critical role when the OEM has no property rights on the initial product and the used cores still have a high residual value at the end-of-cycle or end-of-life. Toner cartridges manufacturers are for instance competing against independent remanufacturers cannibalizing their demand for new, more expensive, cartridges (see chapter 4). The role of procurement is also to set incentives to reclaim the valuable 2

Sustainable development is deﬁned as the ability to ”meet the needs of the present generation without compromising the ability of future generations to meet their own needs” (The World Commission on Environment and Development, 1987).

1.1 Sustainable Supply Chain Management

5

Fig. 1.1. Asset recovery processes: Overview (modiﬁed from White et al. 2003)

cores, especially when these are stuck in a retailer’s channel. This situation is particularly critical when an OEM faces distribution returns such as product recalls, unsold items or stock adjustments (de Brito and de Koster, 2004). Blackburn et al. (2004) sustain this point of view by introducing the concept of marginal value of time (MVT). The authors conclude that for goods with a quick residual value loss but high initial value, the reverse supply chain should be reactive (lead time minimizing) instead of functional (cost minimizing). The collection process depends on the organization of the reverse channel (Beullens et al. 2004, Rinschede et al. 1995). On-site collection gives the possibility to manage synergies between forward and reverse distribution since on-site services are often performed by the OEM itself or by sales representatives. Resulting synergies are identiﬁed by Beullens et al. (2004). While on-site collection generally deals with industrial, maintenance-intensive goods, consumer products are often reclaimed through the retailers’ channels, especially when the products are still under guarantee. End users generally dispose of endof-life products to municipal waste systems. In this case, we conclude that manufacturers are not interested in reclaiming their waste ﬂow

6

1 Introduction

since they could have provided incentives to acquire their end-of-life products. This situation can be currently observed in Germany for the electrical and electronic equipment (computers, mobile phones or printers) that are either reclaimed by independent traders for reselling on secondary markets or disposed of. According to Flapper (2003), incentives are the leverage for reclaiming valuable products. The author diﬀerentiates in this context between ﬁnancial and organizational incentives. Financial compensations allow a customer to reduce his ﬁnancial burden, either through a buyback option, rebates for a new product or a cost-free take-back. Buyback options are problematic as they should be set ex-ante although OEMs have no advice about the future value pattern of their products. Flapper points out that customers will make use of this option when they are not able to resell their used product on the secondary market at a higher price than the buyback reward. In other words, if the buyback incentive is set higher than the market price, an OEM will pay more than it would do if procuring on the market. If the incentives are too low, customers will prefer to resell on the secondary market rather than to the OEM. A deposit fee presents similar beneﬁts and shortcomings than a buyback option: Returns will be guided to the secondary or the grey market when the residual value of the product surpasses the deposit fee. Organizational incentives modify the property rights so that customers are obliged to return the used products to the OEM after a given period. In leasing or rental contracts, for instance, the customers pay for using the product instead of buying it. The OEMs involved have therefore the possibility to reduce the insecurity concerning the reverse ﬂow structure (quality, timing, quantity) and are thus able to embed reverse ﬂows into their decision-making process (see Guide and van Wassenhove, 2001 and Fleischmann, 2001 for the impact of insecurity on planning). Once the cores have been returned, the selection process takes place in which the valuable products are identiﬁed and guided to one of the recovery processes or directly resold. We notice that the term ”valuable” also requires to know whether a demand for this product or its parts exists and if the upcoming recovery costs do not surpass the procurement savings for new components. This decision implies a good coordination between procurement (how many parts are required to match the demand?), inbound logistics (is there any part on stock?), marketing (is there currently a demand for the ﬁnal product?) and service (is there a demand for the recovered spare parts?). The ability to manage this information ﬂow may become a core competence for

1.1 Sustainable Supply Chain Management

7

an OEM as it ﬁts in with the criteria of Prahalad and Hamel (1990): Valuable (because it sustains a competitive edge), knowledge-based and hardly imitable for competitors. After having ﬁltered out the recoverable items, an OEM may choose between two alternatives for closing the loop at a component level. Thierry et al. (1995) diﬀerentiate in this case between disassembly and cannibalization. Disassembly consists of removing all the parts from a returned product and reintegrating them into the assembly process whereas cannibalization is equivalent to a very selective disassembly: Only the valuable parts are removed while the residual product is sent to the mechanical processing step. The mechanical processing step encompasses the shredding of residual products or components and the sorting of the resulting material fractions. The shredded residual that cannot be reintroduced as material into the supply chain is either landﬁlled or incinerated. Nevertheless, despite the necessity of resource reintegration, closed-loop supply chains, i.e. supply chains simultaneously carrying both forward and reverse ﬂows, are very seldom run by Original Equipment Manufacturers. Guide (2000) estimates for instance that 95% of the remanufacturing3 programs are not managed by the original producers. To the knowledge of the author, this ﬁgure should be even inferior for recycling programs due to the low level of vertical integration of manufacturers.4 In this context, it is not surprising to see OEMs being compelled by legislators to at least ﬁnance the reintegration of resources and thereby to increase their own µ rate. Table 1.1 gathers the current environmental legislation setting reintegration rates. We notice that the most recent legislation: The Waste Electrical and Electronic Equipment (WEEE) and the End-of-Life Vehicle (EOLV) directives set only material reintegration (recycling) targets and do not mention any product or component reintegration target. This apparent discrepancy between the critical role of long-term resource availability for a ﬁrm’s success and the low attention manufacturers have paid to this problem by now is the initiator of this monograph. According to van Wassenhove and Guide (2005), the reverse logistics research does not provide currently insight into this contradictory situation due to the research emphasis on operational and 3 4

We will consider remanufacturing and refurbishing as synonym in the following. German car manufacturers generate for instance only 25% of their products’ value (VDA, 2002) so that an improvement of the recyclable fraction of their products does not impact directly an OEM’s proﬁts.

8

1 Introduction Table 1.1. European and German asset recovery legislation Year Law Recovery target (in % of weight) 2002 directive WEEE computers, tv sets: 0.65, 0.8 2000 directive EOLV 0.85 (until 2005) then 0.95 (after 2005) 1998 BattV (batteries) none, but reverse chain setup 1991 VerpackV (packaging) glass: 0.75, paper: 0.7, aluminium: 0.6 – tires no legislation up to now

tactical planning problems. The following analysis, however, states that the setup of closed-loop supply chains will be driven by strategical issues in the near future. As a result, the present monograph, entitled Strategic Closed-Loop Supply Chain Management, aims at providing methods to assess the strategic relevance of asset recovery and at developing methods to determine the optimal setup of closed-loop supply chains. For the sake of practical relevance, we will align the structure of the monograph to the decision-making process of Original Equipment Manufacturers willing to investigate the potential of circular supply chains. This decision-making process will be structured around fundamental questions managers are expected to answer prior to running a sustainable supply chain. The answer to these questions will give us the possibility to present the existing research from a practitioner’s viewpoint as well as to develop, if necessary, applicable concepts for manufacturers.

1.2 Outline of This Monograph The ﬁrst part of the monograph is dedicated to the decision-making process leading to the setup of a closed-loop supply chain by a generic Original Equipment Manufacturer. In accordance to the strategic management literature (Porter, 1996, Christopher, 1998), the ﬁrst question managers will have to deal with can be formulated as follows: Does a closed-loop supply chain ﬁt with our corporate strategy? Compaq provides a good example for illustrating the pertinence of this question. Before its merger with HP, Compaq used to run an asset recovery center (AMRO) in the United States which has been closed down although it turned beneﬁts. Sarkis (2003) notices ”As of winter of 2000, Compaq decided to close down the AMRO facility and outsource

1.2 Outline of This Monograph

9

the demanufacturing function to DMC [...] Even though the facility was proﬁtable, it did not meet Compaq’s strategic vision [...] they decided to focus internal resources elsewhere.” Fujitsu-Siemens Computers and the BMW Group also run own recovery centers in Paderborn (FSC) and Munich (BMW). However, these centers are too small to impact on a ﬁrm’s operating results and exist mostly because of marketing reasons and to underline the environmental goodwill of these companies. As the outcome of these plants is not reintegrated into their own supply chain, these manufacturing network cannot be considered as being ”closed”. On the other side, some manufacturers of complex products have set up reverse supply chains and recover value with success from their returned products. The recovery activities of copier (Oce, Xerox), electrical equipment (OMRON, see Kuik et al., 2005) or tire manufacturers (Michelin) are closely linked to their forward supply chain and are of crucial importance for their operating proﬁts. As the share of refurbished products in the copiers’ and truck tires’ total sales often rises above 30%, the terminology closed-loop supply chain seems to be adequate to depict the recovery activities of these companies. Obviously, asset recovery is in this case considered as a strategic competitive driver for these companies. Porter (1996) deﬁnes strategy as a set of coherent measures necessary to reach above-average proﬁtability. He notices that companies should either follow a cost leadership strategy or provide a unique utility to customers for which these are ready to pay a premium (diﬀerentiation strategy). The decision-making process should therefore be concerned with the development of a competitive advantage over competitors. Hence, chapter 2 will investigate to what extent closed-loop supply chains might contribute to the achievement of a competitive strategy. We will show that the integration of asset recovery requires a coordination of all the activities within the value chain to turn the reverse logistics eﬀorts into cash. The identiﬁed strategic ﬁt of asset recovery will trigger a further question: Is it proﬁtable to run a closed-loop supply chain? Original Equipment Manufacturers are compelled to create value for their shareholders. As a consequence, the strategic ﬁt is a prerequisite but is not suﬃcient to justify the setup of asset recovery activities since managers have to understand the drivers determining the proﬁtability of resource recovery. Given this, chapter 3 develops a mathematical closed-loop model which purpose is to catch the dynamics of resource reintegration while remaining at an aggregated level. This simpliﬁcation concern is justiﬁed by the fact that manufacturers often do not have

10

1 Introduction

access to reliable past data to investigate the ﬁnancial impact of a closed-loop and will therefore give up their investigations at this level. The model presented in this chapter aim to tackle this hurdle as it relies on generally available ﬁgures to provide insights into the ﬁnancial impact of closed-loop supply chains. Despite its simple structure, the model enables us to draw conclusions about current research questions in the reverse logistics ﬁeld: 1. When is it worthwhile for an OEM to recycle on his own? Does the legislation provide enough incentives to do so? 2. How important is the cannibalization eﬀect between new and recovered products after the introduction of a refurbishing program? Is it proﬁtable to refurbish despite this cannibalization eﬀect? 3. How should incentives be aligned within the organization to ensure that a sales division supports the reuse or remanufacturing of its products? Although remanufacturing would be generally proﬁtable for an OEM, some companies (HP, BrB5 ) have introduced an active product return management but have chosen to destroy their returned products instead or refurbishing them. In this context, the following question has to be investigated in order to understand the reasons of this behavior. When should manufacturers destroy their reclaimed products instead of refurbishing them? We will show in chapter 4 that this behavior stems from the competition of independent free-riders refurbishing the products of the manufacturer. To catch the dynamics of this competition, the strategic closed-loop model of the previous chapter will be extended to competitive issues. With help of the resulting mathematical model, we will identify the market patterns by which core destruction is a better issue than remanufacturing. Once the strategic relevance and proﬁtability of closed-loop supply chains have been recognized by managers, these will have to set up a reverse supply chain and to specify the interfaces with the current value chain. As a consequence, the decision-makers will have to solve the following questions: • Which products should be reintegrated, • which level of reintegration is proﬁt-maximizing, and ﬁnally • where the recovery activities should be performed. 5

See chapter 4 for a comprehensive case description of both companies. BrB is an anonymized brake systems manufacturer.

1.2 Outline of This Monograph

11

These decisions imply the realization of investments to run a recovery network eﬃciently. Due to the long-term impact of such investments, information has to be gathered and structured in order to support the decision making process. Domschke and Scholl (2003) depict this business phase as the planning process. The complexity of supply chains requires the use of planning models abstracting from the reality in order to ensure that single investments also match a manufacturer’s corporate objectives. Chapter 5 will present a strategic closed-loop planning model with initial focus on reuse and remanufacturing. After a comprehensive literature overview of the application of strategic network planning to reverse logistics issues, this basic model will be stepwise extended to take the following business aspects into account: 1. Inter-generational component compatibility. 2. Investments in recovery assets (for both remanufacturing and recycling). 3. Plant location and competition impact. After having provided the theoretical background related to the setup of a closed-loop supply chain, the second part of this monograph will apply the concepts of part 1 to real life case studies. The research goal is thereby to provide insights into the practical implementation of strategic network planning models. Chapter 6 will make use of the initial strategic planning model to investigate the proﬁtability of tire retreading. As the remanufacturing rates between passenger car tires and truck tires strongly diﬀers, the planning model will give us the opportunity to identify the reasons for this discrepancy. On a long-term perspective, the recycling of materials might also turn out to be proﬁtable for Original Equipment Manufacturers if the raw material prices continue to soar. Chapter 7 will investigate whether it is proﬁtable for a computer manufacturer to build asset recovery centers to recover products, components or materials. In this case, we will apply a discounted cash-ﬂow approach to assess the return on investment of asset recovery centers. Finally, chapter 8 will summarize the managerial insights obtained from the theoretical framework of part 1 as well as from the two case studies to provide an outlook for further research.

2 Strategic Aspects of Asset Recovery

2.1 Corporate Strategy and Competitive Advantage As stated in the last section, the main goal of private-owned companies is to maximize their proﬁts while ensuring their long-term existence. This proﬁt issues from proﬁtable sales, i.e. sales by which the price paid by the customer (σ) has been superior to the total costs of the product (κ). In a competitive market, however, the question remains why a customer should buy an item from one speciﬁc producer rather than from the others. The economic theory works this problem out by arguing that every customer seeks to maximize his own proﬁt π C which represents the diﬀerence between the utility he obtains through the product Up and the acquisition costs σp (Phillips, 2005). Given this, a simpliﬁed procurement function of a customer can be expressed as follows: max π C = max{Up − σp } p

(2.1)

According to Kotler (1999), the customer utility Up can be broken down into three levels (see ﬁg. 2.1): The core product, the actual product and the augmented product. While the core product, which encompasses its main purpose, does not provide room for diﬀerentiation, the actual product makes it possible for a ﬁrm to increase a product’s utility by adding functional features or by creating a psychological reward for the buyer (for instance through styling, packaging or brand name). In other words, the actual product’s properties help fulﬁlling the higher levels of Maslow’s pyramid of needs (1970). Finally, manufacturers may increase the utility of their products by providing additional services around the core product: Free product installation, reliable availability of replacement parts or for instance quick maintenance services. Such

14

2 Strategic Aspects of Asset Recovery

additional services without psychological references, which aim to create value for the customer by reducing their total life-cycle costs, are gathered by Kotler into the augmented product.

Fig. 2.1. Kotler’s product levels (1999)

In a competitive environment, the success of a ﬁrm depends more of the value π C than on the utility created for the customer. In a market where all producers supply the same level of perceived utility, the competition turns out to be price-driven. The manufacturers will have to set their prices under the competitor’s prices in order to convince customers. In a long term, the retail price σp will only cover the production costs κp , thereby suppressing any proﬁt. The ability to generate a superior value π C for the customer while keeping an above-average proﬁtability (σp κp ) is what Porter calls a competitive advantage. According to Porter (1996), the goal of a strategy is to create, maintain or increase a durable competitive advantage against competitors. In this context, Porter identiﬁes three generic strategies with respect to the strategic focus and the customer scope (see ﬁg. 2.2): • Cost leadership. This strategy consists in being the cheapest supplier over a broad range of market segments. The ﬁrm’s objective is to provide a standard product fulﬁlling only the core utility and thereby to make use of economies of scale and learning eﬀects. A cost leader focuses his resources on measures improving his current productivity level. Despite his cost advantage, the perceived quality of the products should not be signiﬁcantly below the market aver-

2.1 Corporate Strategy and Competitive Advantage

15

age. If the quality gap is too signiﬁcant, customers will expect price rebates that might overcompensate the cost advantage. • Diﬀerentiation. A company acting as a diﬀerentiator provides product characteristics for which customers are willing to pay a premium. As long as these product properties are perceived as unique by the buyers, the ﬁrm is able to charge more than competitors. However, the ﬁrm has to keep its manufacturing costs close to the market average in order to generate a better return on sales than his competitors. Unlike the cost leadership strategy, the diﬀerentiation strategy can be based on several product aspects of the actual and augmented product, thus making room for more than one competitor. • Focus. Instead of trying to supply the whole market, a ﬁrm may focus his eﬀorts on a particular segment that global competitors cannot supply (for examples, see Porter, 1985). A focused ﬁrm takes advantage of the lack of specialization of global competitors. Focused manufacturers can either be cheaper than all-purpose suppliers by removing product functionalities expected from the majority of the customers, or be better positioned by oﬀering targeted product properties which would be too costly for the mass market.

Fig. 2.2. Generic competitive strategies (Porter, 1980)

Porter (1985) observes that ﬁrms trying to obtain several competitive advantages simultaneously have to take contradictory measures and, as a result, remain ”stuck in the middle”. For instance, a differentiation strategy requires e.g. the integration of costly features to

16

2 Strategic Aspects of Asset Recovery

the core product, above-average advertising expenditures or the use of high quality components and materials to ensure a remarkable durability. While customers sensible to the attributes of the actual product are keen on buying this diﬀerentiated commodity, the product will not be chosen by price-sensitive customers putting emphasis on the core utility. These will prefer inexpensive oﬀers without additional features or service. Ries and Trout (1993) observe that customers will prefer manufacturers with a clear positioning. These will be either unique in their prestations and will generate an additional, hard-to-imitate, utility for the customer or will be the least expensive in fulﬁlling the core purpose of a product. Firms stuck in the middle will, according to Porter, neither generate the utility of perfectly diﬀerentiated manufacturers nor be able to produce as cheap as cost leaders. Their proﬁts will be below the market average. To sum up, the role of a competitive strategy is to force managers to commit themselves to one of the three generic strategies and to take coherent decisions in order to generate, keep or increase their competitive edge (Porter, 1996). In another publication, Porter (1985) investigates the origins of competitive edges. In a similar fashion as the resource-based view theory, he postulates that a competitive advantage stems from the way activities are processed within the ﬁrm. For this purpose, he identiﬁes ﬁve sequential core activities: Internal logistics, manufacturing, external logistics (i.e. distribution), marketing and sales as well as services. The marketing and sales activity contributes to the creation of Kotler’s augmented product while the service process is dedicated to the creation of actual product characteristics. Every core activity consumes four kinds of resources managed by the support activities: Materials (procurement), workforce (human resource management), information (ﬁrm infrastructure) and equipment (technology development). The linkage between these activities, known as Porter’s value chain, is illustrated in ﬁgure 2.3. Porter’s value chain highlights the fact that a strategy has to be concretized at the process level because a company’s competitive advantage has its origin in the way the core and support activities are performed and coordinated. Taking advantage of economies of scale is one of the means to achieve a cost-leadership strategy. For instance, standardized products with few variations enable OEMs to obtain better conditions on the components’ prices and to beneﬁt from higher learning eﬀects. The minimization of the time-to-cash cycle is another means to improve an OEM’s cost position since it reduces the amount of inventory required among the supply chain. The reduction of a ﬁrm’s

2.2 Closed-Loop Supply Chains and Competitive Strategies

17

Fig. 2.3. Value chain according to Porter (1985)

lead times also necessitates the coordination of several activities for an eﬀective implementation: Drastic reduction of setup times (manufacturing, technology development, human resources), supply chain planning systems and lean management (all activities). A diﬀerentiation strategy with focus on customers’ costs minimization also implies a set of coordinated measures depending on the customers’ value chain. In the case of industrial equipment, the spare parts availability during the whole product life cycle will be decisive due to the opportunity costs of production breakdowns. The minimization of the total costs of ownership is the mean buying argument for industrial customers and is achieved by aligning the technology development, procurement and service activities.

2.2 Closed-Loop Supply Chains and Competitive Strategies As mentioned before, asset recovery will only be embedded into the strategic planning process when its contribution is relevant to the achievement of a competitive edge. Nevertheless, this integration should be coupled with a coherent set of measures within the value chain in order to reclaim the beneﬁts of asset recovery. Hence, the asset recovery processes will also have to be integrated into the current value chain of the company. The present section will therefore investigate how to ﬁt

18

2 Strategic Aspects of Asset Recovery

the value chain to reach a competitive advantage with help of closedloops. 2.2.1 Cost Leadership Until now, the existence of closed-loops supply chains stems mostly from the savings generated on the procurement of modules and materials. This situation is particularly obvious when the remaining value embedded into the reverse ﬂows amounts for a signiﬁcant fraction of the initial input costs, as e.g. in the tire industry (see chapter 6 or IFEU, 1999) or in the paper industry (Braunmiller abd W¨ ohrle, 1999). Ayres et al. (1997) speak of double dividends when the reuse of resources is both economically successful and ecologically sound. According to Porter and van der Linde (1995) and Romm (1999), the relevance of resource eﬃciency to support a cost leadership strategy will grow in the future as the scarcity of the resources increases. Unfortunately, the double dividends assumption does not always hold, especially in the computer industry. Grote (1994) performs a simpliﬁed life-cycle assessment for computers (see table 2.1) in which comes out that the resource consumption of the extraction and the production phases can be partly avoided by remanufacturing computers. Table 2.1. Energy consumption and CO2 emissions in a computer’s life-cycle, Grote (1994) Extraction Production Use Recovery Energy (in KWh) 2,325 3,010 500 15 CO2 emissions (in kg) 1,400 1,815 0.3 4

However, the current remanufacturing rate in the computer industry is below 1% (see chapter 7) because computer manufacturers do not consider closed-loop supply chains as a strategic alternative despite the evident resource spoilage. As a consequence, the legislator has passed environmental laws (WEEE, ROHS) to internalize at least a fraction of the environmental costs to the computer manufacturers. Thus, the increase of the disposal costs might become a reason for OEMs to start closed-loop supply chains: An integration of the recovery activities could turn out to be cheaper than the passive ﬁnancing of the collection and recycling processes by external ﬁrms. Chapter 7 will analyze this assumption for the computer industry.

2.2 Closed-Loop Supply Chains and Competitive Strategies

19

Spengler and Schr¨ oter (2005) denote the potential of parts cannibalization in reducing storage costs for the end-of-life service. In fact, even after a product line has been discontinued, the OEM is compelled by law to provide spare parts during a given lap of time. The determination of a last lot is a crucial issue because of the costs for over- and understocking (Teunter and Fortuin, 1999). Spengler and Schr¨ oter (2003) and Fleischmann et al. (2003) present practical applications where spare parts are recovered from used products if there is a demand for these stock keeping units. This ﬂexible replenishment strategy reduces the size of the last lot and may avoid the setup of an additional lot in the future. End-of-life reverse ﬂows are not the only potential source of savings. Commercial returns are becoming a preoccupying issue as customers tend to give their products back as soon as these no longer meet their expectations (see Davey et al. 2005, Kostecki, 1998). In general, industry competitors face an equivalent return rate within the warranty time (Rogers and Tibben-Lembke, 1999). The way ﬁrms deal with such perishable items is decisive1 because these returns induce additional costs. Asset recovery revenues (e.g. through reuse or savings) give the possibility to compensate reverse logistics costs and, consequently, to improve the positioning of a cost leader. Since these cost savings opportunities have to be concretized in the value chain, we now need to analyze the required measures to be taken concerning the support and core activities in order to exploit the savings potential. Support Activities Preliminary to the launch of a closed-loop supply chain, company-wide measures have to be taken in order to ensure the proﬁtability of resource reintegration on a day-to-day basis. As chapter 3 will show, the acceptance of reverse ﬂows as a potential rather than as a burden requires to change the behaviors within the company and especially the unique priority given to the forward ﬂows. First, the upcoming organizational changes consequently to the introduction of asset recovery necessitate a credible commitment from a company’s top management.2 As more than 80% of a product’s costs are already set before the production 1 2

We refer to Blackburn et al. (2004) for a detailled analysis of the residual value loss patterns. We refer to Harvard Business School (2003) for a broader description of change management processes, especially at the workforce level. Thierry (1997) describes extensively the organizational changes subsequent to the introduction of a closedloop supply chain at Xerox.

20

2 Strategic Aspects of Asset Recovery

launch, the ﬁrst changes within the corporation will concern the product development activities of the company (Coenenberg, 2003). With help of illustrative examples, Navin-Chandra (1994) and Penev and De Ron (1995) show the impact of design decisions on the eﬀorts necessary to reclaim the value embedded in a returned appliance. On the basis of these ﬁndings, the German engineering lobby (VDI) has published a guideline, called Design for Recycling, VDI 2243, which gives advice on design measures apt to increase the proﬁtability of resource recovery (VDI, 2002). The main product design measures consists of: • Restricting the range of materials used, • Enabling easy separation of the components and materials, • Providing product-inherent information in order to facilitate the product handling and to accelerate the disassembly process. The design for recycling (DfR) guideline is focused on minimizing the processing costs for the disassembler and thus seeks to minimize the burden consequent to the recycling laws mentioned in the introduction. Nevertheless, a closed-loop supply chain requires eﬀorts from the support activities that go beyond such product adaptations. In fact, the contribution of the procurement and of the technology development activities is decisive for creating value from product returns. On the one side, corporate buyers have to give emphasis on a component’s quality and its suitability for refurbishing operations instead of focusing exclusively on the initial procurement cost. On the other side, these should be aware of the reintegration potential of the component in order to avoid paying to much for useless product longevity. In this context, a detailed discussion about the impact of higher remanufacturability levels will be provided in the next chapter of this monograph and thus provide insights into the trade-oﬀ between extended servicelife and its related costs. As we will see in chapter 5, the successful reintegration of used components also depends on the technical compatibility between older and brand new parts. Ferrer (2001) and Krikke et al. (2004) point out that components’ modularity is a prerequisite for high range remanufacturing programs. The insecurity related to the timing, quantity and quality state of reverse ﬂows turns out to be a challenge related to the information available in the forward and reverse pipeline. Fleischmann (2001) notices that this insecurity increases the planning complexity of the supply chain and therefore leads to higher processing costs. Especially the erratic supply of refurbishable cores induces increased inventory levels of both new and recovered parts and generates hidden adaptation

2.2 Closed-Loop Supply Chains and Competitive Strategies

21

costs within the organization. Hence, the OEM might take advantage of resource recovery when the product returns and their quality are known in advance. Given this, Hall (2001) estimates that the success of remanufacturing is conditioned to the existence of leasing programs suppressing this source of uncertainty. Despite the availability of informations related to the reverse ﬂows, a cost-leader ﬁrm needs to ensure that the eﬀorts spent in disassembling selected parts pay oﬀ, i.e. that a demand exists for such a speciﬁc used part. Thus, an integration of the reverse ﬂows into the current supply chain planning systems is necessary to match supply and demand of cores. Unfortunately, despite intensive research eﬀorts on the extension of MRP II systems towards product recovery (see Rautenstrauch, 1997 and Guide, 2000), Rogers and Tibben-Lembke (1999) as well as Chouinard et al. (2005) observe that current information systems seldom store the data concerning product returns, especially because returns are not a corporate priority and are, for this reason, stored in ad-hoc local databases without link to the main ERP system. By doing this, manufacturers not only hinder themselves starting business relevant recovery programs but also renounce to learn from product returns. According Davey et al. (2005), the HP printer returns are in 70% of the cases not related to manufacturing defects but are returned for a wide range of convenience reasons: Opportunistic behavior of the customer, setup diﬃculties or stock surplus of retailers. Without integrated information systems, HP would not be able to understand the reasons of the majority of their returns and therefore could not quantify the beneﬁts of improving the installation routines of his printers or the shadow costs of a liberal product returns policy accorded to the retailers. Klausner and Hendrickson (2000) demonstrate the impact of a data logger containing all the relevant information concerning the service life at the point of return. Depending on the previous utilization of the returned product, the motor is removed and reintegrated in product lines of diﬀerent quality. Core Activities In addition to the structural measures described above, a manufacturer has to adapt the current organization of his core activities to reclaim the previous investments in product development and information technology. The launch of remanufacturing programs begins with the setup of an active product returns management as described in Thierry et al. (1995). Depending on the contractual agreements with customers, a manufacturer will either rely on buyback incentives (in

22

2 Strategic Aspects of Asset Recovery

case of product selling) or on leasing to reclaim the proﬁtable products and guide the other ones into cost-minimizing processing chains. Souza et al. (2003) notice that manufacturers should run a reactive reverse chain to reclaim the value of the returned products, especially because of the high depreciation rates. Spare parts management, which is a task of the core activity service, plays a role in the product recovery process as this activity is at the interface between the customer and the supply chain. Worn parts can be collected during maintenance jobs and, in some cases, replaced by refurbished parts (see for instance Mercedes Benz, Driesch et al., 2005). Concerning the long-term supply strategy of spare parts, end-of-life product returns represent an inexpensive supply source as a manufacturer might reduce his working capital during the maximal legal warranty time (see Spengler and Schr¨ oter, 2003, 2005). Once the end-of-cycle and the end-of-life products have been identiﬁed and reclaimed, the disassembly process takes place. With help of information technology and a disassembly-friendly product architecture, workers remove the valuable parts in a short lap of time and without damages. After a functionality test, the parts are either stored for an imminent reintegration or directly forwarded to the production process. The integration of reverse ﬂows into the corporate information systems enables the company to learn from the previous returns and, for instance, to give up previous refurbishing tasks when the reclaimed parts turn out to fail during their second life-cycle. Subsequent to the product refurbishment process, the incentives between sales divisions have to be aligned in order to ensure that the recovered products, which represent a cannibalization danger for the lines of new products, are enough marketed to be sold at satisfactory prices. Chapter 3 will investigate transfer price schemes able to align incentives between sales divisions and ensure that the divisions selling new items also support remanufactured products. The set of measures listed above aims to show the necessity to take coherent decisions within the complete value chain in order to ensure that closed-loop supply chains also lead to the achievement of a costleadership edge. While cost savings are the main driver for starting a product recovery program, a coherent extension of the current information systems, design for recycling guidelines and a strong commitment of the top management are decisive factors to minimize the shadow costs of closed-loop supply chains. Figure 2.4 gives a graphical overview of the required measures to achieve a cost-leadership value chain on the basis of asset recovery.

2.2 Closed-Loop Supply Chains and Competitive Strategies

23

Fig. 2.4. Cost leadership with asset recovery: Value chain

2.2.2 Diﬀerentiation While a cost leadership strategy seeks to minimize the cost structure of an OEM, a competitive strategy based on diﬀerentiation seeks to provide unique characteristics for the customers. Closed-loop supply chains can contribute to provide two potential diﬀerentiation drivers with an industry-wide scope: Take-back services and quality. Take-Back Services Apart from cost reductions, a manufacturer is able to provide valuecreating services for customers depending on the timing of the return ﬂows. In the European Union, the most obvious diﬀerentiation driver is issued from end-of-life reverse ﬂows, customers in the United States pay more attention to liberal return policies for products under guarantee and end-of-cycle products (Kopicki et al., 1993, Rogers and TibbenLembke, 1999). This behavioral diﬀerence stems from the existence of more restrictive environmental laws in Europe. The recent European product stewardship laws (WEEE, EOLV) not only internalize the reclaiming and recycling costs to the manufacturer but also indirectly regulate the return channels. As a result, customers face technical and ﬁnancial hurdles to get rid of their products: End-of-life products have to be delivered to collection points and customers have to pay a processing fee to dispose of their appliances. These hurdles signiﬁcantly extend the market sojourn of electrical and electronic items (see ﬁg.

24

2 Strategic Aspects of Asset Recovery

2.2). Although the majority of computers has an expected service life of four to ﬁve years, these remain at least three more years stored before being disposed of. Table 2.2. Use pattern estimation for personal computers, Williams and Kuehr (2003) Lifetime Years Reuse Recycle Stockpile Landﬁll 1st cycle 4 45% 5% 45% 5% 3 x 20% 70% 10% 2nd cycle stockpile 3 x x 20% 80%

Original Equipment Manufacturers might beneﬁt from the lack of convenience of current collection systems by providing a take-back service for end-of-life products. Investigations of Schulz (2004) document that customers are willing to pay up to hundred Euro in order to avoid the administrative deregistration steps of a vehicle. A similar service is also commonly provided by retailers of bulky consumer goods such as refrigerators or furnitures when a new appliance is procured (Herold and Kovacs, 2005). Product returns, regardless of their nature, generate additional costs. Hence, a liberal return policy might be costly but presents a higher utility for buyers of goods whose quality can only be estimated after the purchase.3 Rogers and Tibben-Lembke (1999) perform a survey in which manufacturers consider reverse logistics and more liberal return terms as a means to retain buyers and increase the customer lifetime value. Nevertheless, promoting product returns might become problematic due to the danger of abuse (Guide and van Wassenhove, 2003). Promoting a liberal return policy within a diﬀerentiation strategy turns out to be a very challenging option that should be followed when the chances of reuse are good. The next chapter will provide insights into the ﬁnancial impact of convenient product returns with help of an strategic assessment model. Quality Diﬀerentiation through Brand Protection In the last years, some ﬁrms, amongst others the HP toner cartridges division or the German brake systems manufacturer BrB, have engaged into an active product return management targeting the protection of 3

Kotler (2003) speaks of goods high in experience qualities.

2.2 Closed-Loop Supply Chains and Competitive Strategies

25

their brand’s quality reputation. Both ﬁrms diﬀerentiate themselves by providing high printing quality (HP) or technologically advanced security features (BrB) but have faced competition of remanufacturers supplying refurbished items at low cost. Reclaiming used cores at the end of the life-cycle has turned to be a necessity in order to avoid this competition.4 Despite the success of refurbished cartridges (remanufacturers cannibalize HP’s product lines), HP prefers to shred returned cores in order to reintegrate the materials (plastics and metal) into his own production process. This destruction of potentially reusable items can be justiﬁed since the higher prices of new appliances (in this case cartridges) are expected to give a signal concerning the quality of a product (Leavitt, 1954). Brucks et al. (2000) explain this situation with the bounded rationality of buyers. As potential buyers are not able to take into account all product characteristics, they tend to reduce the choice complexity by emphasizing one easily measurable criterium which is the retail price.5 The competition through cheaper refurbished items with the same branding obliterates this signaling and presents a threat to the competitive positioning. This competition preemption strategy and its public marketing have the objective to demonstrate that HP’s customers do not deserve the minor quality provided by reﬁlled cartridges. The case of BrB diﬀers because of the safety issues related to brake systems. BrB started an exchange program for worn brakes in 2003 after having observed the development of free-riders cleaning and reselling still worn brakes at half the original price. Although BrB’s product stewardship would not cover accidents due to defective brakes, the risk of negative publicity would be too high in case of an accident and forced BrB to dry out the grey market. Impact on Value Chain Unlike cost leadership strategies, the two diﬀerentiation strategies presented above do not require such an integrated matching between the value chain activities. In fact, applying all cost-leadership measures proposed previously in order to achieve a quality diﬀerentiation strategy would lead to an over-investment situation. The take-back strategy for instance implies two diﬀerent implementations depending on the type 4 5

The competition between OEM and remanufacturers will be investigated in detail in chapter 4. We refer to Rao and Monroe (1989) for a comprehensive literature overview on the correlation between price and a customer’s quality perception.

26

2 Strategic Aspects of Asset Recovery

of reverse ﬂows. Warranty returns represent a signiﬁcant burden because the original equipment manufacturer has in general to replace the defective product by a new one without obtaining any additional revenue. As a consequence, the ﬁnancial burden related to take-back options will decrease as soon as the producer is able to reclaim more value from these returned items. The focus of the value chain should be, to some extent, similar to the focus of cost-leadership strategies. As the reintegration of parts or products remains an exception rather than the rule, the focus of the technology development activities should not spend signiﬁcant eﬀorts enforcing product modularity. The introduction of end-of-life take-back programs as a diﬀerentiation driver leads to a slightly diﬀerent implementation focus as the manufacturer is concerned with servicing a customer who wants to gets rid of a product without residual value.6 Hence, remanufacturing is excluded and the marginal value of time is low. Design for recycling (task of the technology development activity) provides the leverage to minimize the costs of such convenience services. The use of valuable and separable materials (scope of the procurement activity) lowers the reintegration costs and, depending on the importance of the embedded metallic fraction, might also enable a recycler to turn proﬁts out of the recovery process (as e.g. in the car industry, see Schulz, 2005, or in the computer industry, see Pepi, 1998). The quality diﬀerentiation strategy relies, in a closed-loop context, on free-riders preemption to protect the brand. Thus, the value chain implementation should focus on generating means to avoid the presence of external remanufacturers. First, the marketing activity needs to convince customers of the minor quality of remanufactured items. Furthermore, the refurbishing costs carried by the free-riders need to be increased in order to hinder refurbishers making money out of a manufacturer’s appliance. Lexmark, a toner cartridges OEM, has introduced a proprietary chip forbidding any service life extension (Majumder and Groenevelt, 2001). BrB produces smaller series of replaceable parts to increase the tooling costs of refurbishers. Chapter 4 will provide further details on the eﬀectivity of several manufacturers’ preemptive strategies. Despite the role of technology in preempting competition, the technological means should be aligned with the procurement strategy in order to impact the external competitors. The case of copier manufacturers is thereby illustrative of a missing integration: The development of toner cartridges has been mainly outsourced to a small range of worldwide suppliers who now provide the parts for the orig6

A positive residual value is a sign that a demand exists for the product concerned.

2.2 Closed-Loop Supply Chains and Competitive Strategies

27

inal equipment manufacturers as well all their competitors, including the refurbishers. As a result, the quality of the refurbished cartridges is equivalent to the brand new ones (see chapter 4) and the proprietary technologies in use have always been neutralized until now. While BrB has successfully dried out the grey market of used brake systems, HP’s competitors are still performing well. We will show in chapter 4 that printer manufacturers should remanufacture on their own to maximize their corporate proﬁts. 2.2.3 Focused Strategies On the basis of the case studies available, we identify two potential focused competitive strategies making use of closed-loop supply chains: The presence on secondary markets and the diﬀerentiation through a green image. Cost Leadership: Secondary Markets Commodities with a long service life often change hands. These used items are traded in secondary markets in which buyers have smaller budgets. In general, it is possible to segment a market into a premium segment giving emphasis on Kotler’s actual product characteristics, and a low-end segment more interested in a product’s core utility. In the computer case depicted in table 2.2, desktop computers may yield a second life-cycle dedicated to run oﬃce applications. As OEMs base their business on product sales instead of service sales, they transfer all property rights to their customers.7 Buyers are free to decide when and to whom their goods should be resold. As a consequence, OEMs do not control the secondary markets and forgo potential sales opportunities by price-sensitive buyers looking for a simple home oﬃce desktop. Hence, the choice of secondary markets as a strategic scope should be interpreted as a focused cost leadership strategy followed by IBM in Japan or Apple in the US (Apple, 2005). Nevertheless, despite the proﬁtability of product reuse, the market size of secondary segments will remain limited in the future. Packard (1960) and Bellman (1990) use the concept of obsolescence to explain why used products cannot satisfy all market segments. In their opinion, the residual value loss pattern is determined by three types of obsolescence: 7

A comprehensive introduction to the theory of property rights is provided by Furubotn and Pejovich (1972) as well as Ebers and Gotsch (1999).

28

2 Strategic Aspects of Asset Recovery

• Psychological obsolescence. A product might become worthless because of fashion eﬀects, although it still works perfectly. As an example, mobile phones are exchanged every two years by the German airtime providers although the majority of the two-years cellular phones are fulﬁlling their core utility: Making and receiving phone calls. Nevertheless, since these phones are considered as a fashion widget, providing a brand-new mobile phone is a prerequisite for the acquisition or prolongation of airtime contracts. • Qualitative obsolescence is linked to the situation where a product wears out and stops servicing. Owners’ needs can then be satisﬁed only through a replacement buy. Investigations of Lund (1977) related to qualitative obsolescence patterns over time point out that the failure distribution is similar to a bath-tube curve, i.e. the qualitative obsolescence takes place at the beginning and at the end of the service life, either because of manufacturing failures or of wearing exposure. • Functional obsolescence. Consequently to a modiﬁcation of customers’ needs, a current product may not provide enough utility anymore, either because of lacking functionality or inferior performance. Thus, customers will have to buy new goods in order to close the gap between the utility obtained from their current equipment and the utility provided by up-to-date product releases. Diﬀerentiation: Green Image The green product image is often cited as a potential selling argument in the case of a diﬀerentiation strategy (Carter and Ellram 1998). While the will to buy environmentally friendly products is widely documented, the market for these products often remains a niche. Oetzel (1997) points out the discrepancy between beliefs and attitudes of customers with respect to environmental attributes. He notices that customers are aware of the environmental problems but will not necessarily act in consequence. Oetzel’s argumentation is based on the conﬂicts between environmental issues and other competing values and beliefs (e.g. convenience, modernity, leisure). When the psychological obsolescence of remanufactured item is too signiﬁcant for potential customers, buyers might omit the correlation between their attitude (get a new product instead of a refurbished one) and the impact on the environment (increased resourced consumption) to avoid mental conﬂicts.

OEM Strategy Product level Activity Source Oce cost leadership core reuse, refurbishing Krikke et al. (1999) Xerox cost leadership core reuse, refurbishing Kerr and Ryan (2001) Michelin cost leadership core reuse, retreading Michelin (2004) Bosch cost leadership core refurbishing Klausner et al. (2000) HP diﬀerentiation (quality) actual recycling Davey et al. (2005) BrB diﬀerentiation (quality) actual recycling chapter 4 Omron diﬀerentiation (service) augmented cannibalization Kuik et al. (2005) IBM diﬀerentiation (service) augmented cannibalization Fleischmann et al. (2003) Agfa-Gevaert diﬀerentiation (service) augmented refurbishing, cannib. Spengler and Schr¨ oter (2005) Mercedes diﬀerentiation (service) augmented refurbishing Driesch et al. (2005) FSC focus (service) augmented refurbishing, cannib. Podratzki (2003)

Table 2.3. Closed-loop supply chains and competitive strategy: Case overview

2.2 Closed-Loop Supply Chains and Competitive Strategies 29

30

2 Strategic Aspects of Asset Recovery

The low importance of a green image during the procurement decision is documented by a survey of Saied and Velasquez (2003) investigating the decision criteria for the procurement of desktop computer in the Japanese home-oﬃce market. While 90% of the interviewees cite price and performance, 29% evoke design and 16% evoke brand image, only 5.8% cite environmental issues as a buying motive.8 Similar results are provided by Schulz (2005) for passenger cars. Although environmental issues might concern customers, customers will procure a product for the direct utility it provides. As long as the environmental impacts of a procurement decision are not clearly quantiﬁable during the buying decisions, green products will remain a niche. To summarize, manufacturers face a wide range of opportunities to turn product returns into a sustainable competitive advantage. While OCE, Xerox and Michelin conﬁrm the sustainability of cost-leadership strategies, other manufacturers of complex products prove that diﬀerentiation may, to some limited extent, support a company’s strategic positioning (HP, BrB). However, the contribution of closed-loop supply chains to the realization of a competitive edge depends on the ability to take the right decisions concerning the value chain implementation of the strategy. A clear strategic focus should prevent the original equipment manufacturer from investing in useless activities with respect to the strategic purpose of the closed-loop.

8

The survey allowed multiple answers.

3 Strategic Impact of Closed-Loop Supply Chains

The presumed contribution of asset recovery programs to the achievement of a competitive edge is a prerequisite for investigating the utility of OEM-driven closed-loop supply chains. Nevertheless, as Original Equipment Manufacturers commit management and ﬁnancial resources to the achievement of closed-loop supply chains, the beneﬁts of asset recovery have to be also economically proven. The present chapter will therefore deal with the quantitative approaches enabling a strategic closed-loop accounting. We will ﬁrst provide an overview on the quantitative models already published in this ﬁeld and ﬁlter out the characteristics necessary for the development of a general purpose decision-making model. This model will be then presented in section 3.2 and managerial insights concerning the incentive role of green fees, the trade-oﬀ between cannibalization and revenue per new item sold (RNIS) as well as intra-organizational incentives will be drawn in section 3.3.

3.1 Literature Overview As already stated in the previous section, an OEM-driven asset recovery policy is mainly economically motivated. In this context, the use of quantitative models is a prerequisite for assessing the proﬁtability of such sustainable programs. However, despite the wide range of quantitative publications in the reverse logistics ﬁeld, only few are dedicated speciﬁcally to ﬁnancial topics. St¨ olzle (1996) as well as Dutz and Femerling (1994) provide both an overview of the processes in the reverse pipeline, from acquisition to reintegration, as well as an activitybased costing framework. In the same fashion, Tibben-Lembke (1998)

32

3 Strategic Impact of Closed-Loop Supply Chains

presents a total costs of ownership (TCO) approach helping to identify the role of reverse logistics costs in current accounting systems. Guide and van Wassenhove (2003) follow the same principles in showing the impact of reverse logistics on the economical value-added (EVA) of a ﬁrm. Nevertheless, what these publications have in common is that they are only dealing with reverse logistics accounting rather than with closed-loop accounting. The frameworks presented are thus only adapted for a marginal costs analysis since they model neither eﬀects such as cannibalization nor the necessity of disposing an unused core surplus when this cannot be reintegrated into the production process for marketing reasons. Ferrer (1997) proposes a mathematical approach for comparing the expected revenues and costs for retreading tires that includes a rejection rate and thus takes partially into account a disposal cost for core surplus. Although the ad-hoc structure of his model makes it diﬃcult to generalize his approach, Ferrer provides a true closed-loop model and tries thereby to ﬁnd out an optimal number of cycles for a tire. The approach of Siestrup and Haasis (1997) is similar to Ferrer’s with respect to the modeling of cycles but is more adapted for discounted cash-ﬂow computations, due to the integration of time in their model. However, a shortcoming of both models is the absence of market segmentation. In fact, the authors assume that reused or remanufactured products are not exposed to any kind of obsolescence and compete with brand new products. This problem has been identiﬁed by Geyer and van Wassenhove (2002) as well as Lebreton and Tuma (2003) who limit the remanufacturing pipeline through a bottleneck on the demand’s side. To summarize, a comprehensive closed-loop should be developed with respect to the following aspects: • The mathematical framework should encompass costs and revenues simultaneously to enable realistic return on investment computations. In fact, the setup of a durable asset recovery program is coupled with an initial ﬁnancial burden (design for environment, setup of reverse logistics network including recovery plant) that has also to be recovered. Concerning the revenues, one requires a diﬀerentiation between new and recovered ﬂows due to their value discrepancy in the customers’ eyes. As recovered goods are exposed to obsolescence, these are considered as inferior and need signiﬁcant rebate to be sold. In this context, reverse logistics models with a lot-sizing determination approach are not apt for the assessment of closed-loops as these only take costs into account and thus assume the same

3.2 A Generic Closed-Loop Strategic Model

33

market value for new and old (e.g. Richter, 1996 and Inderfurth, 2002). • The interdependencies between the three core recovery paths have to be accurately depicted because an increase of the reuse ﬂow reduces in the short-term the quantity of products available for remanufacturing and recycling. Furthermore, reintegrated products may compete directly against new products in the lower segments. However, the cannibalization danger might be balanced by the fact that cheaper oﬀers may generate additional demand. • The asset recovery probability depends on sequentially linked key factors such as return probability, technological reintegration alternatives, functional state of a product and the demand segmentation as developed in Geyer and van Wassenhove (2002) and Lebreton and Tuma (2002, 2003). • The primary focus of product stewardship laws is, as already stated in the previous section, to close the supply chain at a material level when it is not proﬁtable to do so. An assessment of the cost impact of material recycling on a ﬁrm’s proﬁt is, in this context, useful in order to ﬁnd out the best response to legislative constraints.

3.2 A Generic Closed-Loop Strategic Model The following framework is embedded into a closed-loop model fulﬁlling the requirements listed above. It is important to notice that we assume the existence of two market segments in which the competition nature diﬀers. The competition between new and recovered items only takes place in the low-end segment. Figure 3.1 illustrates the ﬂows and connections of the strategic model. The mathematical model contains several assumptions to limit the scope of our studies and guarantee the solvability of the model. First, we consider a monopolistic manufacturer, this assumption can be made since management accounting schemes address internal decisionmaking processes only (Johnson and Kaplan, 1988). Furthermore, we build a steady-state model that does not take time into account. We consider products that can be reused or remanufactured only once, such as copiers (Krikke et al., 2005), computers (White et al., 2003), mobile phones (Guide et al., 2005) or diesel motors (Driesch et al., 2005). The OEM is able to sell all recovered products, either through reuse or remanufacturing, with a rebate of λ% on the price of new items.

34

3 Strategic Impact of Closed-Loop Supply Chains

Fig. 3.1. Closed-loop accounting model

The costs of the production processes (materials, components, assembly) represent the value-added of each process. This aggregation level has also been chosen by Ayres et al. (1997) and Ferrer (1997) to build their models. From a strategic point of view, the concept of value-added is helpful to communicate the economical and ecological savings of asset recovery as it documents the potential savings without requiring precise information on speciﬁc materials or components. The danger of cannibalization, as we will investigate in section 3.3, is one of the most relevant hurdle for the implementation of closedloop supply chains. To catch the competition eﬀects between new and recovered products, we include two market segments. The premium segment accepts only new items while the low-budget segment prefers cheaper recovered items and spend the rest of their budget for procuring new items. Atasu et al. (2005) adopt the same modeling approach to analyze the cannibalization eﬀects between new and remanufactured products and depict the customers willing to buy refurbished items green customers while these are only price-sensitive. It is important to notice that we will not consider any green market segment per se because of the very small size of this potential segment (see section 2.2). The mathematical model is directed with help of ﬁve sequential key parameters which are all positive.

3.2 A Generic Closed-Loop Strategic Model

35

Technical parameters ρ : probability of return ω : probability of reuse θ : probability of remanufacturing β : reused fraction in remanufactured product PR µ : recyclable fraction of returned materials

The ﬁnancial parameters include all retail prices and direct costs, hence the objective function πCLSC represents the overall contribution margin. The budget of the segments φH and φL are expressed in currency units. Financial parameters σN : retail price of new products λ : price rebate for recovered products (reuse, reman.), in % of σN κP N : production costs for new items κP R : refurbishing costs for product returns C κN : production/procurement costs for new parts κM N : procurement costs for virgin materials κM R : procurement costs for recycled materials κρ : reverse logistics costs κO : disposal costs (landﬁll/incineration) φH : budget of high-end segment φL : budget of low-end segment

36

3 Strategic Impact of Closed-Loop Supply Chains

Variables PNH : sales of new products in the high-end segment PNL : sales of new products in the low-end segment R : quantity of reused products PR : quantity of refurbished products H CN : quantity of procured/manufactured new components for the high-end segment L CN : quantity of procured/manufactured new components for the low-end segment MN : procurement quantity of raw materials MR : procurement quantity of recycled materials RFN : quantity of returned products in the ﬁrst cycle O : quantity of material land ﬁlled or incinerated

The proﬁt of the investigated OEM πCLSC is computed with help of equation 3.1. Revenues for new and recovered products are reduced by procurement/production costs for materials, parts and ﬁnal products P ρ as well as recovery pipeline costs (κM R , κR and κ ) and landﬁll costs O κ . max ! πCLSC

πCLSC = σN · PNH + PNL + σN · (1 − λ) · R + PR − κM N · MN

C H L P H L P −κM R · MR − κN · CN + CN − κN · PN + PN − κR · PR

−κρ · RFN + ρ · (R + PR ) − κO · O

(3.1)

After a model transformation, both σN and σN · (1 − λ) will become variables in a proﬁt maximization problem. Thus, the model has to take into account the budget restriction of the customers in order to deliver realistic results. The negative correlation between retail price and demand in both high and low-end segments is modelled with help of equations 3.2 and 3.3 respectively. φH and φL represent thereby the customers’ budget in a given segment. While the high-end segment is only supplied with new products PNH , reused R and refurbished products PR cannibalize the demand for new products in the low-end segment PNL . In this segment, we assume that the price diﬀerence λ · σN compensates the product obsolescence and therefore convince customers to procure recovered assets when these are

3.2 A Generic Closed-Loop Strategic Model

37

available. Unlike Atasu et al. (2005) and Toktay and Ferguson (2005), we model the markets with help of a budget which is fully spent during the optimization. This assumption simpliﬁes the determination of a feasible market equilibrium while taking the price sensitivity of the customers into account. φH = σN · PNH

φL = σN · PNL + σN · (1 − λ) · PR + R

(3.2)

(3.3)

The ﬂow RFN , which represents the quantity of reintegrable products, either on the product or the component level, is determined through the return probability ρ. As depicted in eq. 3.4, RFN encompasses reintegrable products after their ﬁrst utilization phase issued from the sales of new products PNH and PNL . The reintegrable product returns are either guided into the reuse path (eq. 3.5) or into the refurbishing path (eq. 3.6) depending on the values of the reuse probability ω and the remanufacturing probability θ respectively.

RFN = PNH + PNL · ρ

(3.4)

R = PNH + PNL · ρ · ω

(3.5)

PR = PNH + PNL · ρ · θ

(3.6)

The residual return ﬂow, gathering return ﬂows from the second cycle ρ · (R + PR ) and non-reintegrable products from the ﬁrst cycle (1 − ω − θ) · RFN , is available for material recycling. However, only a fraction µ of this ﬂow can be reused in the production process. The resulting quantity MR is computed in equation 3.7 while the residual materials O are landﬁlled or incinerated (eq. 3.8). MR = ρ · µ ·

PNH + PNL · (1 − ω − β · θ) + R + PR

O = ρ · (1 − µ) ·

PNH + PNL · (1 − ω − β · θ) + R + PR

(3.7)

(3.8)

The composition of the products is calculated sequentially in the equations 3.9 to 3.11. All new parts contain either virgin MN or recycled MR materials while we assume that no quality diﬀerences exist between both alternatives (eq. 3.9). First cycle products PNH and PNL are assembled with exclusively new components (eq. 3.10). Refurbished

38

3 Strategic Impact of Closed-Loop Supply Chains

products PR require new low-end components but their amount is limited to the fraction (1-β) which represents the fraction of non-reusable parts (eq. 3.11). H L + CN = MN + MR CN H CN = PNH + PNL L = PR · (1 − β) CN

(3.9) (3.10) (3.11)

All ﬂow variables are depending on the initial quantity of the new products sold (PNH +PNL ), on their retail price as well as on the technical parameters. This deterministic structure of the problem enables us to simplify the model to the proﬁt function πCLSC presented in equation LC 3.12. QN (eq. 3.13) represents the quantity of new items sold while πN depicts the life-cycle value of a new item (eq. 3.14). LC · Q πCLSC = f (σN , λ, ρ, ω, θ, β, µ) = πN N

QN = PNH + PNL =

φH φL − (1 − λ) · φH · ρ · (ω + θ) + σN σN · 1 + (1 − λ) · ρ · (ω + θ)

(3.12)

(3.13)

LC = σ · 1 + (1 − λ) · ρ · (ω + θ) − κM · 1 + ρ · θ · (1 − β) − ρ · πN N N

µ · (1 − ω − θ · β + ρ · (ω + θ)) − κM R ·ρ·µ· 1−ω−θ·β

P P +ρ · (ω + θ) − κC N · 1 + ρ · θ · (1 − β) − κN − κR · ρ · θ

−κρ · ρ · 1 + ρ · (ω + θ) − κO · ρ · (1 − µ) · 1 − ω − θ · β

+ρ · (ω + θ)

(3.14)

LC , we notice the assignment of Concerning the life-cycle value πN process costs for all potential recovery strategies to the initial product sales. These are balanced with the additional revenues obtained from reuse and remanufacturing. On the basis of this uniﬁed approach, we are able to compare the impact of the three core asset recovery strategies on a ﬁrm’s proﬁt with the help of a modiﬁcation of the key factors. Table 3.1 speciﬁes the key factors settings for the investigated scenarios. Scenario ini depicts the situation of most of the OEMs prior to the nineties when no regulation was pending. Subsequent to the introduction of product stewardship laws, OEMs face an internalization of the

3.3 Closed-Loop Supply Chains: Managerial Insights

39

Table 3.1. Technical parameter settings of the investigated scenarios

Scenario ρ ini reg rec reu rem all

Key factors ω θ β µ

0 0 >0 0 >0 0 >0>0 >0 0 >0>0

0 0 0 0 0 0 0 0 >0 0 0 0 >0>0 0 >0>0>0

asset recovery costs. These are then compelled to ﬁnance the collection and disposal of their products (sc. reg). One of the main goals of recovery costs internalization is to encourage the use of recyclable materials (Corbett and van Wassenhove, 1993). Scenario rec assesses therefore the real inﬂuence of product stewardship laws on recycling incentives. By closing the loop on a product level (sc. reu) or on a component level (sc. rem), OEMs might yield better ﬁnancial results. These results can be reached either through additional sales in lower segments or through a reduction of the disposal costs. Environmentally proactive ﬁrms such as Xerox (Kerr and Ryan, 2001), or IBM (Fleischmann et al., 2003) seem to follow scenario all as these companies close their loop at the product, component and material levels simultaneously.

3.3 Closed-Loop Supply Chains: Managerial Insights 3.3.1 The Impact of Green Fees on Asset Recovery Original equipment manufacturers would use recycled materials without being constrained to when these would be cheaper than virgin M materials (κM R < κN ). While this might be true for ferrous metals, a broad scope of materials, especially plastics, cannot be sold at competitive prices by now. In this case, product stewardship laws try to balance this cost disadvantage with the introduction of a green fee paid by the OEM (Corbett and van Wassenhove, 1993). To analyze the incentive role of a green fee, we will compare in the following the proﬁts reg ini of a manufacturer before (πCLSC ) and after (πCLSC ) the introduction of a fee. Finally, we will identify the patterns necessary for a proﬁtable rec ). internalization of recycling (πCLSC

40

3 Strategic Impact of Closed-Loop Supply Chains

Equation 3.15 represents the manufacturer’s initial income. It is obvious that no asset recovery costs are appended to the production and procurement costs. The production quantity of new items amounts to the budget of both segments divided by the retail price for new items.

φH + φL

ini C P πCLSC = σ N − κM N − κN − κ N ·

(3.15) σN The manufacturer’s proﬁt after the introduction of a green fee is extended as follows: reg πCLSC

= σN −

κM N

−

κC N

−

κPN

−ρ· κ +κ ρ

O

φH + φL

·

σN

(3.16)

reg ini − πCLSC , is depicted in equation 3.17. The impact on proﬁts, πCLSC

φH + φL

ρ O · ∆reg ini = −ρ · κ + κ

σN

(3.17)

We notice that a green fee has a negative impact on the OEM’s margin when the reverse supply chain costs are positive κρ + κO > 0. Nevertheless, the manufacturer has two possibilities to reduce this burden: Improve the product’s residual value (decrease κO ) or support system losses (decrease ρ). Since the collection cost κρ are, in general, very low compared to the initial retail price σN , the manufacturer may enable external recyclers to beneﬁt from used products as soon as −κO ≥ κρ , i.e. when the disposal revenues overcompensate the reverse logistics costs. By increasing the recovery value of their end-of-life products, producers of passenger cars and computers manage to avoid the payment of a recycling fee. The revenues from the gold and ferrous metals embedded in catalytic converters and in motherboards (Behrend et al., 1997) overcompensate the logistics and disassembly costs. A passive behavior concerning the product returns is another means to keep product stewardship’s burden low. An OEM may also inﬂuence customers’ behavior if he does not promote the possibility to return their end-of-life product. In fact, customers would either stockpile the computers or resell them to brokers trading with less-developed countries. The probability that a used core leaves the system ρ → 0, however, is strongly linked with the residual value κO of these items. For instance, estimations show that 23 of the BMW cars sold in Germany are exported to a non-UE country after a service life of ten years (Kircher, 2004). Hence, BMW has only to ﬁnance the recovery of ρ = 13 of his end-of-life cars.

3.3 Closed-Loop Supply Chains: Managerial Insights

41

In the reg scenario inspected above, the green fee would not inﬂuH L ence the turnover of the company ( φ σ+φ ) if the producer does not N pass the green fee to the buyers. However, as reverse logistics and recovery costs do not present a critical issue to OEMs due to their low fraction in the initial total production costs: 1% for desktop computers (Kriebel, 2004), 0.5 to 1% for passenger cars (Kircher, 2004), OEMs could keep the margin of the ini case. Nevertheless, this measure has reg:pass a negative impact on the proﬁts πCLSC of the manufacturer (see eq. 3.18), especially in the presence of high return rates and signiﬁcant green fees. Passing the green fee to the buyer impacts negatively the proﬁt function (if κρ + κO > 0) but does not threaten the strategic positioning as long as the green fee remains insigniﬁcant. reg:pass ∂πCLSC

∂ρ

κρ + κO · φH + φL

C P = σ N − κM N − κN − κ N · −

σ N − ρ · κρ + κ O

2 (3.18)

The most eﬃcient way for the lawmaker to operationalize sustainable development consists, in the short-term, in avoiding market-related incentives and in forcing the use of recycled materials into the producrec tion process. πCLSC (eq. 3.19) depicts this situation and provides a real internalization of the recovery costs on the OEM’s side. rec πCLSC

M C P = σ N − κM N · (1 − ρ · µ) − κR · ρ · µ − κN − κN

−ρ · κρ + (1 − µ) · κO

φH + φL

·

σN

(3.19)

Every increase of a product’s recyclable fraction pays when the recovered materials price is cheaper than the price for virgin ones plus the disposal costs (eq. 3.21). Hence, we notice that the disposal costs are internalized into the procurement costs for new materials. φH + φL ∂π rec M O · = ρ · κM N − κR + κ ∂µ σN

∂π rec M O ≥ 0 ⇔ κM R ≤ κN + κ ∂µ

(3.20) (3.21)

We now investigate the proﬁt impact of additional returns by differentiating π rec by the return factor ρ. According to 3.23, an additional return is proﬁtable when the ability to take advantage of the

42

3 Strategic Impact of Closed-Loop Supply Chains

price diﬀerential between recyclable and virgin materials is higher than the collection costs. Moreover, it turns out that products with a very low recyclable fraction (µ → 0) are seldom proﬁtable to recover when ρ O the reverse logistics and disposal costs are positive ( κ +κ → ∞ µ M M O κN − κR + κ ). ∂π rec = ∂ρ

µ·

κM N

−

κM R

+κ

O

−κ −κ ρ

O

∂π rec M O ≥ 0 ⇔ µ · κM − κ + κ ≥ κ ρ + κO N R ∂ρ

·

φH + φL (3.22) σN (3.23)

In addition to the internalization of disposal costs, a green fee has a noticeable impact on the decision to start reintegrating materials. In the following, we calculate the proﬁt impact ∆rec ini of the switch from a non-regulatory situation to recycling and the impact from an already existing product stewardship to recycling ∆rec reg . In equation 3.24, the proﬁt diﬀerence is equivalent to the ﬁnancial burden related to the green fee (see also eq. 3.17). The proﬁtability yield is therefore higher when an OEM starts a recycling program after the introduction of product stewardship laws rather than before.1

∆rec ini

=

rec πCLSC

−

ini πCLSC

M = ρ · µ · κM N − κR −

ρ · κ + (1 − µ) · κ ρ

∆rec reg

=

rec πCLSC

−

reg πCLSC

= ρ·µ·

κM N

−

κM R

O

φH + φL

·

σN

+ρ·µ·κ

O

φH + φL

rec ρ O ∆rec · reg − ∆ini = ρ · κ + κ

σN

·

φH + φL σN (3.24)

Given the previous results, it is possible to conclude that a green fee improves the attractiveness of recyclable materials by increasing the procurement price of new materials. This internalization of the disposal costs, however, does not automatically provide enough incentives rec to reintegrate materials into the production process. When ∂π∂ρ ≤ 0, the existence of compelling recycling targets for OEMs turns out to be necessary to ensure a sustainable behavior.

1

If the sum of collection and disposal costs is positive, κρ + κO > 0.

3.3 Closed-Loop Supply Chains: Managerial Insights

43

3.3.2 Managing the Cannibalization Eﬀect Unlike recycling, reuse and remanufacturing have a direct, negative impact on the turnover of new products. Nevertheless, examples show that product-based recovery strategy might increase an OEM’s overall proﬁt (see for instance Debo et al., 2005 or Kerr and Ryan, 2001). Since the products are sold twice, these generate a higher revenues per new item sold (RNIS). However, OEMs have to oﬀer a rebate λ in the second life-cycle and carry the costs of the reverse pipeline so that the total life-cycle value of the product might then decrease. Thus, the focus of our investigations will be to identify the key factors determining the proﬁtability of reuse and remanufacturing. First, we demonstrate the negative impact of product-based recovery strategies on the sales of new products. We recall here to eq. 3.13 for the deﬁnition of QN . Equation 3.25 shows the negative impact on the original turnover evoked above as −(1 − λ) · ρ ≤ 0. ∂QN φH + φL ∂QN (1 − λ) · ρ ≤ 0 (3.25) = = − 2 · ∂ω ∂θ σN (λ − 1) · ρ · (ω + θ) − 1 We now seek to ﬁnd out when reuse is proﬁtable despite evident cannibalization eﬀects. For this purpose, we use the parameter settings reu . of scenario reu (see table 3.1) and obtain the proﬁt function πCLSC reu πCLSC

C P = σN · 1 + (1 − λ) · ρ · ω − κM N − κN − κN

−ρ · κ · (1 + ρ · ω) + κ · (1 − ω + ρ · ω) ρ

O

· QN (3.26)

The marginal value of a reusable product (eq. 3.27) is positive as long as Ω is negative. On the basis of the case studies available, it M is possible to state that the production costs κPN + κC N + κN are for passenger cars and desktop computers at least ﬁfty times higher than the collection and disposal costs. Assuming a return rate of 100% and a rebate of 50%, we can conclude that reuse is in this case a proﬁtable alternative (Ω = −0.5 · 100 − 0.5 · 1 + 0.5 · 1 = −50 < 0). reu φH + φL ρ ∂πCLSC ·Ω = − 2 · ∂ω σN ρ · ω · (λ − 1) − 1

(3.27)

44

3 Strategic Impact of Closed-Loop Supply Chains

M O ρ Ω = (λ − 1) · κPN + κC N + κN + (ρ · λ − 1) · κ + ρ · λ · κ

(3.28) Hence, equation 3.27 is positive when Ω < 0 as ρ ∈ [0, 1]. Reuse is thus proﬁtable when the rule expressed in equation 3.29 holds, i.e. when the adjusted savings on production costs do not surpass the eﬀective reverse logistics costs. In this context, we can state that products with high initial production costs and low reverse logistics costs are generally proﬁtable to reintegrate. Nevertheless, the reintegration of cheap consumer products with comparatively high return costs will not obligatory create value since the relative attractiveness of reuse decreases as reverse logistics costs increase.

C P ρ O (1 − λ) · κM − κO N + κN + κN ≥ ρ · λ · κ + κ

(3.29)

Logically, the proﬁt impact of a rebate increase is negative (eq. 3.30) as Λ is always positive (eq. 3.31). reu φH + φL ρ·ω ∂πCLSC ·Λ<0 = − 2 · ∂λ σN ρ · ω · (λ − 1) − 1 M O Λ = κPN + κC N + κN + ρ · κ · (1 − ω + ρ · ω) +ρ · κρ · (1 + ρ · ω) > 0

(3.30)

(3.31)

These results provide insight into an current research question concerning the valuation of reverse ﬂows. In fact, manufacturers recovering their products need to valuate them in their balance sheet as these are considered as assets. Teunter (2001) as well as Teunter and van der Laan (2004) recently investigated valuation methods for inventories of recoverable products with help of stochastic inventory models. The authors conclude that the value of a recoverable item is equal to the margin provided by this item when resold. With help of the present model, we are able to state that the book value inv reu of potentially reusable items is equal to the marginal proﬁt generated by these products. Thus, π reu . inv reu = CLSC ω In the present model, remanufacturing diﬀers from reuse as we consider that refurbished items are repaired and their defective parts removed. The fraction of the reusable parts has then to be taken into account. The resulting proﬁt remanufacturing function is depicted in eq. 3.32.

3.3 Closed-Loop Supply Chains: Managerial Insights

45

rem C πCLSC = σN · 1 + (1 − λ) · ρ · θ − κM N + κN · 1 + ρ · θ · (1 − β)

−κPN − κPR · ρ · θ − ρ · κρ · (1 + ρ · θ) + κO ·

(1 − θ · β + ρ · θ)

· QN

(3.32)

If all parts are reused (β = 1), then we obtain a similar proﬁt funcreu . We notice the emergence of additional refurbishing costs tion as πCLSC P savings on components and materials κR as well as fraction-dependent M C β · (ρ · θ) · κN + κN . Moreover, the OEM saves ρ · κO · θ · β on the disposal costs. Since the potential additional sales are diﬃcult to quantify for manufacturers, these often motivate their remanufacturing program only with the savings realized (see e.g. Maslennikova and Foley (2000) for Rank-Xerox or Davey et al. (2005) for Hewlett Packard). The durability of components is critical for the success of remanufacturing (Stahel, 1986) since manufacturers have to carry disassembly costs to replace the defective or obsolete parts. This durability can be correlated with the fraction of components that can be reused β after having been returned (Geyer and van Wassenhove, 2002). The marginal value of an increase of the reusable fraction is expressed in eq. 3.33. We notice that the left term of eq. 3.33 is always positive when C O κM N + κN + κ > 0. As a result, ceteris paribus, it is always proﬁtable to increase the remanufacturable fraction of a product.

rem ∂πCLSC

∂β

=

C O ρ · θ · κM N + κN + κ

1 + (1 − λ) · ρ · θ

·

φH + φL σN

(3.33)

Let us now assume that longer lasting components are more expensive to produce, we state therefore that the manufacturing costs of parts are positively correlated with the reintegration level β. We have C C κC N = ι · (1 + β) with ι > 0 representing the initial cost for a component inapt for reuse. This positive correlation has been observed for instance in the tire industry by Lebreton et al. (2004) and by Debo et al. (2005). Equation 3.33 is then transformed to eq. 3.34 to obtain the resulting proﬁtability yield (eq. 3.35).

rem∗ ∂πCLSC

∂β

=

O − ιC · (1 − 2 · ρ · θ · β) ρ · θ · κM N +κ

1 + (1 − λ) · ρ · θ

·

(3.34)

46

3 Strategic Impact of Closed-Loop Supply Chains

φH + φL σN rem∗ ∂πCLSC >0 ∂β

⇔

β>

O ιC − ρ · θ · κM N +κ

2 · ρ · θ · ιC

(3.35)

O In order to draw conclusions from these results, we express κM N +κ C 2 M O as a percentage of the one-way component costs ι . Hence, κN +κ = τ · ιC with τ ∈ [0, ∞[. Equation 3.35 can be rewritten as follows: rem∗ ∂πCLSC >0 ∂β

⇔

β>

(1 − ρ · θ · τ ) 2·ρ·θ

(3.36)

We conclude that when the fraction of remanufacturable items ρ · θ is insigniﬁcant (→ 0), an OEM should not improve the fraction of reusable items in his products because he is not able to reclaim the beneﬁts of this strategy. Furthermore, when the value-added of a component is low compared to the material and disposal costs (τ 0), the proﬁtability yield for a durability improvement is lower. In other words, the attractiveness of durable items rises with the savings realized through component reuse. While the previous reuse scenario did not enclose repair costs, remanufactured products have been partly dismantled and their defective parts have been replaced. The disassembly process induces important costs, especially when the product has not been designed for the purpose of providing an easy access to the replaceable parts (VDI, 2002). It is generally recognized that a modular design is helpful for this purpose due to the complexity reduction of the dismantling process (Krikke et al. 2004). In a similar fashion as in eq. 3.27, we aim at ﬁnding out the inventory value of a return ﬂow dedicated to remanufacturing rem ∂πCLSC = inv rem . This value is positive when Θ is negative. Once again, ∂θ it is obvious that a rebate λ and the refurbishing costs κPR (including disassembly) have a negative impact on the book value inv rem while an increase of component reusability (β) is, ceteris paribus, beneﬁcial for inv rem . rem φH + φL ∂πCLSC −ρ ·Θ = 2 · ∂θ σN ρ · θ · (λ − 1) − 1 2

(3.37)

Recall that the component costs do not include the material costs necessary to produce a part but solely the value-added of the component manufacturing process.

3.3 Closed-Loop Supply Chains: Managerial Insights

47

C P P Θ = (β − λ) · κM N + κN − (1 − λ) · κN − κR −

κO · (ρ · λ − β) − κρ · ρ · λ

(3.38)

Equation 3.38 enables us to estimate a priori the proﬁtability of C remanufacturing when κM N + κN is predominant. In fact, as soon as the rebate λ is higher than the value of the replaced parts (in % of the initial material and component costs), remanufactured products have ∂π rem < 0. a negative impact on the overall proﬁt since Θ > 0 ⇒ CLSC ∂θ In order to illustrate these ﬁndings, we now apply Θ to the computer remanufacturing case study of chapter 7. The required information is gathered in table 3.2. We take into consideration an OEM leasing highend computers (power) with a lease term of t years and a yearly value loss of 50%. The replaced parts account for 10% of a power computer’s initial value. We obtain Θt = 110 − 1100 · 0.5t . Table 3.2. Leasing and refurbishing of power computers: Aggregated data P M ρ 1 κC N 1,100 EUR κN 20 EUR κN 0 EUR t ρ β 0.9 λ 1 − 0.5 κ 10 EUR κO 10 EUR

The leasing and refurbishing of power computers is proﬁtable when the lease term is shorter or equal to three years (Θ ≥ 0). The attractiveness of power lease returns will be conﬁrmed with help of a detailled strategic planning model in chapter 7. Table 3.3. Proﬁtability of computer lease remanufacturing by age: First estimations 1 year 2 years 3 years 4 years Θ1 = 439 Θ2 = 164 Θ3 = 26.5 Θ4 = −42.5

Despite cannibalization eﬀects and lower retail prices for recovered products, product-based asset recovery might be proﬁtable under the speciﬁc cost settings identiﬁed above. Generally speaking, products with a high embedded value and not subject to psychological obsolescence are particularly proﬁtable to recover. As a rule of thumb, we can state that remanufacturing is worth investigation when the rebate λ for a refurbished product is smaller than the fraction of the components being

48

3 Strategic Impact of Closed-Loop Supply Chains

replaced (β). With the help of the closed-loop model, we have also been able to provide insight into the valuation of returned assets for accounting purposes. Returned assets should be therefore valued at their marginal contribution to a company’s proﬁts. 3.3.3 The Role of Intra-Organizational Incentives Alignment OEMs implementing closed-loop supply chains tend to create an independent division dedicated only to asset recovery. This measure is twofold motivated: First, the focus of the supply chain workforce is set on forward ﬂows which represent the ﬂows a company bases its business on. Reverse ﬂows are often considered as an unnoticeable burden and remain too long in the reverse pipeline, therefore suppressing any possibility to turn these assets into cash. The setup of an independent division with the objective of maximizing the proﬁts from the product returns provides enough incentives to reduce the value losses in the reverse chain. Rogers and Tibben-Lembke (1999) point out the existence of logistics providers specialized on this type of activities and acting as a proﬁt-center. Daimler-Benz (Driesch et al., 2005), HP (Davey et al., 2005), ReTreadCo (Debo and van Wassenhove, 2005), Fujitsu-SiemensComputers (Podratzki, 2003) or Compaq (Sarkis, 2003) also document the existence of separated proﬁt-centers. We will show in the following that this strategy may pay when asset recovery remains a marginal activity which does not present serious cannibalization danger for the turnover of new products. Nevertheless, intra-organizational conﬂicts arise when the asset recovery division gains in importance. These conﬂicts are similar to the competition between OEMs and the independent remanufacturers that will be treated in the next chapter with exception of one aspect: While a independent remanufacturer reclaims a fraction of the proﬁts of an OEM, a manufacturer-owned refurbishing center just transfers the profits within the ﬁrm. In a pioneering work, Toktay and Wei (2005) identify the risk of suboptimal decisions within the company due to a failing cost allocation scheme. The authors propose to pass a fraction of the production costs to the remanufacturing division but agree on the fact that this solution is diﬃcult to apply. We propose the setup of an internal transfer price σT in order to avoid the allocation of the production costs. This transfer price represents a general compensation which has the advantage of being easier to implement. With help of the present model, we are able to calculate price limits for σT ensuring the satisfaction of both divisions.

3.3 Closed-Loop Supply Chains: Managerial Insights

49

We consider two divisions, one focused on the sales of new products I and the other, AR, dedicated to asset recovery. We compare the proﬁts of both divisions before (ini) and after the introduction of product reuse (reu). We consider that conﬂicts will be avoided when both divisions generate higher proﬁts after the introduction of prodreu − π ini ≥ 0. Assuming that uct reuse, i.e. πIreu − πIini ≥ 0 and πAR AR division I only carries the costs directly assigned to the production of new items, reuse is the proﬁt functionof the division under Hproduct φ +φL reu M C P πI = σN − κN − κN − κN · QN . Since QN ≤ σN , division I

φ +φ C P would lose σN −κM −QN after the introduction N −κN −κN · σN 3 of product reuse. Hence, the introduction of a compensation price is necessary to compensate division I for this proﬁt loss. We include this compensation as an additional revenue in πIreu . The proﬁt modiﬁcation πIreu − πIini is given in eq. 3.39. H

L

πIreu − πIini = −

C P (1 − λ) · ρ · ω · σN − κM N − κN − κ N − σ T

1 + (1 − λ) · ρ · ω

·

φH + φL σN (3.39)

C P πIreu − πIini ≥ 0 ⇔ (1 − λ) · ρ · ω · σN − κM N − κN − κN ≤ σ T

(3.40)

Division I has an incentive to support reuse when the transfer price M C P is at least higher than (1 − λ) · ρ · ω · σN − κN − κN − κN which also represents the proﬁt loss for I through reuse (eq. 3.40). Reuse reu are positive. Ψ is proﬁtable for the AR division when its proﬁts πAR represents the contribution margin of product reuse from which the transfer price σT is deducted. reu πAR =

Ψ φH + φL · 1 + (1 − λ) · ρ · ω σN

(3.41)

Ψ = σN · (1 − λ) · ρ · ω − κρ · ρ · (1 + ρ · ω) − κO · ρ · (1 + ρ · ω − ω) − σT reu ≥ 0) when the transfer The AR division operates proﬁtably (πAR price is lower than the contribution margin of the division which is the marginal value of an additional cycle (see eq. 3.42). 3

ini ini As the asset recovery division does not exist initially, πAR = 0 and πIini = πCLSC (see eq. 3.15). The asset recovery division AR reclaims the residual positions of reu , i.e. the recovery -related costs and revenues. the proﬁt function πCLSC

50

3 Strategic Impact of Closed-Loop Supply Chains

σT ≤ σN · (1 − λ) · ρ · ω − κρ · ρ · (1 + ρ · ω) − κO · ρ · (1 + ρ · ω − ω) (3.42) A satisfactory transfer price for both divisions exists when equations 3.40 and 3.42 are simultaneously respected. This condition holds when the transfer price can be set between the lower and upper-bounds C P deﬁned above. Assuming (1 − λ) · ρ · ω · σN − κM N − κ N − κN ≤ σ T ≤ σN · (1 − λ) · ρ · ω − κρ · ρ · (1 + ρ · ω) − κO · ρ · (1 + ρ · ω − ω), a satisfactory transfer price for both divisions can be set when equation 3.43 is true.4

C P O κρ + κO · (1 + ρ · ω) ≤ (1 − λ) · ω · κM (3.43) N + κN + κN + ω · κ

On the basis of these results, a satisfactory transfer price σT for both divisions is to be found when the beneﬁts of reuse (procurement and disposal savings, minored by the rebate on reused items) are superior to the additional costs of reuse (reverse logistics and disposal costs). As a result, this transfer price exists when reuse turns out to be proﬁtable for the company as a whole. Without a compensation of the cannibalization eﬀect through additional revenues for the division in charge of the initial sales, product reintegration will lead to internal conﬂicts. Incentive alignment is thus a critical issue to avoid intra-organizational conﬂicts hindering the development of closed-loop supply chains.

4

C P (1−λ)·ρ·ω· σN −κM N −κN −κN

≤ σN ·(1−λ)·ρ·ω−κρ ·ρ·(1+ρ·ω)−κO ·ρ·(1+ρ·ω−

C P ρ ω) ⇔ (1−λ)·σN ·ρ·ω−(1−λ)· κM N +κN +κN ·ρ·ω ≤ (1−λ)·σN ·ρ·ω−κ ·ρ·(1+ρ·

C P ω)−κO ·(1+ρ·ω−ω) ⇔ κρ ·ρ·(1+ρ·ω)+κO ·(1+ρ·ω−ω) ≤ (1−λ)· κM N +κN +κN ·ω

4 Competition in Closed-Loop Supply Chains

The American Remanufacturing Association (AMRA) estimates the number of people employed in asset recovery companies to surpass 500,000 people in the United States of America (Steinhilper, 1998). The fact that these companies are, in their great majority, private-owned, not subsidized and not bound to Original Equipment Manufacturers, is an obvious proof pointing out the ﬁnancial attractiveness of refurbishing. Nevertheless, OEMs are often reluctant to start an own asset recovery program before it becomes clear that free-riders are cannibalizing their own products. In this context, section 4.1 will provide examples of manufacturers facing the competition of external ﬁrms and their reaction to this cannibalization threat. Generally speaking, an OEM can either try to preempt or to accept the competition of independent refurbishers. Furthermore, the OEM faces the choice of either destroying the reclaimed cores or to overtake remanufacturing activities. A various range of publications which aim at providing insights into these problems will be presented in section 4.2. With respect to the shortcomings of previous approaches and the managerial requirements to quantitative decision models, we extend the strategic model developed in chapter 3 to competitive issues. After this, we will draw conclusions in section 4.3 concerning the optimal reaction of a manufacturer facing competition in closed-loop supply chains.

52

4 Competition in Closed-Loop Supply Chains

4.1 External Competition as a Signal of Proﬁtability 4.1.1 Evidence from Current Practice Toner Cartridges Manufacturers of printing systems, for instance HP or Canon, face a price-based competition in their primary market: the laser printer market. In their buying decision, buyers of such devices usually only take the retail price into consideration and tend to oversee the total lifecycle costs. Thus, OEMs accept to lose money in the primary market (printers) to generate their proﬁts in the after-market when customers are already locked in. However, the proﬁtability of this strategy is being threatened by remanufacturers selling similar cartridges at a lower price and therefore partly cannibalizing OEMs’ sales. Although remanufacturers are not able to copy the cartridge architecture, they are able to refurbish them by replacing three parts: tank (toner container), wiper blade (which removes the toner surplus from the paper) and the organic photo conductor (opc). Since both remanufacturers and OEMs procure their parts from the same suppliers, quality discrepancy is minimal between new and refurbished cartridges. Eﬀorts of some manufacturers to limit the remanufacturability of their products by implementing proprietary technologies have failed so far. Technical solutions such as proprietary chips have for instance been forbidden by local regulators or hindered by non-governmental organizations.1 In order to redirect the return ﬂows to their own recovery centers, all printer manufacturers now provide pre-stamped packaging for returning used cartridges and some even oﬀer vouchers for the next cartridge. According to the environmental reports of HP (HP, 2004), returned cartridges are shredded and recycled. The reclaimed materials (especially plastics) are then used for the production of new toner cartridges. Brake Systems A worldwide supplier of brake systems located in Germany, faces a similar problem as the laser printer OEMs. Remanufacturers in developing countries have started reclaiming brake systems from end-of-life trucks in order to resell them through uncontrolled channels under the OEM’s name for a fraction of the initial price. For security and image reasons, this ﬁrm has introduced a deposit system in which retailers are 1

Majumder and Groenevelt (2001) provide a comprehensive history of the American legislation.

4.1 External Competition as a Signal of Proﬁtability

53

Table 4.1. Competition matrix for laser printers Actor Market OEM 3rd Party Initial printer + new cartridge – After sales (high-end) new cartridges – After sales (low-end) reﬁll reﬁll

obliged to send the used cores back to an OEM-controlled return point center. Although a great majority of the returned cores are shredded and molted nowadays, the OEM is looking for new ways to remanufacture brakes in order to satisfy the after-sales demand in developing countries. Table 4.2. Competition matrix for brake systems Actor Market OEM 3rd Party Initial brake system (built-in) – After sales (high-end) brake system – After sales (low-end) – refurbished brakes

Tires Tire retreading has already been practiced for long. Chapter 6 attests in this context that retreaded tires increase resource eﬃciency by a factor of four since the value embedded in the casing is retained. The current market share of retreaded truck tires yields approximatively 42% in Germany and 12% for summer car tires (see chapter 6). The relative importance of remanufactured tires can be explained by the fact that these tires have the same quality as low-end tires and hence provide a cheap alternative for buyers. As the retreading process is not speciﬁc,2 free-riders already entered the secondary market, despite very low batch sizes compared to the OEMs. Thus, tire manufacturers are forced to compete against free-riders to protect their margins. Michelin, for instance, owns a brand which only supplies consumer retreads. 2

The process is well known and the technology is publicly available.

54

4 Competition in Closed-Loop Supply Chains

OEMs also provide truck carriers with service contracts in which a ﬁxed number of kilometers is guaranteed, including two or three retreads (Deierlein, 1997). Car spare parts Spare parts sales represent 10-20% of an automobile manufacturer’s total sales (Ihde et al., 1999). OEMs try to keep control over this secondary market and hinder the activity of third parties. The European directive 98/71/EC which has initially been promoted by the automotive lobby, could give an OEM exclusive design property rights on his visible parts. The market for these protected parts such as car panes, lighting and body parts accounts for 10 billion Euros in the European Union (GVA, 2004). Due to a high degree of freedom for each European country concerning the application of the directive, OEMs have not managed to create a monopolistic situation in the spare parts market but only yield a total market share of 74% for visible parts and even as little as 55% when including all other wearing parts (GVA, 2004).

4.1.2 OEMs’ Competitive Leverages Generally speaking, an OEM has the choice between three competitive strategies: do nothing and left the market for used items to refurbishers, accept the presence of these free-riders and compete with them, or otherwise preempt the entry of external competitors. In the case of a competition between OEM and an independent remanufacturer called EXT , we identify two potential leverages to improve a manufacturer’s situation against EXT . First, he can inﬂuence the return behaviour of customers by either providing a buyback incentive for end-of-cycle items or charging a deposit fee on the retail price. Another possibility to lower the competition intensity is to complicate the disassembly process and therefore to reduce the proﬁt potential of the free-rider (see Toktay and Ferguson, 2005). Instead of competing, one manufacturer can decide to preempt the entry of free-riders in order to avoid sharing the residual value of used items with them. Leasing is for instance a contractual means to block other ﬁrms’ access to the used cores which is used by copier producers (Thierry et al., 1995) or tire producers (Deierlein, 1988). An OEM may also introduce a buyback incentive which surpasses the cost margin of external remanufacturers. By doing this, rational customers would return the cores solely to the OEM while free-riders could not bid as

4.1 External Competition as a Signal of Proﬁtability

55

much without making losses. An ex-ante deposit at the preemption level should also yield to similar results. Finally, the OEM might prefer to destroy all cores instead of reusing them, as for instance the brake systems manufacturer mentioned above does. Intuitively, this strategy seems suspect since the existence of external refurbishers is a sign that a demand exists for reclaimed items. In order to analyze which strategy is proﬁt-maximizing for Original Equipment Manufacturers, we ﬁrst have to structure the evoked cases which will be investigated and compared in section 4.3. The competition between OEM and independent remanufacturers is not new and has been therefore object of numerous publications during the past ﬁfteen years. The next section will present their main insights and give emphasis on their ability to determine which strategy listed in table 4.3 should be followed by Original Equipment Manufacturers. The scenarios mentioned in table 4.3 will be then benchmarked with help of chapter 3 to determine under what extent a complete entry preemption is superior to an incomplete entry preemption (compete strategies). Table 4.3. OEM’s competitive options against external remanufacturers Strategy

Leverage

do nothing x compete buyback option compete ex-ante deposit fee compete increase disassembly costs compete destroy cores preempt destroy cores preempt leasing preempt buyback option preempt ex-ante deposit fee preempt cut component supply

OEM’s proﬁt function passive πOEM cpt:buy πOEM cpt:dep πOEM cpt:dis πOEM cpt:des πOEM pre:des πOEM pre:lea πOEM pre:buy πOEM pre:dep πOEM pre:cut πOEM

56

4 Competition in Closed-Loop Supply Chains

4.2 How to Deal with Independent Refurbishers: A Literature Overview While the development of ﬁrm-centric decision models has been one of the main focuses of the reverse logistics research,3 very little attention has been paid to the way a manufacturer should deal with external companies, especially when these are facing the competition external refurbishers. To the knowledge of the autor, six publications address speciﬁcally this topic. These will be presented below.4 Hollander and Lasserre (1988) analyze the reaction of a primary producer facing competition through recyclable materials. The authors show that OEMs have an incentive in reclaiming the scrap materials in order to preempt market entry of independent recyclers. Due to his monopolistic position, the primary producer can pass the reclaiming costs to end-customers, so that ﬁnally the threat of entrants may rise the price above the full-monopoly level. Grant (1999) investigates the market power of a main American virgin aluminium producer before World War II. His simulations encompass a price-leader (producer of virgin metal) and a price-taking metal recycler. Surprisingly, he shows that the competition in the secondary market was, during that period, welfare-reducing because recycled aluminium used to be more costly in production than virgin one. Debo et al. (2005) introduce the concept of technology choice in order to assess remanufacturing strategies. The authors identify the incentive problem linked to the remanufacturability of a product: the remanufacturability level is positively correlated with the production costs. Thus, OEMs must burden higher costs in the ﬁrst product cycle but can provide cheaper products in the next periods and then appeal new segments. Debo et al. develop their model in a monopolistic environment but point out that a consideration of free-riders is necessary in the next periods since the OEM will not necessarily be able to harvest his eﬀorts in the second period when free-riders refurbish his products. Majumder and Groenevelt (2001) address the modeling of competition in secondary markets and consider one OEM and one remanufacturer in a two-period game. The purpose of the model is to determine 3 4

We refer to de Brito and Dekker (2004) for a current models’ overview. Savaskan et al. (1999) investigate which actor should overtake the collection of used products. Depending on the competitive structure of the collection network, the authors conclude that in the presence of a multitude of retailers, an OEM should collect used products itself and avoid using retailers as intermediaries. Since the authors do not deal with the OEM/remanufacturer competition, this publication is not mentioned in the literature overview.

4.2 How to Deal with Independent Refurbishers: A Literature Overview

57

the Nash Equilibrium for the second period when both actors compete against each other. The outcome of the game is determined by three variables: the OEM’s retail price and his production quantities for new and for refurbished items. The remanufacturer only reacts to the OEM’s decisions. Majumder and Groenevelt compare the situations in which an OEM remains passive, acts as a monopolist or accepts competition. They show that, in case of competition in the secondary market, OEM and remanufacturer should cooperate in order to keep the return incentives low. Nevertheless, the monopolistic solution would be, as expected, proﬁt-maximizing for the OEM. Robotis et al. (2004) investigate an independent remanufacturer reselling its products in secondary markets. The authors’ approach, which can be interpreted as a newsvendor problem, makes a distinction between two quality levels on both input and output ﬂows. Robotis et al. conclude that resellers should procure cores in a good quality state rather than only reclaiming cheaper cores of minor quality. They also introduce as model variable a yield quality level that help to identify the minimal quality level of a core for a proﬁtable refurbishing. A more comprehensive analysis has been performed by Toktay and Ferguson (2005) who investigate under which conditions entry preemption (with either own remanufacturing or core destruction) is more proﬁtable than competing against free-riders. With help of a two-period game, Toktay and Ferguson conclude that products cheap to remanufacture should be remanufactured internally because of their attractiveness for independent refurbishers. The authors underline the misperception of OEMs when assessing the attractiveness of asset recovery: The proﬁtability yield is lower for external refurbishers than for manufacturers which have to charge additional opportunity costs for lost sales. However, due to the complex structure of the decision model, it is diﬃcult to give general insights about the best response to external competitors’ threat. This shortcoming, which is common to Majumder and Gronevelt as well as to Toktay and Ferguson, is mostly issued from the two-period structure of the decision models in which manufacturers are taken as the dominant actor in a Stackelberg duopoly. Given the limitations of multi-periodic models, we will extend the closed-loop model of chapter 3 to ﬁnd out which competitive strategy is optimal for an Original Equipment Manufacturer.

58

4 Competition in Closed-Loop Supply Chains

4.3 Analyzing Best Responses Strategies for Manufacturers We extend the strategic model developed in chapter 3 to integrate the potential competition of an external remanufacturer. We assume that the OEM reclaims a fraction α of the returns while the competitor EXT obtains (1 − α). Furthermore, we split the collection costs κρ into transportation costs κRF and incentive costs of the OEM (κIo ) and of the external remanufacturer (κIe ). Given this, we divide the rem of the monopolistic manufacturer into former proﬁt function πCLSC rem rem . two proﬁt functions πOEM and πEXT rem πOEM

C = σN · 1 + (1 − λ) · ρ · α · θ − κM N + κN · 1 + ρ · α · θ

·(1 − β) − κPN − κPR · ρ · α · θ − κRF + κIO · ρ · α · (1 +

rem · Q ρ · θ) − κ · ρ · α · (1 − θ · β + ρ · θ) · QN = πOEM N O

(4.1)

rem C πEXT = σN · (1 − λ) · ρ · (1 − α) · θ − κM N + κN · ρ · (1 − α) · θ ·

(1 − β) − κPR · ρ · (1 − α) · θ − κRF + κIE · ρ · (1 − α) · (1 +

rem · Q ρ · θ) − κO · ρ · (1 − α) · (1 − θ · β + ρ · θ) · QN = πEXT N

(4.2) Assuming that refurbishing is proﬁtable but that the OEM does nothing to control and reclaim his returned items, his return fraction α would be zero and the external remanufacturer would reclaim all cores. passive In this case, the OEM’s proﬁt function would be πOEM .

passive C P = σ N − κM πOEM N − κN − κN · QN

(4.3)

By doing this, the OEM transfers the proﬁts from remanufacturing to the free-rider. This passive strategy leads also to a worse proﬁt ini than in the monopolistic remanufacturing case (πCLSC ) as the margin rem of remanufacturing πEXT is positive. Logically, this case is also worse ini where a manufacturer neither than the monopoly case scenario πCLSC H L remanufactures nor carries a green fee since QN ≤ φ σ+φ . AccordN ing to these observations, the OEM should act in order to improve

4.3 Analyzing Best Responses Strategies for Manufacturers

59

his competitiveness in the presence of external competitors. The ﬁrst possibility is to accept the competition and to provide an incentive for customers to return the products back. 4.3.1 Competitive Asset Recovery Strategies One straightforward strategy to reclaim cores is to provide incentives for customers to do so. This strategy is followed for instance by manufacturers of toner cartridges who try to avoid that independent reﬁllers get access to the empty cartridges (B¨ urgermeister et al., 2003). Tire manufacturers also oﬀer a buyback option for end-of-cycle tires depending on their age (Ferrer, 1997). Applying this strategy to an cpt:buy rem . which is equivalent to πOEM OEM’s proﬁt function, we obtain πOEM Starting from a passive strategy, the manufacturer’s proﬁt will increase cpt:buy passive − πOEM > 0) if the collection and acquisition costs are lower (πOEM than the contribution margin of remanufacturing, i.e. when remanufacturing is also proﬁtable for the free-rider (eq. 4.4).

κIo + κRF ≤

θ · σN · (1 − λ) −

−

κO

κM N

+

κC N

· (1 − β) −

κPR

1+ρ·θ · (1 − θ · β + ρ · θ) 1+ρ·θ

(4.4)

For an OEM, the main shortcoming of buyback incentives is that he competes with the same means as external remanufacturers despite his strong initial position. A deposit on the sales of new products (ex ante deposit) would oblige the free-rider to pay this fee to the customer back to reclaim the core. Nevertheless, this operation would be neutral for initial sellers. As a results, switching from an ex post buyback option to an ex ante deposit would improve an OEM’s proﬁt function by

cpt:dep cpt:buy − πOEM = κIo · 1 + ρ · θ · α · (1 − λ) · QN ≥ 0 πOEM

(4.5)

which corresponds to the additional revenues charged on every unit sold by the Original Equipment Manufacturer. Thus, manufacturers are always better oﬀ applying a deposit fee rather than competing without one. Since 5

cpt:dep ∂πOEM ∂κIo

= 1 + ρ · α · θ · (1 − λ + ρ) − 1 > 0,5 we notice that

∀(ρ < 1) ∩ (α < 1) ∩ (θ > 0)

60

4 Competition in Closed-Loop Supply Chains

the deposit is only partly neutral for manufacturers but has a positive contribution. In fact, the lower the return rates in an OEM’s channel, the higher the deposit’s contribution to proﬁt. This situation has also been observed in Germany after the introduction of deposit fee for one way packaging. German retailers, who collected the fee, have generated more than 450 million EUR deposit surplus during the ﬁrst year after introduction as customers still threw away their packagings instead of returning them (estimation by WDR, 2004). The refurbishing activities performed by free-riders are possible as long as the disassembly process is manageable in a short lap of time and does not require speciﬁc equipment. Cartridges reﬁllers, depending on their level of competence, will only reﬁll the empty cartridges with toner and perhaps replace some generally available parts of the cartridge to ensure a satisfactory printing quality. Tire retreading is also a well-known technology which can be easily performed by small independent companies as long as these have access to stress-reliable rubber. In this context, increasing free-riders’ costs provides a legal hurdle to hinder their activity. Manufacturers can make use of their product development know-how to increase the diﬃculty of the disassembly process. Such a measure would increase the refurbishing costs κPR by a factor δ > 1 but would impact both actors. The proﬁt function of an OEM already following a deposit strategy would decrease by κPR · δ · α for every new unit sold. An alternative strategy followed by the brake systems manufacturer BrB or the printing division of Hewlett-Packard consists in the destruction of the cores collected. This strategy seeks to reduce the number of reusable cores available on the market and therefore to increase the turnover of new products. The OEM discards the returned cores with a rate of 1 − τ so that the fraction ρ · α · τ is resold in the second , cycle by the manufacturer. The quantity of new items sold Qcpt:des N depicted in eq. 4.6, is greater than QN when τ < 1. As a consequence, the more cores are reclaimed, the higher is the turnover of new items as

∂Qcpt:des N α

> 0.6

Qcpt:des = N

φH + φL

σN · 1 + (1 − λ) · ρ · θ · (1 − α + α · τ )

(4.6)

Logically, the quantity of new items sold is negatively correlated with τ but is mostly determined by the fraction α reclaimed by the OEM. Therefore, it seems obvious that a manufacturer following a core 6

∀(ρ > 0) ∩ (θ > 0) ∩ (λ < 1)

4.3 Analyzing Best Responses Strategies for Manufacturers

61

destruction strategy should reclaim the majority of the cores in order to . The manufacturer’s proﬁt signiﬁcantly increase his initial sales Qcpt:des N cpt:des cpt:des function changes to πOEM = πOEM · Qcpt:des if the manufacturer also N removes from the market the fraction refurbished by the free-rider.

cpt:des C = σN · 1 + (1 − λ) · ρ · α · θ · τ − κM πOEM N + κN · 1 + ρ · α ·

θ · (1 − β) · τ − κPN − κPR · ρ · α · θ − κRF + κIo · ρ · α ·

1 + ρ · (1 − α) · θ + ρ · α · θ · τ − κO · ρ · α · 1 − θ · β · τ

+ρ · θ · (1 − α + α · τ )

· Qcpt:des N

(4.7)

Given eq. 4.7, we are able to ﬁnd out whether core destruction is beneﬁcial by diﬀerentiating the proﬁt function on τ . Hence, if a decrease of cpt:des , a manthe reintegrated core fraction τ has a positive impact on πOEM ufacturer should prefer this strategy instead of performing refurbishing. For reasons of clarity, we make use of C = 1 + (1 − λ) · ρ · θ · (1 − α + α · τ ) which represents the average number of use phases (or cycles) serviced by a new product. We further isolate themarginal income per reman rem M C = σN · (1 − λ) − κN + κN · (1 − β) − κPR − ufactured unit M I

κRF + κIo · ρ · α − κO · (ρ · θ − β). We simplify 4.8. cpt:des ∂πOEM

∂τ

cpt:des ∂πOEM ∂τ

to obtain eq.

φH + φL · ρ · α · θ cpt:des rem · C · M I − (1 − λ) · π (4.8) = OEM σN · C 2

cpt:des ∂πOEM π cpt:des M I rem <0 ⇔ < OEM (4.9) ∂τ 1−λ C As the ﬁrst term of 4.8 is always positive due to our parameter settings, we observe that an OEM is better oﬀ destroying his collected cores when the adjusted marginal income is lower than the average income per use phase (see eq. 4.9). In other words, ceteris paribus, the attractiveness of core destruction is positively correlated with the price rebate λ. In the case of a refurbished product without rebate (λ = 0), we derivate from eq. 4.9 the insight that asset recovery is proﬁtable when the marginal income surpasses the average income per use phase. This situation happens when the margin generated on the sales of new items surpasses by far the margin on refurbished cores (e.g. for low-end consumer appliances such as brown goods).

62

4 Competition in Closed-Loop Supply Chains

We can apply and verify these insights to the toner cartridges case for which we use the ﬁndings of B¨ urgermeister et al. (2003) who analyzed the economics of toner cartridges refurbishment. Table 4.4 gathers the parameters of a HP LaserJet 4050 toner cartridge (10,000 pages) and are based on experts interviews with independent refurbishers. Table 4.4. Parameter settings for toner cartridges C σN 100 κM 5 κP 13 θ 1 N 2 κN N O κRF 1 κIo 6 κP 10 κ 2 β 0.5 R H L λ 0.3 ρ 1 φ 1,000 φ 4,000 α 0.5

Using the available information, the isolated marginal income of remanufacturing is M I rem = 53 Euros while the price for reﬁlled cartridges amounts for 70% of the initial price (1 − λ = 0.7). The avcpt:des erage income per unit sold πOEM equals 73.7 + 21.2 · τ and the average number of cycles amounts to 1.28 + 0.28 · τ . Given fλ=0.3,τ = (1−λ)·π cpt:des

OEM = 53− 51.59+14.84·τ M I rem − C 1.28+0.28·τ . fλ=0.3,τ is strictly positive and concave on τ ∈ [0, 1] so that cartridge refurbishing is proﬁt-maximizing under these settings.7 HP would be therefore better oﬀ refurbishing his toner cartridges rather than shredding them. We now determine the price yields for which EXT enters the market. As EXT does not bear costs for the initial cycle, its margin is equivalent to the marginal income of refurbishing of the OEM: M I rem = 83 − 100 · λ. This ﬁgure is positive as long as the price for reﬁlled cartridges remains above 17 Euros.8 Therefore, external refurbishers will compete against manufacturers when customers are willing to pay more than 17 Euro for reﬁlled items. It is interesting to notice that the proﬁtability yields of cartridge refurbishing diﬀer between the OEM and the free-riders. Hence, the proﬁtability indicator fλ,τ is negative when the price rebate rises above 0.54779 . Thus, the manufacturer should destroy the reclaimed cartridges when the reﬁlled cartridge would be sold for 45 Euros or less. Figure 4.1 depicts the OEM’s proﬁtability patterns fλ,τ which is negative when core destruction is ﬁnancially more advantageous than refur2 7 ∂ πOEM ∂2τ 8 rem

cpt:des

9

) 593.6 = − (1.28+0.28·τ + 7.84·(73.7+21.2·τ <0 )2 (1.28+0.28·τ )3 MI ≤ 0 ⇔ λ ≥ 0.83 ⇔ 1 − λ ≤ 0.17. fλ=0.5477,τ = 0 ∀τ .

∀τ ∈ [0, 1].

4.3 Analyzing Best Responses Strategies for Manufacturers

63

bishing. We notice that, all things being equal, the level of the fraction returned to the OEM τ plays a minor role in the value of fλ,τ .

Fig. 4.1. Proﬁtability yields of core destruction with respect to τ and λ

The present computations enable us to draw preliminary conclusions concerning the beneﬁts of generic competitive strategies. First, an Original Equipment Manufacturer should make use of his market power to collect an ex ante deposit fee under the threat of potential entrants. Products with a margin higher in the reintegration phase than in the initial phase should be rather reintegrated while products with a low margin in the second cycle and a high initial margin should be rather destroyed. An increase of the refurbishing costs would be avoided as this measure penalizes both actors and is heavily criticized by non-governmental organizations for environmental reasons.10 4.3.2 Entry Preempting Asset Recovery Strategies An OEM can avoid the existence of proﬁt-squeezing free-riders by preempting their entry in the market. Copier manufacturers apply contrac10

See e.g. Majumder and Groenevelt (2001).

64

4 Competition in Closed-Loop Supply Chains

tual means to reclaim their products at their end-of-cycle. The introduction of leasing contracts has, for instance, not been only motivated by the necessity to manage the timing of the reverse ﬂows but also because independent refurbishers were already reselling used copiers (Thierry et al., 1995). As a result, the proﬁt function in case of leasing pre:lea rem πOEM is equivalent to the monopolistic proﬁt function πCLSC investigated in chapter 3. A similar monopolistic position might also be achieved by cutting the supply of spare parts for refurbishing purposes. Nevertheless, this strategy solely goes up when the spare parts required are diﬃcult to imitate, for instance because of patent laws owned by the OEM or when the manufacturer is able to contractually forbid his suppliers to sell the replaceable parts to unaﬃliated refurbishers. Most manufacturers of toner cartridges have focused their competencies on the development of high resolution printing systems and have outsourced the manufacturing of their cartridges’ components to suppliers (B¨ urgermeister et al., 2003, Siestrup, 1999). This low vertical integration level leads to a situation where cartridges producers cannot forbid refurbishers getting access to the modules required for the update of a used cartridge. This strategy also fails in the BrB case as refurbishers limit themselves to the disassembly and cleaning of the parts, thereby avoiding to replace any worn or defective part. The resulting cores are far from complying with BrB’s security requirements but are sold with success through grey markets to customers in developing countries (Seitz, 2005). For similar reasons, tire manufacturers not involved in retreading prefer to supply retreaders with original, compatible, rubber for the tread application. This measure leads to a better bonding quality of the tread rather than with all purpose rubber mixes. The probability of service failure of retreaded tires and the resulting problems with quality signaling are minimized consequently. Hence, technical means to reduce the competition seem to be inadequate for Original Equipment Manufacturers. While we assumed in the previous section that the return incentives were low enough to allow remanufacturers to turn proﬁts, an OEM is able to set the return incentive at a level for which it is impossible for free-riders to enter. The preempting return incentive would at least amend the proﬁt of an independent remanufacturer and thus amount for κImax (eq. 4.10).

κImax =

C P σN · (1 − λ) · θ − κM N + κN · θ · (1 − β) − κR · θ

1+ρ·θ

4.3 Analyzing Best Responses Strategies for Manufacturers

−

κO · (1 − θ · β + ρ · θ) − κRF . 1+ρ·θ

65

(4.10)

Assuming that the manufacturer would oﬀer κImax as buyback option and that all cores are returned, his proﬁt function would be reduced pre:buy M C P to πOEM = σN −κN −κN −κN ·QN . Thus, the income of the second phase would be amended, the OEM generates as much proﬁts as in the passive . Given this, a preempting buyback option, which passive case πOEM is worse as any competition case, should not be followed to preempt pre:buy cpt:buy − πOEM between a competitionentry. In fact, the proﬁt gap πOEM based and a preempting buyback strategy is always negative when a remanufacturer would turn beneﬁts.11 As a conclusion, the presence of independent refurbishers is a sign of the proﬁtability of remanufacturing. Given this, an OEM has to react in order to protect his proﬁts. The present investigations have shown that two alternatives may help the manufacturer in this aspect. • Leasing does not transfer to the user the right to resell the lease object which remains in this context the property of the manufacturer. With the help of such contractual means, manufacturers preempt legally the entry of free-riders while avoiding to charge a deposit fee. Most of the functional goods identiﬁed by Lebreton and Tuma (2003) are leased due to the attractiveness of remanufacturing. • Deposit fees. When customers are not willing to lease products or to accept service-selling concepts, an ex-ante deposit fee might hinder the possibility of free-riders to make proﬁts while ensuring high return rates. Core destruction appears to be attractive when OEM generate better margin with their initial rather than with the recovered products. Nevertheless, according to the data available, it seems that the proﬁts of refurbished cartridges are too high to justify their destruction. The computations also show that manufacturers relying on ex post buyback incentives to guide product returns are better oﬀ keeping the incentives low in order to save their margins.

11

pre:buy cpt:buy πOEM − πOEM =

(φH +φL )·ρ·α σN ·(1+ρ·θ·(1−λ))

RF β) − κP + κIo R ·θ− κ

C · − σN · (1 − λ) · θ − κM N + κN

· (1 + ρ · θ) − κO · (1 − θ · β + ρ · θ) .

· θ · (1 −

5 Strategic Network Planning in Closed-Loop Supply Chains

Once the strategic pertinence and the ﬁnancial potential of closed-loop supply chains have been identiﬁed by a manufacturer, his managers have to decide whether to setup a reverse network or not. The preparation phase preliminary to the decision-making process, also called planning process (Domschke and Scholl, 2003), requires the use of mathematical models in order to ensure that the decisions are made with respect to their impact on the whole supply chain (Chopra and Meindl, 2001). The existence of reverse ﬂows represents an additional vector of complexity as the interfaces between forward and reverse supply chains should be also determined (Fleischmann, 2001). According to Gutenberg (1983), only a holistic planning approach, called simultaneous total planning and encompassing all supply chain decisions from the network setup to the daily routing planning, would ensure the consistency of the decisions. Since holistic planning is not adequate for real life applications,1 pragmatic planning systems follow a hierarchical structure in line with the management hierarchy within corporations (Anthony, 1965, Schneeweiss, 1999). The seminal work of Hax and Meal (1975), providing a hierarchical production planning framework for a tire manufacturer, has been followed by numerous articles extending this hierarchical decision-making framework (see Steven, 1994 or Miller, 2000 for a literature review). Following this logic, Miller (2000) and Fleischmann et al. (2005) classify the supply chain tasks with help of two dimensions also depicted in ﬁgure 5.1: The scope of the decisions (i.e. which time horizon is taken into account), and the supply chain process concerned (procurement, manufacturing, distribution and sales). The relevance of this classiﬁcation for practical applications is conﬁrmed by the fact that 1

We refer to Steven (1994) for an analysis of holistic planning’s shortcomings.

68

5 Strategic Network Planning in Closed-Loop Supply Chains

the majority of advanced planning systems (APS) providers have also adopted this framework for their planning software (Rohde et al., 2005). According to Fleischmann et al. (2005), the strategic network planning process deals with the setup of capacities and the identiﬁcation of profitable ﬂows of materials, components, work-in-process and ﬁnal products. Since the decisions related to the network infrastructure commit the ﬁrm on the long-term and generally bind high amounts of capital, the strategic planning process is a task of the top management. On the basis of the existing network, the objective of tactical supply chain planning is to ﬁnd a production and distribution plan enabling an eﬃcient use of the available capacities. The directives obtained from the tactical planning process provide a means to coordinate the supply chain ﬂows on a yearly basis (Chopra and Meindl, 2001). Finally the operational planning process applies the production and distribution plans issued from the tactical planning level to concretize the day-to-day utilization of the available capacity.

Fig. 5.1. Supply chain planning matrix (modiﬁed from Rohde et al., 2005)

Given the objectives of this monograph, we will limit our research scope to the strategic planning process. In this context, the ﬁrst objective of this chapter is to assess to what extent closed-loops would modify the nature of the strategic network planning process. For this purpose, the decisions related to the setup of a reverse network have to be identiﬁed to develop practical strategic planning models. By now, only Schultmann et al. (2002) have investigated the implementation of reverse logistics into supply chain planning systems. Nevertheless,

5 Strategic Network Planning in Closed-Loop Supply Chains

69

their proposed extension of the supply chain planning matrix is mainly focused on disassembly planning and provides restricted insights into the other strategic decisions faced in closed-loop supply chains. The Copy Magic case of Thierry et al. (1995) and Thierry (1997) depicts the sequential extension of a manufacturer’s planning scope towards an integrated closed-loop network planning. Initially, Copy Magic started with the reuse and remanufacturing of end-of-cycle copiers when free-riders were already reclaiming and refurbishing Copy Magic’s used appliances with success. With respect to the external competition, the initial question of the Original Equipment Manufacturer was to determine if refurbishing would be better for the contribution margin of the company than leaving the segment to competitors (see chapter 4). After the setup of an active return management scheme, the manufacturer faced more returns than he could reintegrate and had thus to deal with a core surplus which had to be disposed of. This unforeseen problem had to be solved in order to minimize the side costs of remanufacturing and thus keep the proﬁtability of the product recovery program at a satisfactory level. Consequently, the manufacturer started to improve the recyclability of his products so that the material ﬂows could also be reintegrated instead of being landﬁlled at high cost. Once the proﬁtable products and recovery activities had been launched successfully, the OEM was able to improve the eﬃciency of his network by reengineering the ﬂow structure of the reverse chain. In fact, the ad-hoc development of the asset recovery program led to a situation where the recovery nodes could be moved to other places in order to save costs. Oce, which has passed through the same evolution process towards closed-loop, refurbishes for instance a part of his product in the Czech Republic instead of Holland due to the lower labor costs in Eastern Europe (Krikke et al., 1999). As the copier industry shows, closed-loop supply chains are developed sequentially starting from an isolated business opportunity assessment (here refurbishing) to a comprehensive network planning embedding logistics and processing costs (see ﬁg. 5.2). Each closed-loop development milestone mentioned above triggers new problems that need to be tackled. It is interesting to notice that each step towards an integrated planning of the closed-loop addresses a speciﬁc supply chain dimension: The product (step 1), the process (step 2) and eventually the location (step 3). As a consequence, strategic closed-loop network planning does not diﬀer from the classical strategic network planning described by Goetschalckx and Fleischmann (2005). In fact, their scope is equivalent since strategic closed-loop network

70

5 Strategic Network Planning in Closed-Loop Supply Chains

Fig. 5.2. Steps towards closed-loop supply chains: An OEM’s perspective

planning is concerned with the deﬁnition of the location and activity of asset recovery centers2 as well as with the identiﬁcation of economically recoverable products3 . On the basis of the three dimensions depicted in ﬁgure 5.3, we are now able to structure the existing publications related to closed-loop strategic network planning.

Fig. 5.3. Strategic network planning dimensions in closed-loops

2 3

the supply chain nodes the supply chain ﬂows

5.1 Strategic Closed-Loop Network Planning: A Review

71

5.1 Strategic Closed-Loop Network Planning: A Review Product Selection As already evoked in chapter 2, OEMs are seldom vertically integrated and generally outsource the development of components to suppliers (Prahalad and Hamel, 1990, VDA, 2002). As a result, while the recycling of materials would not impact directly a manufacturer’s proﬁt structure, the reintegration of used components or products would affect his sales and is therefore of prime interest for Original Equipment Manufacturers. Given the immediate business impact of remanufacturing, a wide range of publications provides strategic planning models assessing the proﬁtability of a potential core refurbishing program for an isolated product. Siestrup and Haasis (1997), Ferrer (1997) and Lebreton et al. (2004) present quantitative approaches to determine the optimal number of retreading cycles for truck tires with respect to the discounted life-cycle value of a new tire. Nevertheless, these three approaches are not taking into account the cannibalization eﬀect between new and retreaded tires and are assuming de facto the existence of two independent markets. In a recent article, Fandel and Stammen (2004) have published a strategic network planning model which has the particularity of integrating the product development phase within a proﬁt-maximizing approach. Nevertheless, one of the main shortcomings of their model lies in the diﬃculty of retrieving the data sets required. While the strategic assessment model developed in chapter 3 gives advice about the proﬁtability impact of product-based closed-loop supply chains, it remains at an aggregated level. The decision to start an asset recovery program aiming at refurbishing used products needs further investigation and especially the integration of time in order to catch the dynamics of remanufacturing. Remanufacturing programs, for instance, have to cope simultaneously with the psychological, qualitative and technological obsolescence (see chapter 2) which are also closely interacting and varying over time. The complexity of component reintegration issues makes it necessary to deal with these issues within a comprehensive framework in order to identify which critical factors enable or hinder the reuse of components. Since the scope of the products studied is wide (see de Brito and Dekker, 2004 as well as Flapper et al., 2005, for a comprehensive overview), these often present few commonalities concerning their characteristics. Hence, product-based cluster building seems a priori inappropriate to make a judgement concerning the reuse potential of a component. Despite this diversity, envi-

72

5 Strategic Network Planning in Closed-Loop Supply Chains

ronmental publications of Stahel (1986), Steinhilper (1998) or Souren (2002) provide a framework structuring the life-cycle of every remanufactured product. The authors identify three phases with respect to the whole product life cycle: Production which transforms resources into a ﬁnal product, utilization which transforms a ﬁnal product into a used product and recovery which reclaims the resources from these returned items (see ﬁg. 5.4).

Fig. 5.4. Generic life cycle phases and remanufacturing key factors

The strategic planning approaches of Geyer and van Wassenhove (2002) as well as Lebreton and Tuma (2002, 2003) adopt the life-cycle paradigm presented in ﬁgure 5.4. The authors identify thereby the existence of three interconnected key factors at the end of each life-cycle phase: The return behavior of customers, the technological evolution and ﬁnally the market segmentation. While Geyer and van Wassenhove (2002) are developing a cost-minimization model in which recovered products have the same retail value as new ones, Lebreton and Tuma (2002, 2003) develop a proﬁt-maximization linear programming model that will be used in this chapter as the basic approach for closed-loop strategic network planning models. Recovery Path Determination In the case of a steady increase of the raw material prices, the setup of a proprietary supply chain where resources are retained in the supply chain and thus remain the property of the OEM might turn out to be a proﬁtable business on a long-term perspective. Manufacturers could, in this situation, avoid procuring the required materials on the markets since at least a fraction of these would have been kept within their own supply chain. Consequently, the determination of the recovery path and especially the decision to start recycling will depend on

5.1 Strategic Closed-Loop Network Planning: A Review

73

an OEM’s expectations concerning the price evolution of raw materials, the disposal costs and the internal prices for recovered materials. Generally speaking, an inhouse recycling will be proﬁtable when the sum of the virgin material price and its disposal cost rises above the internal costs for recycled materials (see chapter 3). Furthermore, the mutual exclusivity between remanufacturing and recycling has to be taken into account since remanufacturing impacts the residual ﬂow of components available for recycling (see ﬁg. 5.5).

Fig. 5.5. The role of material markets in proprietary closed-loops

By now, the determination of an optimal level of recovery is coupled to the disassembly planning research which seeks to determine the optimal disassembly depth for a given product. Spengler (1994) proposes a mixed-integer program based on Leontief’s linear activity analysis to ﬁnd the most proﬁtable level of disassembly of houses. Penev and de Ron (1995) propose a dismantling framework derived from a travelling salesman problem. Krikke et al. (1998) present a heuristic to ﬁnd out a proﬁtable disassembly plan. The authors adopt an equivalent approach to Spengler but enable thereby more complex disassembly sequences. Inderfurth et al. (2004) also develop a disassembly planning model close to Spengler’s model but embed also inventory management issues. The results of these planning models enable a decision-maker to ﬁnd out whether remanufacturing, recycling or a direct disposal would be proﬁt-maximizing. However, the isolated focus on disassembly planning to identify the optimal recovery path has a very critical shortcoming. In fact, all four approaches assume that the outcome of the dismantling process is sold to an anonymous market at a given price. While this assumption holds well for recycling or common mass components, an important fraction of an OEMs’ components are generally too speciﬁc to have a real market. The demand for reclaimed Mercedes engines depends for instance on the motors’ wearing patterns of the ”compati-

74

5 Strategic Network Planning in Closed-Loop Supply Chains

ble” cars currently in service. If this demand falls short, the marginal income for a reclaimed engine (its internal price) will tend to zero. As a consequence, the determination of an optimal recovery path has to be linked with the production process in which the disassembly outcome should be reused. Location Planning For manufacturers, the decision to set up a reverse channel also depends on the location of the asset recovery centers. Due to the manual nature of refurbishing, labor costs often account for the main cost driver in product-based asset recovery programs (Steinhilper, 1998). A delocalization of the disassembly and recovery activities in low cost countries could then be the only alternative to turn beneﬁts from remanufacturing. However, the optimal location of an asset recovery center is also inﬂuenced by the total transportation eﬀorts to feed the closedloop. If the main source of returned products is the main customer for refurbished items, then the cost advantage of low cost countries has to be balanced with the resulting unnecessary transport of goods and the sunk costs for the setup of an oﬀshore plant (Krikke et al., 1999, Fleischmann, 2001). Bloemhof-Ruwaard et al. (1996) ﬁrst observed that the existence of reverse ﬂows apt to be recovered might inﬂuence on the optimal location of the manufacturing plants. By analyzing the European paper industry, the authors conclude that paper recycling targets will force the producers of virgin pulp to move their production capacities from Scandinavia to central Europe where the main markets are located. Fleischmann (2001) conﬁrms these expectations on the basis of numerical results for the paper and copier industry. Furthermore, Fleischmann compares the cost-optimal network structure obtained by a sequential network design and by an integrated network design (i.e. when the production and recovery nodes are deﬁned simultaneously). He concludes that the network structure resulting from a sequential optimization is close to the integrated case when the production nodes are already located in the main markets. With the rising importance of environmental and legislative issues, a wide range of location planning models has been published during the past ten years to complete these preliminary observations. Most of them are variations of the warehouse location problem. We will limit our analysis to the models depicting real closed-loop supply chains and leave therefore the strategic planning models dealing only with the setup of reverse logistics networks out of our investigation scope.

5.1 Strategic Closed-Loop Network Planning: A Review

75

Kroon and Vrijens (1995) present a location planning model targeted to reusable containers, so that the application ﬁeld for OEMs is limited to packaging issues. Thierry (1997) develops a linear programming model based on his experience with Copy Magic/Xerox (copier manufacturer) to determine the optimal recovery path of end-of-cycle copiers. The author’s optimization model is restricted to repair and remanufacturing and assumes that reclaimed products are as good as new. Krikke et al. (1999) discuss a further business case concerning copier remanufacturing by Oce. The recovery path is thereby set in advance and limited to refurbishing for which a speciﬁc demand exists. Moreover, a discarded machine is not returned directly to Oce but brought back to the local operating company at the end of the lease contract or through active buyback measures. The local subsidiary decides whether to recover the returned machine by himself or to redirect the copier to a recovery location of the parent company. As Oce builds up a stock of returned machines to feed his pipeline without interruption, the parent company runs a pull system in contrast to his operating companies who face mostly deterministic lease returns (push system). Although the company has a worldwide presence, the recovery activities are only implemented in Europe. The goal of the analysis is to optimize a subnetwork of Oce’s global supply chain since only the location of the preparation and re-assembly units should be determined. As a consequence, the reverse chain is optimized separately from the forward chain. Assuming that processes and transportation means have no capacity limitations, the model can be interpreted as an uncapacitated two-level location-allocation problem. Jayaraman et al. (1999) present a mixed-integer programming model for the location of a given set of distribution and remanufacturing locations, type of products to remanufacture and the amount of initial stock necessary to feed the supply chain. In this context, the authors apply a cost-minimizing capacitated two-level location-allocation problem. Like for Krikke et al. (1999), time is not taken into account and the demand for remanufactured goods is given without link to the initial sales ﬂow. This situation stems from the fact that this case depicts an independent cellular phones refurbisher and not the OEM. Fleischmann (2001) develops a closed-loop strategic model which is generic enough to encompass all recovery alternatives. Fleischmann applies his network model to the paper and to the copier industry in order to investigate, amongst others, the diﬀerence between a sequential and an integrated design of the closed-loop. Once again, the network

76

5 Strategic Network Planning in Closed-Loop Supply Chains

planning model designed by the author is a cost-minimization problem for which a perfect equivalence assumption between new and recovered ﬂow is assumed. To conclude, the location planning models reviewed above have in common the shortcoming of being focused only on location planning issues. Consequently, these leave proﬁtability issues out of their analysis and do not question the utility of an asset recovery program for the original equipment manufacturer. The cost-minimization objective (coupled with the resulting assumptions) conﬁrms this point of view. As the cases of Copy Magic and Oce show us, the three dimensions of strategic closed-loop supply chain planning should be at least partly linked together during the decision-making process. However, every case of application is unique due to its political relevance within the corporation and, for instance, the degree of freedom concerning the network design. A comprehensive but adaptable planning model is thus necessary to fulﬁll a decision-maker’s needs while ensuring that the model provides applicable results. On the basis of these requirements, we will present in the following a generic OEM-centric strategic network planning model that will be stepwise extended to recovery path decisions as well as location planning issues. The modular nature of the upcoming model will ensure its applicability to the business cases presented in the second part of this monograph.

5.2 A Generic Strategic Network Planning Model Evaluating the remanufacturing potential of an end-of-cycle or end-oflife product requires a comprehensive approach coping with the complexity of this issue. Furthermore, extending the supply chain towards an integrated loop is a challenge whose ﬁnancial burden may outweigh its utility. In this context, it seems necessary not only to rely on qualitative approaches within the decision-making process but also to produce quantitative results. Linear programming appears thereby to be an adequate means to produce the required ﬁgures. Basically, the planning model should ﬁnd out whether it is proﬁtable for an OEM to launch or intensify remanufacturing programs. If the answer is negative, the model should also provide enough information to enable the identiﬁcation of the key factor undermining the proﬁtability of product refurbishing.

5.2 A Generic Strategic Network Planning Model

77

5.2.1 The Key Factors of Remanufacturing Market Segmentation The presence of a marketing channel for refurbished goods is a prerequisite for implementing remanufacturing operations. Remanufactured products are often considered to be cheap and obsolete although these products fulﬁll the same basic requirements. The high-end market segment, for which innovation and trends are a critical decision factor, tends to ignore this supply source. For this reason, remanufactured goods can only reach a part of the customers, especially low-end ones, who are keen on buying goods at a favorable price. Given this, a planning model has ﬁrstly to take market diﬀerentiation into account and secondly, it has to integrate an upper bound for the demand of a product, which meets a given quality expectation. In case of diverging reverse ﬂow patterns among market segments, it appears useful to differentiate these into single segments in order to estimate the impact of each segment on component reintegration. Return Flow Structure The construction of a reverse logistics network initially presents a critical ﬁnancial burden so that OEMs are keen on cooperating with already established logistics providers to minimize their sunk costs.4 Considering the uncertainty surrounding the amount and timing of the return ﬂows, producers are logically reluctant to start a high range remanufacturing program. Two reasons contribute to maintain this uncertainty at a critical level: The absence of past data documenting the return behavior and the demand for used items, as well as the product selling strategy which induces products return too late to be remanufactured.5 Reverse ﬂow management is a recent preoccupation for most of the ﬁrms confronted with either more restrictive legislation (e.g. car industry, mobile phones) or competitive pressure (e.g. copiers, truck tires). As a result, little attention has been given to the acquisition of past data so that manufacturers are not able to provide their corporate planning with reliable information. Publications focusing on reverse ﬂow forecasting (Kelle and Silver, 1989, Toktay et al., 2004) build on Goh’s approach, which addressed returnable containers (Goh and Varaprasad, 4 5

As e.g. in the car industry where BMW and Mercedes cooperate with their retailers and local recyclers (see Kircher, 2004, Schulz, 2004). Kostecki (1998) investigates the reasons for the late returns. His investigations will be summarized in chapter 7.

78

5 Strategic Network Planning in Closed-Loop Supply Chains

1986) or equivalent items such as one-way cameras (Toktay et al., 2004). Nevertheless, one of the limitations of this forecasting method is its lack of reliability when past data is not available. On the other hand, Gohbased algorithms, unlike exponential smoothing or time series, respect the correlation between sales ﬁgures and return ﬂows. Fuzzy logic approaches, which combine expert knowledge and experience gained by other industries, may be a way to cope with information lacks and to get reliable data (Marx-Gomez, 2001). For end-of-life product returns, the return distribution can be estimated with help of a normal or a gamma distribution as ﬁgured in ﬁg. 5.6 (see Kriebel, 2004 for the computer industry and Kircher, 2004 for passenger cars).

Fig. 5.6. Generic return distribution

By basing their business on product selling, OEMs complicate product return management. The sale of a product implies the transfer of the whole property rights to the buyer and in particular the right to resell its good without referring to the OEM.6 The ﬁrst buyer can try to extend the product life or to resell the product as long as its value is positive. However, the timing of the reverse ﬂows may be in both cases too late, furthermore an OEM has then no direct means to inﬂuence product returns in his favor. A second core ﬂow, parallel to reverse ﬂows occurring at the product end-of-life, supplies remanufacturing activities during the utilization phase. Unlike product returns, single core returns are more driven by functional than psychological obsolescence, i.e. part returns are caused by a service failure. Lund (1977) identiﬁes three phases in the working life of a core. Each phase presents a 6

See Furubotn and Pejovitch (1972) and Ebers and Gotsch (1999) for a comprehensive explanation of the property rights theory.

5.2 A Generic Strategic Network Planning Model

79

single failure reason with respect to the service time. Manufacturing errors and material failures explain early life defects; then, after a long period of lowest defect incidence during which the defects are almost always caused by unusual stress, wearing makes the failure rate rise again (see ﬁg. 5.7). Ihde et al. (1999) argue that with the increasing quality eﬀorts in manufacturing processes, early failures tend to disappear, although the part’s reliability still diminishes over time. Under this quality aspect, the attractiveness of a used part is maximal when remanufacturers can ensure that this core will be returned before the wear out phase begins, thus assuring at least the same service quality as new parts.

Fig. 5.7. Wearing pattern of durable goods, Lund (1977)

Reintegration Potential Taking used products back is a prerequisite for remanufacturing but is not decisive. Technical factors also strongly inﬂuence the remanufacturing probability of a core. We identify hereby three constraints that tend to reduce the reuse opportunities: • Technological evolution. Functional obsolescence, as quoted in Bellmann (1990) and Packard (1960) may occur when a returned core does not ﬁt the requirements for being remanufactured, despite its working state. This may be due to problems regarding the functionality of an item, its performance (e.g. processing speed) or the absence of adapted interfaces.

80

5 Strategic Network Planning in Closed-Loop Supply Chains

• Lacking focus on design for recycling. Design for Recycling (DFR)7 results from the necessity of improving the proﬁtability of the disassembly process. Navin-Chadra (1994) shows the impact of design decisions on the eﬀectivity of disassembly processes and highlights the role of design guidelines in order to help avoiding core damage during the recovery process with the help of removable connections. Furthermore, a modular product structure helps reduce the process complexity faced by the disassembler and thereby contributes to maintain disassembly costs sensibly lower than purchasing costs for new parts (Krikke et al., 2004). • Wearing. A third hurdle on the way to reintegration is the wearing state of the parts, also called qualitative obsolescence. Due to wearing during the utilization phase, the parts’ performance may decrease over time, once reaching a point where a reintegration appears neither economically justiﬁable nor defendable in view of security concerns. In the automotive industry for instance, OEMs refuse to use remanufactured break systems in their ﬁrst-equipment channel. Despite these diﬃculties, it is important to notice that some parts are not subject to strong wearing and, as a consequence, ought to last much longer than their planned life-cycle. As an example, the service time of a processor is much longer than the life-cycle of a computer in which it has been integrated ﬁrst. The reintegration of reclaimed components is not only limited by the disassembly eﬀorts but also depends on the technological evolutions of a product. Whereas recovery models, mostly focusing on materials recycling, assume that a fraction of the reverse ﬂows can be reintroduced in the supply chain, little attention has been given to the remanufacturing opportunities of a distinct component. In order to develop a practical model, technological evolutions complicating or disabling the remanufacturing option of a part should be taken into account. This model extension enables the OEM to estimate the impacts of a product upgrade on inbound ﬂows. An isolated improvement of a given key factor increases the potential throughput of the following generic processes. A tire manufacturer is able to facilitate the reuse of car tire casings by strengthening the sidewalls (see chapter 6). However, the question raises whether a single adjustment of a key factor makes the utilization of the whole remanufacturing potential reachable. In fact, the answer depends on the structure of the input ﬂows which is ﬁnally coupled with the key factors 7

We refer to the technical report VDI 2243 (VDI, 2002) concerning recyclingoriented product development for further information.

5.2 A Generic Strategic Network Planning Model

81

Fig. 5.8. Impact of wearing and technological evolution on the reintegration probability

settings for the processes located upward in the supply chain. Experience shows that these input ﬂows represent a bottleneck that cannot be removed with the help of a local key factor change downwards. For this reason, a modiﬁcation of a key factor should take into consideration the whole parameter setting, beginning with the customer acceptance for used items and the ﬁnancial impacts of such modiﬁcations. Finally, as long as a strategically sound and organizationally coordinated asset recovery program has not been set (see chapter 2), the OEM’s ability to signiﬁcantly inﬂuence part reuse will remain low. 5.2.2 Optimization Model The following linear programming model encompasses the remanufacturing key factors as well as monetary parameters to gain ﬁrst insights into the proﬁtability of remanufacturing. The choice of linear programming as support for our investigations, instead of simulation or agentbased systems, has one main advantage: The mathematical model returns the optimal sales mix with respect to the parametric constraints. The results are therefore directly appealing manufacturers because of their optimality. Moreover, the lack of data considering customers’ behavior and the necessity of delivering quantiﬁable results for OEMs makes the use of agent-based systems diﬃcult to apply in a strategic planning context. We consider an OEM acting as a monopolist on his market whose goal is to obtain a proﬁt-maximizing product mix that also includes remanufactured products. This model is kept in its

82

5 Strategic Network Planning in Closed-Loop Supply Chains

generic form in order to enable comparisons between case studies. Its purpose is to cluster the products into weak and strong remanufacturing candidates. Obviously, strong remanufacturing candidates ought to improve an OEM’s overall ﬁnancial performance. Figure 5.9 illustrates the model’s ﬂows and their interactions.

Fig. 5.9. Strategic network planning model: Overview

Index Sets c : set p : set q : set t, t’ : set

of of of of

components products quality levels time periods

5.2 A Generic Strategic Network Planning Model

83

Variables βp,q,t Xp,t Nc,p,t Ac,p,t ,t RFc,t ,t Rc,t ,t Ic,t ,t Oc,t ,t

: market share of product p in segment q during period t : ﬂow of produced good p in period t : ﬂow of new components c in period t used to make item p : ﬂow of reused components c in product p of period t’ into period t (remanufacturing or spare part) : reverse ﬂow of component c of period t’ into period t : reintegrated ﬂow during period t of component c introduced in period t’ : inventory of component c from period t’ stored in period t : ﬂow of components c, aged t-t’, leaving the system

Monetary Parameters σp,t κPp,t κN c,p,t κRF c,t ,t

κR c,t ,t κO c,t ,t κIc,t ,t

: retail price of product p in period t : variable cost per unit of manufacturing product p in period t : cost of a new component c for product p during period t : reclaiming costs for component c issued from end-of-life product return ﬂows after t-t’ periods (acquisition, transportation, disassembly) : recovery costs for component c from period t’ in period t (testing, cleaning, repair) : recycling cost for component c sold in period t’ and processed in period t : holding cost for part c from period t, aged t-t’ periods

Technical Parameters demq,t αc,p max βp,q ρp,t ,t r θc,t ,t

δc,p,t ,t

: demand of service units from quality q during period t : sourcing matrix, number of components c in product p : upper bound for market share of product p in segment q : return probability of product p sold in period t’ during period t : working probability of part c introduced in period t’ and returned after t-t’ periods of use (no worn part) : binary indicator allowing the reuse option of component c in product p after t-t’ periods of use

84

5 Strategic Network Planning in Closed-Loop Supply Chains

Objective Function M ax !

CMt

t

CMt =

p

−

σp,t · Xp,t −

κPp,t · Xp,t −

p

κRF c,t ,t

· RFc,t ,t −

c,t

−

κN c,p,t · Nc,p,t

c,p

κR c,t ,t

· Rc,t ,t −

c,t

κO c,t · Oc,t ,t

c,t

κIc,t ,t · Ic,t ,t

(5.1)

c,t

The producer’s goal is to maximize his contribution margin CMt with a given selling price σp,t for a good p in period t on the one hand, and the costs induced by production, components’ procurement, reclaiming, recovery and disposal on the other hand. Although remanufactured products are generally cheaper than new ones, the marginal income per remanufactured unit might be higher thanks to savings on the cost side. The diversity of reverse logistics processes makes it diﬃcult to encompass each speciﬁcity of a supply chain; therefore, we have limited R RF the cost parameters to two generic data sets: κRF c,t ,t and κc,t ,t . κc,t ,t encompasses all process costs required to make a component available, i.e. acquisition costs, transportation and disassembly. κR c,t ,t comprises the postponed steps which are necessary to ensure a component’s reintegration into the production or utilization process (sorting, testing, cleaning and upgrading). This aggregation level allows a ﬂexible adaptation to the speciﬁcity of every case study. Constraints Xp,t =

demq,t · βp,q,t

∀p, t

(5.2)

q

βp,q,t = 1

∀q, t

(5.3)

p max βp,q,t ≤ βp,q

∀p, q

(5.4)

Equation 5.2 expresses the matching of customers’ needs in a given segment with the supplied products (e.g. for cars: The demand for superior quality with up-to-date technology matches with a new premium car). Variable βp,q,t also describes a product’s market share in a speciﬁc segment and period. Since βp,q,t is variable, equation 5.3 ensures

5.2 A Generic Strategic Network Planning Model

85

that the sum of all market shares in a segment q reaches 100% whereas max for a market share. equation 5.4 sets a potential upper-bound βp,q This restriction especially applies to remanufactured items which face psychological obsolescence. Depending on the market structure, a product may compete in several market segments: Remanufactured toner cartridges, for instance, partly cannibalize the demand for new OEM cartridges. RFc,t ,t =

∀c, t , t

αc,p · Xp,t · ρp,t ,t

(5.5)

p

RFc,t ,t = Rc,t ,t + Oc,t ,t

∀c, t , t

(5.6)

r Rc,t ,t ≤ θc,t ,t · RFc,t ,t

∀c, t , t

(5.7)

Equation 5.5 computes the product returns with help of the reverse ﬂow distribution ρp,t ,t . Equation 5.6 expresses the fact that returned parts RFc,t ,t are either disposed of (i.e. landﬁlled or recycled) or processed for reuse (Rc,t ,t ). In equation 5.7, the reintegrated cores Rc,t ,t r are limited by parameter θc,t ,t which takes into account that some of the reclaimed cores, depending on their origin and age, may be too worn or damaged to be reintegrated. Ic,t ,t−1 + Rc,t ,t = Ic,t ,t +

Ac,p,t ,t

∀c, t , t

(5.8)

p

Ac,p,t ,t ≤ δc,p,t ,t · (Ic,t ,t−1 + Rc,t ,t ) αc,p · Xp,t = Nc,p,t +

Ac,p,t ,t

∀c, p, t , t (5.9)

∀c, p, t

(5.10)

t ≤t

Equation 5.8 allows the storage of parts to be reused in later periods. Equation 5.9 sets an upper-bound for component reintegration, whereby ﬂow Ac,p,t ,t comprises only parts that are technologically reusable. The key factor and binary parameter δc,p,t ,t indicates which sourcing combinations are allowed, i.e. in which product a given component c can be reused after t-t’ periods. Finally, equation 5.10 computes the overall part demand which is either accommodated by new parts (Nc,p,t ) or by reclaimed ones (Ac,p,t ,t ). 5.2.3 An Inter-Generational Compatibility Extension The generic model presented in the previous section has been customized for a project with a ﬁrst tier supplier of the automotive in-

86

5 Strategic Network Planning in Closed-Loop Supply Chains

dustry. The product lines of the car and truck manufacturers have an average market life cycle of ﬁve years. However, the products are still used during at least ten or ﬁfteen years after introduction (Ihde et al., 1999, Hauser and R¨ ottchen, 1995). Hence, automotive suppliers have to guarantee that exchange parts will be available after the initial product line has run out. In order to minimize the production costs for inactivated product lines, OEMs might reclaim the broken parts and remanufacture them in order to avoid the setup costs induced by the production of a new lot. Inventory management strategies for discontinued spare parts have been investigated by Spengler and Schr¨ oter (2003, 2005) as well as Teunter and Fortuin (1999). To some extent, the publications mentioned provide promising results concerning the savings potential of spare parts. Although the supplier’s products are improved stepwise during the life-cycle, these have common parts that could be reused into newer product lines. The objective of the automotive supplier was to investigate the proﬁtability of remanufacturing with respect to the potential compatibility of the parts within product generations. The required components may be either procured as new Nc,p,t or reclaimed internally A∗c ,c,p,t ,t (eq. 5.11). In this case, a binary intergenerational compatibility parameter γc ,c,p has been introduced to allow the use of components c from older generations into newer products p (eq. 5.12). M represents a very high number. Finally, the inventory constraint is updated in equation 5.13 to take the intergenerational usage of parts accurately into account. αc,p · Xp,t = Nc,p,t +

A∗c ,c,p,t ,t

∀c , c, t

(5.11)

c ,t

A∗c ,c,p,t ,t ≤ M · γc ,c,p

∀c , c, p, t , t

Ic,t ,t = Ic,t ,t−1 + Rc,t ,t −

A∗c,c ,p,t ,t

(5.12) ∀c, t , t

(5.13)

c ,p

Table 5.1 shows an extract of the compatibility matrix where the initial part c can be reused as part c into product p. old gen depicts an item of a previous generation while new gen depicts the product currently in use. The on-stock module K may replace T K in the previous generation’s motors but not in the current ones (γK,T K,newgen = 0). While module K can replace module T K in the products from the older generations old gen 2, K cannot replace module T K in the newer products (i.e. γK,T K,newgen1 = 0). A similar case is described by Driesch et al. (2005) concerning the engine refurbishing activity of Mer-

5.3 Extensions to the Generic Strategic Planning Model

87

Table 5.1. Intergenerational parts’ compatibility between products γc ,c,p Component (c’) Target use (c) Product (p) Value K TK K TK

TK TK TK TK

old gen 2 old gen 2 new gen 1 new gen 1

1 1 0 1

cedes Benz (MB). MB guarantees a spare parts supply for twenty years after product introduction. Within this lap of time, the design of the motors has been steadily improved so that the reintegration scope of the returned engine components is technologically limited but feasible. The authors estimate the amount of motors processed to 50,000 units per year.

5.3 Extensions to the Generic Strategic Planning Model In the following, we will present two model extensions relying upon the strategic remanufacturing planning model developed previously. While the ﬁrst extension will provide support in deciding whether to reintegrate recycled materials in addition to the used products and components, the second model extension will put emphasis on the location planning of resource recovery centers. 5.3.1 Recovery Path Determination An internal remanufacturing program and a product stewardship law have in common that they force the manufacturer to internalize the costs of material recycling. To minimize the resulting ﬁnancial burden, some OEMs concerned have started developing products following the design for recycling guidelines (see chapter 2). Given this, the question turns up why manufacturers should not reclaim the beneﬁts of such development eﬀorts on their own. Assuming that the market prices for virgin and recycled materials would keep increasing continuously during the next twenty years,8 the setup of a proprietary supply chain might protect a manufacturer from future material price increases. Unlike remanufacturing, the realization of recycling programs is coupled with the development of an infrastructure with signiﬁcant investments 8

Consequently to the increased demand from China and India.

88

5 Strategic Network Planning in Closed-Loop Supply Chains

and operating ﬁxed costs (Hansen, 1999). Hence, the attractiveness of recycling investments should be assessed with help of a discounted cash ﬂow analysis. Integration of Recycling Figure 5.10, which depicts the additional ﬂows of the strategic planning model (thick lines) and the binary activity indicators (spots), illustrates the mutual exclusivity of remanufacturing and recycling ﬂows. When both recovery paths are ﬁnancially attractive for the OEM, an internal competition might occur which can be easily solved by analyzing the optimization results of the strategic model. While component reintegration underlies psychological obsolescence, virgin and recovered materials fulﬁll the same purpose and comply with equivalent quality requirements (except for plastics). Therefore, we will assume in the folN for lowing model extension that the two inbound material ﬂows (Mm,t R virgin materials m, Mm,t for recovered ones) are identical with exception of their price.

Fig. 5.10. Recovery path selection: Extended model ﬂows

In order to take the evolution of the material composition over time, the period index t is appended to the material sourcing coeﬃcient ηm,c,t (eq. 5.14). The ﬂow of material leaving the system to be landﬁlled or incinerated is now Dm,t and accounts for the component surplus Oc,t ,t R (eq. 5.15). Current technolominus the ﬂow of recycled materials Mm,t gies do not automatically enable a recycler to reclaim the totality of a product’s content, however, the recoverable material fraction steadily increases after the introduction of design for recycling guidelines dur-

5.3 Extensions to the Generic Strategic Planning Model

89

ing the product development phase.9 In order to take this eﬀect into account during the planning process, a factor µm,t is introduced and expresses the reintegration upper-bound for a material m processed in year t (eq. 5.16). µ is expected to increase over time to reach 100% on the long term.

N R Nc,p,t · ηm,c,t = Mm,t + Mm,t

∀m, t

(5.14)

R Oc,t ,t · ηm,c,t = Mm,t + Dm,t

∀m, t

(5.15)

c,p

c,t

R Mm,t ≤

Oc,t ,t · ηm,c,t · µm,t

∀m, t

(5.16)

c,t

Investments and Depreciations While the great majority of the strategic closed-loop planning models takes only variable and ﬁxed costs into account, the investments in recovery assets are important to separate from the ﬁxed costs in order to catch the corporate taxation eﬀects of depreciations (Goetschalckx and Fleischmann, 2005). Fleischmann et al. (2005) as well as Ferber (2005) propose to model the regular asset depreciations with help of binary variables (in our case ΩtM for recycling factories and ΩtR for remanufacturing lines) and a depreciation schedule dprt ,t . Figure 5.11 illustrates the relation between Ω and the depreciation scheme. The initial investments amount for inv M for a recycling plant and inv R a product recovery center. Once the investment has been performed (i.e. Ω = 1), equation 5.17 computes the annual depreciation amount DP Rt . DP Rt =

t

+

M inv M · (ΩtM − Ωt −1 ) · dprt ,t

t

inv R · (ΩtR − ΩtR −1 ) · dprt ,t

(5.17)

This modeling approach requires to forbid the closedown of an asset when it has been opened in previous periods.10 For this purpose, the 9 10

See for instance Podratzki (2003) who documents the recyclability improvements by Fujitsu-Siemens-Computers from 1990 to 2002. If not, the depreciations could become negative if a recently opened plant would be closed.

90

5 Strategic Network Planning in Closed-Loop Supply Chains

R M following logical constraints ΩtR ≥ Ωt−1 ∀t and ΩtM ≥ Ωt−1 ∀t are appended to the strategic model. Due to the existence of a time lap b between the investment start and the beginning of the activity, this time lap is deducted in the capacity constraints (eq. 5.18 and 5.19).

R M Mm,t ≤ capM · Ωt−b

∀t

(5.18)

∀t

(5.19)

m

R Rc,t ,t ≤ capR · Ωt−b

c,t

The depreciation scheme is activated only in the period when the binary investment variable (here Ω) switches from 0 to 1. The situation at the beginning of the planning period (by which no recovery asset M = Ω R = 0). exists) has to be modeled explicitly (Ωt=0 t=0

Fig. 5.11. Depreciation modelling framework (modiﬁed from Ferber, 2005)

Objective Function With respect to the ﬁnancial impact of recycling related investments, the integration of the ﬁnancial dimension in the context of strategic network planning is required to ﬁt with managers usual decision tools such as discounted cash ﬂow analysis (Copeland and Antikarov, 2003). This integration has been extensively investigated by Goetschalckx and Fleischmann (2005). On the basis of the authors’ framework, the objective function of the strategic closed-loop planning model is enhanced

5.3 Extensions to the Generic Strategic Planning Model

91

in equation 5.20 and now maximizes the discounted cash-ﬂow after tax DCF of the potential closed-loop supply chain. The parameter wacc depicts the weighted average cost of capital while taxt depicts the corporate tax rate of the manufacturer in period t.

max ! DCF =

EBITt · (1 − taxt ) + DP Rt t

(1 + wacc)t

(5.20)

The discounted cash-ﬂow DCF is an extension of the initial contribution margin CMt from which the variable material κM and landﬁlling costs κD are deducted. The assets set up generate a periodical overhead of κFt M for a recycling plant and of κFt R for a reuse / remanufacturing plant. Eventually, the annual asset depreciations are removed to obtain the earnings before tax EBITt (see eq. 5.21). MR N R D N EBITt = CMt − κM m,t · Mm,t − κm,t · Mm,t − κm,t · Dm,t

−κFt M · ΩtM − κFt R · ΩtR − DP Rt

(5.21)

To compute the net present value of the supply chain, one needs to integrate the investments performed during the optimization to the discounted cash ﬂow. An alternative objective function would then be max ! N P V with N P V = DCF − IN V . The investment expenses are detailled in equation 5.22.

IN V =

M R + inv R · ΩtR − Ωt−1 inv M · ΩtM − Ωt−1 t

(1 + wacc)t

(5.22)

5.3.2 Location of Recovery Centers The previous extension provides means to assess closed-loop supply chains on the basis of discounted cash-ﬂow approaches, nevertheless, the plant location issues mentioned in the literature overview are still missing. Therefore, in order to decide which locations are optimal for the setup of asset recovery activities, we extend the core model to the the geographic dimension z depicting the markets in which the company either manufactures or sells his products. We assume in this case that the OEM might only have one disassembly and one recycling plant in each market investigated. The previous waste ﬂow Oc,t ,t is now broken down into three types of waste ﬂows documenting the product state at

92

5 Strategic Network Planning in Closed-Loop Supply Chains

P C the point of disposal: Complete product (Dz,p,t ,t ), component (Dz,c,t ,t ) M ). The recycling costs are allocated to the quantity or material (Dz,m,t C processed Oz,c,t ,t instead of being assigned to the reintegrable outcome R Mz,m,t in order to allocate the material reintegration costs accurately. The production ﬂow Xz,z ,p,t now represents the quantity of item p manufactured in market z and dispatched to market z in period t. The updated network ﬂows are highlighted in ﬁgure 5.12.

Fig. 5.12. Strategic closed-loop location planning

Optimization Model The market segmentation constraints are subject to minor changes after the introduction of speciﬁc markets. The production quantity dispatched to a market z matches with the demand of this market demz,q,t max (see eq. 5.23, 5.24 and with respect to the local market settings βz,p,q 5.25).

5.3 Extensions to the Generic Strategic Planning Model

demz,q,t · βz,p,q,t

Xz ,z,p,t =

z

∀z, p, t

93

(5.23)

q

βz,p,q,t = 1 ∀z, q, t

(5.24)

max βz,p,q,t ≤ βz,p,q

(5.25)

p

∀z, p, q, t

The product returns in market z are determined by the local return rate ρz,p,t ,t and the quantities initially sold in period t on this market Xz ,z,p,t (eq. 5.26). According to equation 5.27, market z’s returns are forwarded to market z for processing (Rz,z ,p,t ,t ) or disposed of locally in a complete state. RFz,p,t ,t =

Xz ,z,p,t · ρz,p,t ,t

∀z, p, t , t

(5.26)

z

RFz,p,t ,t =

P Rz,z ,p,t ,t + Dz,p,t ,t

∀z, p, t , t

(5.27)

z

Equation 5.28 ensures that the components forwarded to market z and dismantled are either reintegrated as components (Az,c,p,t ,t ), disC posed of (Dz,c,t ,t ) or forwarded to a recycling facility in the same marC ket (Oz,c,t ,t ).

Rz ,z,p,t ,t · αc,p =

z ,p

C C Az,c,p,t ,t + Dz,c,t ,t + Oz,c,t ,t

∀z, c, t , t (5.28)

p

Equation 5.29 sets an upper-bound for the component reintegration R ) and technical obsowith respect to the functional obsolescence (θc,t ,t lescence (δc,p,t ,t ). The removal of the inventory constraint enables the merging of both key parameters into a single constraint instead of two.

A

z,c,p,t ,t

≤

R

z ,z,p ,t ,t

·α

c,p

R · θc,t ,t · δc,p,t ,t

∀z, c, p, t , t (5.29)

z ,p

We state that the disassembly takes place in the same market as the remanufacturing process (eq. 5.30) as the recovered parts reclaimed in market z are also reused in z. New parts Nz,c,p,t are procured when component reintegration is neither technologically nor functionally possible, or when the reclaimed parts are more expensive to recover than the procurement of new ones.

94

5 Strategic Network Planning in Closed-Loop Supply Chains

αc,p · Xz,z ,p,t = Nz,c,p,t +

z

∀z, c, p, t

Az,c,p,t ,t

(5.30)

t ≤t

In a similar fashion as in the previous section, the materials m required for the manufacturing of the components c in market z are N R or recycled Mz,m,t . The component composition rate either new Mz,m,t ηm,c,t remains the same (eq. 5.31). The components rejected for remanC ufacturing Oz,c,t ,t provide a source of materials which depends on a component’s initial composition in period t . The materials inapt for reintegration Dz,m,t are landﬁlled in the market of processing (eq. 5.32). Equation 5.33 sets an upper-bound for the recycling pipeline with respect to a component’s initial composition ηm,c,t and the current material reintegration rate µm,t .

N R Nz,c,p,t · ηm,c,t = Mz,m,t + Mz,m,t

∀z, m, t

(5.31)

C R M Oz,c,t ,t · ηm,c,t = Mz,m,t + Dz,m,t

∀z, m, t

(5.32)

c,p

c,t

R Mz,m,t ≤

C Oz,c,t ,t · ηm,c,t · µm,t

∀z, m, t (5.33)

c,t

The plant location decision relies on a binary capacity constraint similar to the previous section but extended to the geographic situation. Equation 5.34 allows the disassembly of a product p in a recovery R is plant located in market z when the binary activity indicator Ωz,t positive. We assume in equation 5.36 that an asset recovery plant cannot be closed down after its setup. The same constraints apply to the recycling process where the quantity of material processed in a market C Oz,c,t cannot surpass the capacity capM ,t · ηm,c,t z,t in the same c,m,t

market z.

R Rz ,z,p,t ,t ≤ capR · Ωz,t−b

∀z, t

(5.34)

∀z, t

(5.35)

z ,p,t C M M Oz,c,t · Ωz,t−b ,t · ηm,c,t ≤ cap

c,m,t R R ≥ Ωz,t−1 Ωz,t M Ωz,t

≥

M Ωz,t−1

∀z, t

(5.36)

∀z, t

(5.37)

5.3 Extensions to the Generic Strategic Planning Model

95

Extended Objective Function The periodic corporate earnings before tax and amortization function EBITt encompasses the sum of all regional subsidiaries’ earnings before sub which amount for: tax EBITz,t sub EBITz,t =

z ,p

−

F σz ,p,t − κPz,p,t − κTz,z ,p,t · Xz,z ,p,t −

κRF z,p,t · RFz,p,t ,t −

p,t

−

z ,p,t

N κM z,m,t

·

N Mz,m,t

−

−

P P κO z,p,t · Dz,p,t −

p

−κFz,tR

c,p

·

−

κFz,tM

·

M Ωz,t

R∗ κM z,c,t · Oz,c,t ,t

C C κO z,c,t · Dz,c,t −

c R Ωz,t

κN z,c,t · Nz,c,p,t

κTz R,z,p,t + κR z,p,t · Rz ,z,p,t ,t

c,t

m

M M κO z,m,t · Dz,m,t

m

− DP Rz,t

(5.38)

The objective function is extended to encompass explicitly the costs TR F related to the transport of forward κTz,z ,p,t and reverse ﬂows κz ,z,p,t . Parameter κRF z,p,t now only depicts the procurement costs necessary to get access to the used product. The OEM has the possibility to exclude uninteresting cores in the country where the returns take place in order to avoid transporting waste from one country to the other. Since the product would still be in a complete state, the disposal costs for this ﬂow R P amount for κO z,p,t ,t . The disassembly costs κz,p,t are aggregated at the product level. Implicitly, we assume complete disassembly before reintegration. The component surpluses can be shredded but might also C leave the closed-loop without processing at a cost of κO z,c,t representing the market value or disposal fee for recovered parts. The variable recycling costs for reintegrated materials are computed at the compoR and Dz,m,t are treated. Thus, the recycling nent level since both Mz,m,t MR∗ costs are modiﬁed to κz,c,t and assigned to the component in which the materials are embedded. The landﬁlling costs for unused reclaimed maM terials yield κO z,m,t . We notice that the reclaiming costs are accounted to the local subsidiaries in which the returns take place whereas component reintegration is allocated to the subsidiary refurbishing the cores. Although this cost accounting scheme is neutral with respect to the optimization results, this internal accounting policy might provide low incentives for local subsidiaries to reclaim the cores when no remanufacturing center exists in the same market. We refer to chapter 3 for a broader overview on transfer price setting policies able to align incen-

96

5 Strategic Network Planning in Closed-Loop Supply Chains

tives between sales divisions. The local depreciations DP Rz,t depend on the recovery assets set up in market z during the optimization. DP Rz,t =

t

+

t

R R R Ωz,t · dprt ,t − Ωz,t −1 · inv

M M M Ωz,t · dprt ,t − Ωz,t −1 · inv

(5.39)

To demonstrate the practical relevance of the modular strategic planning model presented above, this decision-making framework will be applied in the following to two case studies.

Part II

Closed-Loop Supply Chains: Case Studies

99

Theoretical concepts need to be validated in practice in order to convince managers to make use of them instead of only relying on presumed facts and beliefs. Therefore, the objective of part 2 is twofold: First, we aim at demonstrating the relevance of the decision-making framework developed in part 1. Furthermore, the application of the strategic network planning models to real life case studies will enable us to catch further aspects of closed-loop supply chains’ business dynamics which are diﬃcult to catch a priori with theoretical concepts. Chapter 6 is dedicated to the tire industry. With help of the generic network planning model of chapter 5, we will investigate the remanufacturing potential of car and truck tires. On behalf of the optimization results, we will propose an explanation for the remanufacturing rates discrepancy between truck and passenger car tires. This observation will then be generalized to all complex products and should enable decision-makers to focus their attention on potentially remanufacturable products. Chapter 7 addresses the computer industry. Considering the environmental burden of computers, the European legislation enforces recycling targets. In this context, we will assess to what extent remanufacturing and recycling are proﬁtable for computer manufacturers and apply the discounted cash-ﬂow perspective drafted in the previous chapter. We will observe that closed-loop supply chains are a proﬁtable issue despite the initial investments and will quantify the incentive role of the current take-back legislation (see chapter 3 for preliminary insights).

6 Tire Industry

6.1 Introduction Over 600,000 tons of scrap tires are yearly disposed of in Germany (see IFEU, 1999). Since the introduction of the German Recycling Law (KrW/AbfG) in 1994, particular attention has been paid on the disposal phase of end-of-life tires and its ecological impact has been object of recent studies. Although the German Recycling Law presents a framework that has not been speciﬁcally applied to tire manufacturers by now, the consequences of the life cycle assessments published (IFEU, 1999, Ayres et al., 1997) should not be underestimated as they point out the current resource wastage induced by the current disposal processes. To improve the sustainability of the tire supply chain, the German Recycling Law could be applied in the near future to the tire industry and enforced by setting recycling targets similar to the automobile industry. In order to understand the challenges appearing during the recycling process, a brief overview of a tire’s material composition is necessary. As ﬁgure 6.1 shows, about 45% of a tire consists of rubber compounds and 25% of carbon black. This composition makes tires inappropriate for landﬁll because of their inﬂammability and tendency to ﬁnd their way back up to the surface after having been buried. However, this dumping problematic is not critical since, with a caloriﬁc value of 31,000 kJ/kg, used tires are an appreciated input for incineration processes. In fact, as shown is ﬁgure 6.2, approx. 50% of scrap tires are used as a supply for the German Portland cement kilns. The advantages of this approach are manifold: Rubber combustion generates heat, moreover, incineration residues serve as an input into the cement production without leading to quality loss. In fact, the reclaimed

102

6 Tire Industry

Fig. 6.1. Average material composition of a tire (IFEU, 1999)

steel can substitute iron which is usually required in Portland cement. Dedicated thermoelectric plants present another energy reclaiming alternative (see Ferrer, 1997 and IFEU, 1999). In both cases, used tires lead to signiﬁcant resource savings as tires either replace raw materials such as coal (Portland cement) or avoid generation of further KWh through oil, gas or nuclear power. Considering raw material prices, this substitution also leads to signiﬁcant ﬁnancial savings for plants that apply this recycling approach. Nevertheless, energy reclaiming, despite its resource savings, is not sustainable in the long run as it utilizes only non-renewable materials (with exception of natural rubber). Unlike incineration, material recycling represents an opportunity to keep resources in the supply chain. The composition of a tire indicates that the processing output mostly consists of granulated rubber. Nevertheless, the limits of recycling become clear when considering the physical properties of rubber as the vulcanization process causes a cross-linking of rubber molecules which is not reversible. It is therefore not possible to manage a further molecular bond with other rubber mixes. As a result, a satisfactory chemical cohesion cannot be attained when using reclaimed rubber, therewith limiting its use to tire parts that do not underlie above-average stress (quality requirements exclude reclaimed rubber from high-end tires) or in which virgin rubber does not overperform (connections of rubber to metal). Because of these limitations, the rubber reintegration scope will remain low. Investigations mandated by the German Environmental Administration, among other by IFEU, estimate an upper-bound reaching up to a 1.5% of a tire’s

6.1 Introduction

103

Fig. 6.2. Allocation of scrap tires to recovery alternatives (IFEU, 1999)

weight. Experts do not expect this ﬁgure to increase in the near future (BRV, 2001). Considering the material limits of rubber recycling, downcycling appears unavoidable, i.e. rubber can only be processed into diﬀerent, lower-grade applications such as a bitumen additive in order to extend the service life of the road surface.1 However, two objections can be raised against downcycling despite its contribution to resource replenishment: The German market for scrap rubber is already saturated and this recycling path has no impact on the initial resource consumption in the tire supply chain. So far, it has been diﬃcult to identify a sustainable way to recover resources from used tires. As a matter of fact, Ferrer (1997), Ayres et al. (1997) and IFEU (1999) instead agree about the overwhelming resource savings reached by tire remanufacturing, also commonly depicted as retreading. In this context, Ayres et al. point out that retreading is the only way to partly retain a tire’s added value which represents about 80% of the production costs. Material recycling or incineration can only recover less than 3% of a tire’s value. A look at the life cycle assessment dealing with tire recycling strengthens these ﬁndings: The production of a retreaded car tire, for instance, consumes only one third of the energy input and one fourth of the material input required for a brand new tire (see table 6.1). 1

The additional elasticity of rubber compounds helps extend the surface’s service life.

104

6 Tire Industry

Table 6.1. Resource consumption for the production of a car tire, see Ferrer (1997) Tire type Energy (in kJ) Materials (in kg) New Retreaded

72,000 19,000

8 1.9

A look at the product structure is necessary to understand the overwhelming ecological merits of retreading. A generic tire is composed of several elements that can be aggregated into two parts: Casing and tread. Due to its continuous contact with the road surface, a tread wears out and tends to become ﬂat, then causing insuﬃcient road adherence and reduced braking performance. Fortunately, tires are returned although the tire casing often presents no signiﬁcant damage. As a matter of fact, tire casings are not considered as wear parts since their wear pattern is not directly linked to the tire’s service life but to the user’s driving style.

Fig. 6.3. Tire structure, Continental (2003)

Roughly speaking, retreading means replacing the worn tread with a new one. To be remanufactured, a tire has to pass several steps beginning with an initial inspection where damaged casings are ﬁltered out. The disassembly process consists of removing the worn tread with help of a buﬃng rasp and thus preparing the casing for the application of a new tread. As a next step, both parts are bonded during the vul-

6.2 Model Implementation

105

canization process. At this point, the remanufacturer faces the choice between two types of bonding techniques: mold cure (with an uncured tread subsequently molded) and precure (with an already cured tread ﬁxed by a special bonding rubber). Despite the evident merits of retreading on an economical as well as on an ecological level, remanufacturing rates are varying signiﬁcantly among tire types (see table 6.2). All published studies so far agree that the remanufacturing potential of truck and airplane tires is already exhausted. At a ﬁrst glance, however, the discrepancy between summer car tires and airplane tires is hard to explain since material composition and the retreading technology applied do not signiﬁcantly diﬀer. This provides an interesting basis for investigating the reasons explaining the current unbalanced situation. Table 6.2. Tire remanufacturing rates, IFEU (1999), Ferrer (1997) Tire type

Remanufacturing rate

Summer car tire Winter car tire Light truck tire Truck tire Airplane tire

1% 10% 17% 50% up to 90%

6.2 Model Implementation The ﬁrst step in order to implement the model is to deﬁne an index set that considers the existing trade-oﬀ between accuracy and complexity. The model should therefore be dimensioned as concise as possible without compromising its realism. The trade-oﬀ can be illustrated by the dimensioning of the component set C in which seven diﬀerent tire parts can be taken out of ﬁg. 6.3. We mentioned that these can be aggregated into a casing2 and a tread. This dimension reduction by two-third has no consequence on the results as retreading processes are only concerned with the separation of casing and tread and never aim at reclaiming the parts embedded into a casing. 2

The casing encompasses carcass, inner-liner, sidewall, belt, apex and bead.

106

6 Tire Industry

We focus our analysis on two generic tire sizes which are widely spread in Germany: P175/65R143 for car tires and 295/80R22.5 for truck tires. The diversity of tire producers can be grouped into four homogeneous clusters: High-end, middle-range (MR), import and retreaded tires. Hence, we limit the model’s product set P to four items (one product for each group). According to our investigations, these four products are positioned into three partly coupled market segments q: Premium, budget and low budget (see BRV4 , 2000). The time frame for the observation encompasses 20 periods in order to respect the whole product life cycle of a tire. The period length thereby encompasses 1/2 year for all tires. 6.2.1 Demand Segmentation The ﬁndings from table 6.3 are twofold. First, the demand for passenger car tires is nearly equivalent for Premium as well as for Budget tires. For psychological reasons, buyers prefer to avoid low budget tires. Unlike car tire buyers, truck tire buyers prefer either Premium or Low budget tires. Deierlein (1988) mentions that carriers procure Premium tires for the driving axle of their trucks, not solely because of security concerns but also because these tires are easier resold for retreading. Low segment tires are primarily used for trailers since these face tire failures to a lesser extent. As it is also reglementary forbidden to put retreaded tires on the front axle, the demand is clearly limited for these tires. max , issued from own estimations and checked for The value of βp,q correctness by the BRV, depicts the tire positioning in a segment q. An upper bound of 1 indicates that a good is clearly positioned in this segment while a value of 0 means the opposite. The small price gaps between segments (see table 6.5) partly explains the porosity between segments. In the following, we observe an OEM covering a representative market with the four products depicted in table 6.4. Due to the highly competitive environment and the existence of established seg max ments, βp,q is close to one but keeps a realistic degree of freedom. p

Hence, an OEM cannot signiﬁcantly modify the demand structure on a short run. 3 4

P: passenger car tire, 175: section (= tire) width in mm, 65: aspect ratio in % (section height / section width), R: radial construction, 14: rim diameter in inches. BRV: Bundesverband Reifenhandel und Vulkaniseur Handwerk e.V. - German tire manufacturer lobby

6.2 Model Implementation

107

Table 6.3. Demand segmentation of tire markets in Germany (BRV, 2000 and 2001) Segment Sum. car tires Win. car tires

Truck tires

Premium

e.g. Continental, Michelin, Goodyear, Bridgestone 39.5% 35.5% 36% Budget e.g. Uniroyal, Dunlop, Vredestein, Pneumant, Kleber 38.5% 39.5% 18% Low budget e.g. Sumho, Rigdom e.g. Rigdom, Barum 22% 25% 46%

Table 6.4. Product market share upper bounds in given segments Product Sum. car tires Win. car tires Truck tires High-end Premium 1 Budget 0.2 Low-budget 0 Middle- Premium 0.1 range Budget 1 Low-budget 0.2 Import Premium 0 Budget 0 Low-budget 0.8 Retread Premium 0 Budget 0 Low-budget 0.1

Premium 1 Budget 0.1 Low-budget 0 Premium 0.1 Budget 1 Low-budget 0.2 Premium 0 Budget 0 Low-budget 0.8 Premium 0 Budget 0 Low-budget 0.5

Premium 1 Budget 0.1 Low-budget 0 Premium 0.1 Budget 1 Low-budget 0.3 Premium 0 Budget 0 Low-budget 0.8 Premium 0 Budget 0.3 Low-budget 0.8

German tire retreaders argue that the present segmentation is more driven by psychological obsolescence than by real quality and reliability concerns. In fact, all tires produced in Germany must pass quality tests to be allowed in a given speed segment. Tires within the same speed class have approximatively the same objective quality, no matter if they have been retreaded or not. Import tires, however, do not underlie these restrictions and can be placed into speed classes without strict reliability control by the German authorities. Furthermore, retreaded tires are produced for a maximal speed of 190 km/h (speed class T) which represents a great hurdle since an important fraction of German

108

6 Tire Industry Table 6.5. Average retail prices in Germany in Euros (ADAC, 2002) Product High-end Middle-range Import Retread

Summer car tires Winter car tires Truck tires 72 50 35 37

80 55 45 37

385 324 250 295

motorways are not regulated by any speed limit. Driver’s lobby agitations in reaction to a potential reduction of motorways’ speed limits have shown that Premium and Budget customers are keen on keeping the freedom to drive fast, even if only occasionally. Few information is publicly available concerning the procurement costs κN . Hence, we can solely assess the resource input to obtain nonbiased data and thereby rely on Ferrer’s (1997) estimations of material and energy costs. On the basis of his calculations, we set for car tires a price of 13.50 Euros for the casing and 3.50 Euros for the tread. Truck tires 295/80R22.5 weigh approximatively 58 kg which represents about six times a car tire’s weight. Due to the higher metallic fraction, material and energy costs rise to 98 Euros for a truck casing and 25 Euros for a corresponding tread. Compared with procurement costs, production costs remain in a modest range: Whereas labor costs for premium and middle-range tires yield 2 Euros and for import tires only 1 Euro, retreaded tires involve the highest labor costs with 4 Euros.5 According to a German retreader, labor costs of truck tire manufacturers are about seven times higher than the costs for car tire production. We apply this rule of thumb to assess manufacturing costs κP of truck tires. 6.2.2 Return Flow Timing and Quantities Both the car and truck tire industry have in common the lack of knowledge about return distribution. Nevertheless, aggregated estimations of BRV document the average age of a returned tire and the total number of tires that are recovered in Germany. Summer car tires are returned after 3.5 years, winter tires only after 4.85 years. As mentioned before, the average lifespan of a truck tire is up to 1.8 years. 5

Especially because of the bonding process for tread and casing (curing) as well as because of labor intensive quality controls processes.

6.2 Model Implementation

109

Reclaiming costs κRF can be divided into disassembly, collection and acquisition costs that are required to guide the cores into a retreader’s pipeline. As we already suggested in the introduction, obtaining tires for remanufacturing is not necessarily easy since other industry sectors, especially cement kilns, are also interested in this reverse ﬂow (see ﬁgure 6.2). Once again, the ascertainment of prices for both car and truck tires is subject to diﬀerent patterns. Whereas disassembly and collection costs are quite proportional to a tire’s size (4 Euros for car tires, 14 Euros for a truck tire), only truck cores are traded on a vivid market where prices may yield up to 50 Euros for an original tire, 30 Euros for a tire retreaded once and 15 Euros for a tire already retreaded twice. Acquisition costs for used car tires will be ignored in the following. 6.2.3 Reintegration Potential A core declines in value as a result of constant wearing: The loss of material consistency and the damages caused to the sidewall reduce the reintegration probability of a casing. The legislator has recognized these limitations and forbids tires aged more than 6 years (car tires) or 8 years (truck tires) for resale. Despite the lack of studies concerning reintegration probability distributions, we deﬁne the following distributions concerning parameter θr for the optimization: r θcasing,t ,t =

⎧ ⎪ ⎪ ⎨ truck ⎪ ⎪ ⎩

car

tires : tires : else :

0.9 − 0.6 −

t−t 20 t−t 20

∀ 0 ≤ t − t ≤ 16 ∀ 0 ≤ t − t ≤ 12

0

We point out that these ﬁgures have been tested for plausibility by the German tire lobby (BRV) and a German tire retreader. In addition to wearing, technological restrictions further hinder the reuse of reclaimed casings. In the tire industry case, OEMs are not allowed to use recovered casings into tires that are labelled as new. Hence, we set the respective parameter δ as follows in order to enable recovered casings’ use into retreaded tires (δ = 1) and avoid their use into the production of original tires (δ = 0):

δc,p,t ,t =

1 0

∀c = casing, p = retread, t , t ≥ t else

Subsequently, reintegration eﬀorts have to be performed to refresh the used carcass. After the removal of the worn tread (buﬃng process),

110

6 Tire Industry

the casing is controlled and repaired for further use. The following processes, tread application, curing (binding of casing and tread) and ﬁnal inspection, are included in the production costs for retreaded tires, thereby explaining the sensibly higher manufacturing costs for remanufactured tires. Reintegration costs are charged at 3 Euros for car tires and 8 Euros for truck tires. In case reverse ﬂows are leaving the system, a fee of 10, respectively 1 Euro is charged to OEMs. This corresponds to the transportation costs to a recycler.

6.3 Optimization Results 6.3.1 Scenario Overview In the following, we will analyze the product portfolio of a representative OEM on the German tire market. We assume that this producer provides three products: A high-end tire dedicated to the premium segment, a middle-range tire primarily addressing the budget customers and a cheap imported tire for the low-budget segment. By now, tire OEMs face no regulation related to the recovery of their products unlike in the automotive or computer industry where manufacturers have to ﬁnance the recovery of their products. Scenario (0) therefore computes an initial situation where the investigated OEM pays neither for reclaiming nor for disposal / recycling of his tires. In consequence to the introduction of a product stewardship law, an OEM’s alternatives might be twofold: Scenario (1) consists of doing nothing, i.e. starting to pay a fee on every sold tire to be released from the obligation of recovering tires on his own. The OEM can alternatively choose an oﬀensive recovery strategy, depicted in scenario (2), which consists of starting a remanufacturing program and, to this purpose, add retreaded tires to his product mix. The goodwill issued from the launch of environmental friendly goods gives the OEM a far more important degree of freedom concerning the obligations of product stewardship laws. Legislation may for instance allow OEMs to ﬁnance the recovery of all returns ex-post as opposed to scenario (1) in which the recovery fee is paid ex-ante on all sold tires. As a consequence, system losses would not present a ﬁnancial burden for the ﬁrm. Thus, scenario (3) computes the case where the OEM decides to sustain system loss, representing in this case 50% of the sold items, with help of a passive return strategy. Scenario (4) depicts a strategy opposite to scenario (3) since the producer obtains the tires back earlier, therefore enhancing the number of potentially reusable casings. This strategy,

6.3 Optimization Results

111

however, also increases the number of reclaimed used tires during the optimization horizon.

Fig. 6.4. Optimization results: Contribution margin per tire sold

Instead of adapting parameter ρ by means of an active return policy, one straightforward solution would be to improve the remanufacturability of a casing ceteri paribus by strengthening its sides, therefore requiring the use of more rubber. This case, depicted in scenario (5), leads to a 10% improvement of factor θ but also implies a material costs increase by 10% that cannot be shifted to the ﬁnal customers. In the long run, a further possibility for the OEM might be to modify the demand structure (parameter β) through marketing actions. For this purpose, we compute in scenario (6) an optimal solution for the case where no market limitations (β max ) are present and obtain ﬁrst advices about the products to promote preferentially. All scenarios have been computed in the same fashion for summer car tires (SCT), winter car tires (WCT) as well as truck tires (TT). The results are presented below (ﬁg. 6.4) in a normalized form: The objective function (equivalent

112

6 Tire Industry

to the overall contribution margin) has been divided by the number of tires sold to reach the average marginal income per tire. The corresponding market shares are documented in table 6.6. Table 6.6. Products’ market share among scenarios in % Scenario

0

1

2

3

4

5

6

SCT High-end 47.2 47.2 47.2 47.2 47.2 47.2 100 Mid.-range 35.2 35.2 35.2 35.2 35.2 35.2 0 Import 17.6 17.6 15.4 15.4 15.4 15.4 0 Retread 0 0 2.2 2.2 2.2 2.2 0 WCT High-end 39.45 39.45 39.45 39.45 39.45 39.45 100 Mid.-range 40.55 40.55 40.55 40.55 40.55 40.55 0 Import 20 20 7.5 8.9 7.5 7.5 0 Retread 0 0 12.5 11.1 12.5 12.5 0 TT High-end 37.8 37.8 37.8 37.8 37.8 37.8 100 Mid.-range 30 30 20 23.4 20 20 0 Import 32.2 32.2 0 0 0 0 0 Retread 0 0 42.2 38.8 42.2 42.2 0

After the introduction of a recycling fee, the marginal income per tire shrinks ceteri paribus by 10% (car tires) and 23% (truck tires). This is due to the fact that reclaiming costs κRF and disposal costs κO encompassed in the fee are added to the production costs while the OEM is not able to provide a counterpart on the sales side. The contribution margin improvements computed in scenario (2) have to be mitigated: Despite the presence of retreaded tires, the marginal income per car tire shows no signiﬁcant improvement for car tires. The component savings on casings are compensated by reintegration costs κR and lower prices for low-budget winter car tires (37 Euros instead of 45). Although summer car retreads are higher priced than import tires, their impact on the overall contribution margin is very limited because of their low market share (2.2%). Truck retreads yield better prices than import ones and also induce important procurement savings through casing reuse. The remarkable increase of the marginal income between scenario (1) and (2) (+42 Euros / truck tire) is mostly explained by a higher average unit price (+12 Euros) coupled with component savings of 38 Euros / unit sold and disposal savings (3.5 Euros / tire).

6.3 Optimization Results

113

Scenario (3), computing the impacts of a 50% returns loss, provides the most promising results in case of product stewardship since the marginal income per unit sold rises by 5.5% (SCT), 6.7% (WCT) and by 10.3% for truck tires. The noticeable marginal increase corresponds to the reclaiming cost savings. Furthermore, the market share of retreaded truck tires shrinks from 42.2% to 38.8%, thus indicating that not enough remanufacturable casings are returned. The remanufacturing bottleneck for truck tires in this scenario shifts from parameter β to parameters θ and δ, since not enough remanufacturable casings are available despite signiﬁcant return rates. However, β remains bottleneck in the car tire’s case. The results of scenario (4) also underline that a preponement of the reverse ﬂows has no positive impact on the contribution margin of the ﬁrm. By obtaining used tires earlier, a ﬁrm reclaims more tires during the planning time horizon. Unfortunately, this does not lead to a remanufacturing rates improvement so that the tire surplus has to be disposed of. For the same reasons, a casing strengthening measure as in scenario (5) does not bring the positive eﬀects expected: The additional costs in the ﬁrst life-cycle cannot be amortized in the second cycle. Contribution margin diﬀerences in scenario (2) thereby have their origin in the component costs’ increase, while the other proﬁt and cost blocks remain stable. In order to benchmark the previous scenarios with the ideal case, we remove the market restrictions in scenario (6). Hence, it appears that an OEM should only sell high-end tires, notwithstanding which kind of tire he produces. While this hypothetic scenario is superior to every other scenario for car tires, scenario (3) provides a better margin income per truck tire sold. As a result, a passive return strategy seems to be the right strategy for a truck tire OEM. The following ﬁgure, representing the remanufacturing pipeline, helps visualizing the inﬂuence of every key factor on the remanufacturing rate of a product and especially the bottleneck factor. In the best case, truck tires yield a market share of 42.2% while winter and summer car tires have an upper-bound of 12.5% and 2.2% respectively. Every measure aiming at increasing the remanufacturability of a product, as in scenarios (3), (4) and (5), will fail as long as the market share upper-bound has been already reached and will only generate a core surplus (see ﬁg. 6.5) that has to be either recycled, incinerated or landﬁlled. In spite of the similar marginal income variations among scenarios, retreaded tires impact diﬀerently the car and the truck tire market. While the introduction of retreaded truck tires in all scenarios is obviously better than doing nothing in case of product stewardship (+29%),

114

6 Tire Industry

Fig. 6.5. Tire remanufacturing pipeline

the results are not clearly in favor of retreaded car tires. Given these results, it is not surprising to see truck tire manufacturers already active in the remanufacturing segment and praising the remanufacturability of their products (see for instance Continental, 2003 or Goodyear, 2003). Since the market potential of car retreads is much more restricted than in the truck tires’ case, the hypothetic marginal income improvement per car tire is, for a majority of OEMs, by now not an integration worth. Despite the uncertainty concerning the available data, the results obtained are in accordance with current market observations. Given the model output, the low budget segment provides signiﬁcant growth potential in an already saturated car tire market. Retreaded car and truck tires turn out to be a competitive alternative, although these face higher recovery costs. The observations of Ayres et al. (1997) related to the potential double-dividends thanks to tire retreading seem to be conﬁrmed by our computations: the added value retained in a casing and reclaimed through remanufacturing reduces the manufacturing costs of a new tire while simultaneously improving the sustainability of the tire industry. Unfortunately, tire retreading has already reached its limits with respect to the fraction of the demand willing to buy ”green tires” eventually. To remove this demand bottleneck, one solution could be to underline the functional nature of a tire and to reduce the role of psycho-sociological factors in the procurement process. Otherwise, further measures aiming at increasing the amount of recoverable cores or measures addressing the ecological beneﬁts of tires are not expected to inﬂuence the attractiveness of remanufactured items. To conclude,

6.3 Optimization Results

115

since a 100% remanufacturing rate is neither technologically feasible nor economically justiﬁed, the goal of an OEM should be to determine an optimal return distribution to avoid core surpluses. The strategic planning model’s ability to integrate all key remanufacturing parameters enables a decision-maker to optimize his reintegration pipeline in this aspect. 6.3.2 Closed-Loop Supply Chains and Functional Goods The scenarios computed in section 6.3 have highlighted the fact that the proﬁtability of tire remanufacturing programs obviously depends on the customers’ acceptance for recovered products. The latter has been taken into account in our model through parameter β. Despite their identical product structure and recovery technologies, truck and car tires necessarily diﬀer when it comes to customers’ acceptance. Based on the observations in the tire industry, we draft a hypothesis arguing that products have functional (mostly of objective nature) as well as psycho-sociological characteristics (mostly of subjective nature). Given this, the success of remanufacturing is coupled with the dominance of functional aspects over psycho-sociological ones in the buying decision. While basic needs are covered through the initial functionality of a good (its core utility), OEMs tend to add and promote a subjective utility in order to improve their positioning and therefore achieve a distinct competitive edge. Car tires thereby provide a good example because the core function of a tire, which is to enable the movement of an object on a road, does not leave any room for diﬀerentiation. In fact, an observation of marketing campaigns for passenger car tires shows that only additional needs, mostly of subjective nature, are evoked. In Germany for instance, car tire manufacturers emphasize on security and design aspects while objective factors such as durability, rolling friction or price are seldom highlighted. This might be surprising as long as all new and retreaded tires sold in Europe have to comply with the same reliability requirements. In contrast to car tires, cost and service aspects are especially pointed out when OEMs promote their truck tires. Deierlein (1988, 1997) notices that tire manufacturers act as ﬂeet managers and tire maintenance specialists instead of limiting themselves to product selling. These services, for example, may be extended by a buy-back program for used tires including discount oﬀers for the procurement of new and retreaded truck tires. Thus, whereas the core and augmented product are predominating factors when selling truck tires, the actual product characteristics are decisive in the car tires’ buying decision.

116

6 Tire Industry

The discrepancy between summer and winter car tires can also be explained in a similar fashion. German customers expect to drive faster in summer than in winter so that they are more willing to buy winter retreads. This is, however, a subjective perception since the average speed on motorways remains constant over the year. Moreover, current speed limitations generally hinder car drivers to drive faster than 130 km/h so that the upper-bound of 170 km/h recommended for lowbudget tires is far from being reached. An incentive problem occurs when OEMs try to push their sales since they have to underscore subjective quality aspects in their marketing strategy. By promoting aspects such as prestige, modernity or security, they automatically tend to discredit remanufactured items. Furthermore, technology-based commodities such as mobile phones or, to a lesser extent, computers face short innovation cycles making previous product generations psychologically (through β) as well as functionally obsolete (through θ and δ). This trend toward shorter product life cycles already depicted by Packard in the 60’s (1960) is obviously contradictory to component reintegration eﬀorts. Unlike private customers, ﬁrms focus primarily on cost/performance ratios to guide their decisions. For this purpose, subjective utility is outweighed by price considerations and objective quality in the procurement phase of industrial commodities. The lack of psycho-sociological factors in the decision process hinders the raising of incentive problems between OEMs and buyers. In this context, we deﬁne functional products as goods for which only objective product characteristics and cost/performance aspects play a role in procurement decisions.

6.3 Optimization Results

117

Product type psycho sociological Mobile phones

Photocopiers

Desktop computers Summer car tires Winter car tires

0%

functional

Truck tires Cash distributors Toner cartridges

50%

Returnable containers

Airplane tires

100%

Remanufacturing rate

Fig. 6.6. Present remanufacturing rate for functional and psycho-sociological goods – (functional goods underlined)

As a consequence, psycho-sociological products are goods for which subjective arguments tend to dominate buying decisions: β max will often present the bottleneck for remanufactured products. High-range remanufacturing programs ought to emerge when no incentive problems exist. Therefore, we conclude that only functional products are relevant for component reintegration programs. Material recycling should be preferred for closing the supply chain of psycho-sociological commodities since remanufacturing seems not to be viable on a big scale. Figure 6.6 recapitulates known case studies dealing with component-based recycling strategies. Thus, it is very interesting to notice that all important remanufacturing programs concern functional goods, therefore corroborating our hypothesis.

7 Computer Industry

7.1 The Environmental Challenge of Computer Manufacturing From a sustainability standpoint, the production process of desktop computers represents a huge ecological challenge which needs to be coped with. According to estimations of Williams (2003) and Cole (2003), the production of a 10 kg computer requires 240 kg of fossil fuels, 1.5 ton of clean water and at least 20 kg of chemicals. Unlike refrigerators which consume 20 times as much energy during the utilization phase than in the production phase (Williams, 2003), the manufacturing phase amounts for 90% of a computer’s energy needs during the whole life-cycle (Grote, 1994). Since the materials cited above are not available for free, we can assume that semiconductors manufacturers seek to minimize this resource input. However, the chemical processes required to produce wafers and keep them in a puriﬁed environment are resource-intensive and cannot be sensibly improved in an immediate future. In this context, reclaiming the eﬀorts embedded into used computers turns out to be the most practicable way to reduce the disequilibrated balance between input and output and thus to reach sustainability. The European directive on waste electrical and electronic equipment (WEEE) sets compelling material recycling targets for computer OEMs (see chapter 1.1). Although state of the art recyclers are currently able to recover about 100% of a computer’s metallic fraction,1 material recycling faces two shortcomings. First, it is not possible to fully close the loop, especially because of the diﬃculty to separate plastics from metal on printed boards. Secondly, reclaimed materials provide an in1

Representing 68% of an average desktop weight (Spengler et al., 2003).

120

7 Computer Industry

put for the manufacturing processes so that the environmental burden caused in the production phase remains unchanged. We already noticed in chapter 2 that measures targeting clear eco-eﬃciency improvements have to be taken during the production phase of computers, i.e. with help of components reintegration. In fact, the reuse or refurbishing of used computers might not be only ecologically motivated. The development of second-hand markets is documented in recent publications and can be interpreted as a weak signal (Ansoﬀ, 1975) for OEMs. In Japan for instance, the yearly demand for second-hand desktop computers is estimated to 980,000 computers2 in 2002 (Williams and Kuehr, 2003). Shrinking IT-budgets and the increased rationality when it comes to evaluate the utility of new hardware might explain why buyers often prefer not to keep pace with every innovation cycle. OEMs, among others Fujitsu-Siemens, IBM, Sun, HP or Apple, have recognized this trend and now also oﬀer refurbished computers (see Apple, 2005). On the one side, computer refurbishing seems to be a market niche both ﬁnancially and ecologically attractive but its impact on a company’s proﬁts is unknown in advance. On the other side, manufacturers now comply with the take-back and recycling legislation but outsource this activity to the market and do not perform a closed-loop at the material level. Given this, the following analysis will aim at determining to what extent computer OEMs should create a signiﬁcant closed-loop with respect to the initial investments in recovery capabilities.

7.2 Model Implementation The present study focuses on the remanufacturing potential of desktop computers in a saturated market. According to a survey performed by Saied and Velasquez (2003), the average service lifetime of computers is about six to seven years. In other industrialized countries a similar behaviour of PC users can be assumed. Hence, our investigations encompass 20 years, beginning in 2005. The consideration of several replacement cycles during the optimization should help the model reaching a stable state. The desktop market is divided into two segments, oﬃce and home, in which two product performances are supplied (high-end computers, also called power in the following, and average computers depicted as 2

The same study estimates the yearly demand for new personal computers to be 12 million units/year.

7.2 Model Implementation

121

normal). These computers may be either sold or leased with a term of three years. The OEM has the possibility to refurbish his product returns but the retail price of these items will depend on the initial performance of the computer. Our model considers a monopolist supplying six generic products: Power lease (P L), normal lease (N L), power (P ) and normal (N ) sales, as well as refurbished power (P R) and normal (N R) desktop sales. Since the residual value of refurbished products depends on their former service life, we will specify three sub-products for every refurbished group. A remanufactured PC can either be one (P R1 , N R1 ), two (P R2 , N R2 ) or three (P R3 , N R3 ) years old. As a consequence, the following investigations will encompass 10 diﬀerent products p. A trade-oﬀ appears when we deal with component modelling. On one side, an accurate modelling of component reuse is required to catch the dynamics of computer remanufacturing. On the other side, the complex structure of computers leads to a situation where a model’s precision has to be balanced with its solvability. For instance, the permanent evolution of devices with respect to performance and functionality complicates comparisons between components of diﬀerent aging. A common display of 1993 (CRT 14 inches) cannot be compared with a standard 15 inches TFT of 2003. Given this, a generic computer view enabling time-related component matching would help coping with modelling diﬃculties. We limit our analysis to six components outlined in ﬁgure 7.1. Some input and output devices are excluded because of their marginal fraction in a computer’s value (mouse and keyboard) or because of their accessory nature (scanner and printer). For reasons of close interaction, motherboards and processors are, in the following, considered as a whole. Concerning storage capabilities, our investigations distinguish two types of external disc drives, which are often both available in current computers: One read-only drive and a further drive with writing functionality. Despite steady performance improvements, both drives remain constant in their functionality. Thus, we assume that a computer is always equipped with one read-only device (ROD) and one read/write-device (RWD). This aggregation level is sustainable since the shift from CD to DVD-drives shows the same patterns. It further enables the consideration of future technological shifts that might happen after 2007 although it is currently impossible to forecast which technology will replace digital videodiscs.

122

7 Computer Industry

Fig. 7.1. Investigation scope: selected PC modules, Stahlknecht et al. (2002)

7.2.1 Demand Segmentation According to Williams and Kuehr (2003), most of the demand for desktops (75%) is generated by oﬃce customers, whereas home customers account for 25% of the total market demand. The determination of a proﬁt maximizing product mix is constrained by the potential market share upper bounds published in table 7.1 and derived from Williams and Kuehr. max Table 7.1. Market share upper bounds βp,q (initial scenario)

Product Oﬃce Home Product Oﬃce Home PL P P R1 P R2 P R3

0.1 0.1 0.3 0.2 0.1

0 0.4 0.15 0.1 0.05

NL N N R1 N R2 N R3

0.5 0.5 0.15 0.1 0.05

0 0.7 0 0 0

New power computers are generally available for 1,500 Euros in Europe whereas average appliances are available for 900 Euros (Seitz et al., 2003). Concerning refurbished products, a rule of thumb of the computer industry states that electronic appliances lose one percent of their value per week or approximatively 50% per year (Kriebel, 2004).

7.2 Model Implementation

123

In this context, the retail price of these computers σp∈ref urb ,t+a is a function of their age a and initial value σp∈new ,t where σp∈ref urb ,t+a = σp∈new ,t · 0.5a . For instance, a three year-old power desktop is valued at 1500 · 0.53 = 188 Euros. These estimations match with the observations of Williams and Kuehr (2003) on the second-hand markets. Leased products (PL and NL) are considered as operational lease so that their value depends on the lease duration d and the residual value at the end of the contract. Thus we obtain σp∈lease ,t+d = σp∈new ,t · (1 − 0.5d ): The lease price of a power desktop with a three years contract amounts to 1312 Euros (1500 − 1500 · 0.53 ). The procurement costs for new components (κN c,p,t ) were exposed to strong price variations between 1997 and 2003 and no clear trend can be statistically identiﬁed concerning the evolution of the market prices (Seitz et al., 2003). However, the performance improvements are driven by the concern of maintaining component prices at a constant level. As a consequence, our study relies to the future periods on the average component prices between 1997 and 2003 which are listed in table 7.2. New parts are only required for refurbished computers when the returned parts are defective and excluded from the reintegration process. Table 7.2. Average component prices (1997-2003), in Euro (Seitz et al., 2003) Products Component PL NL P N PR NR cpu ram gra hd rod rwd

440 100 140 225 100 225

265 60 85 135 60 135

440 265 100 60 140 85 225 135 100 60 225 135

265 60 85 135 60 135

265 60 85 135 60 135

7.2.2 Return Flow Timing and Quantities Computer OEMs have not shown great interest for end-of-life reverse ﬂows so far. This lack of initiative leads to late returns so that reverse ﬂows are almost inapt for component reuse and must be therefore recycled. As Matthews and Matthews (2003) document (see table 2.2), private users often store their obsolete computers a few more years.

124

7 Computer Industry

Computer manufacturers prefer not to inﬂuence this customer behavior and reclaim consequently a quite insigniﬁcant fraction of the returns. Fujitsu-Siemens Computers, for instance, processed 2001 about 1% of its sold oﬃce computers in its own recovery centre (Fujitsu-Siemens Computers, 2002). Given this lack of interest by OEMs and an average service life of 6 years, we approximate the probability of return with help of a normal distribution with a mean of 8 years (assuming a stockpile duration of two years) and a standard deviation of 3 years which includes the fact that computers might be stockpiled after the end of their service life. Leased desktops (PL, NL) are returned after three years of use (d = 3).

ρp,t ,t =

⎧ (t−t −8)2 − 1 ⎪ √ 18 ⎪ · e ⎨ 3 2π ⎪ ⎪ ⎩

1 0

∀p ∈ {P, N, P R, N R}, t ≥ t ∀p ∈ {P L, N L}, t = t + 3 else

7.2.3 Reintegration Potential As mentioned before, the reintegration potential of a reclaimed part depends on two factors: Wearing and technological evolution. The ﬁrst prerequisite is, in fact, that the core still works and that it presents enough service life expectancy to overcome a next utilization phase without failing. The probability that a core fulﬁlls this requirement is R . The investigated parts are set expressed with help of parameter θc,t ,t under two diﬀerent wearing patterns. Parts with a steady performance such as cpu, ram, rod, rwd or graphic card are working in a way that almost no wearing signs might be perceived by the user. Unlike this ﬁrst group, the performance of hard drives declines considerably over time. The reintegration probability of a wearing-exposed item is computed with the same function as in the tires’ case.3 Due to data security concerns, g4 is valued with 0.5 for hard drives, i.e. these are automatically expelled from remanufacturing after two years of use. For the parts not exposed to noticeable stress but which may also fail, g is valued with 0.025. The technical reintegration parameter δc,p,t ,t determines the possible assignments of reclaimed components c to the desktop type p. Used parts with an age of a years can only be reused in refurbished products of the same age or younger (P Ra , N Ra ). 3 4

We refer therefore to section 6 for modeling issues. Representing the wearing exposure of a part.

7.2 Model Implementation ⎧ ⎪ ⎪ ⎨1

δc,p,t ,t =

1

⎪ ⎪ ⎩0

125

∀c, p ∈ {P Ra }, t ≤ t + a ∀c, p ∈ {N Ra }, t ≤ t + a else

The collection costs for used desktop computers account for 10 Euros which are broken down equitably between the six parts, i.e. κRF c,t ,t = 1.67 (∀c, t , t). The disassembly of computers has to be processed manually in order to avoid damages. Given this and the costs for testing reintegrable cores, we assume that every reused part induce reintegration costs κR c,t ,t of 5 Euro each. Useless parts are disposed of O at a price κc,t ,t which is estimated at 0.5 Euro per part (Klatt, 2003, Pepi, 1998). Finally we estimate that the assembly costs κPp,t are equivalent for both new and refurbished computers, and yield 20 Euros per computer. Considering that all reclaimed parts should be removed and tested before reuse, the assembly eﬀorts should not diﬀer between new and older cores as they provide the same functionality. On the basis of Klatt’s (2003) and Pepi’s (1998) estimations, we asN sess the cost of new materials for a computer κM m,t to 3.4 Euros per computer. We limit the scope of the materials analysis as we gather all computers materials into one aggregated block. As a consequence, a desktop’s materials mix is supposed to remain constant over the 20 years of the investigations. The variable costs for reclaiming one kg of recycled material amounts approximatively for 0.5 Euro (Pepi, 1998). Following the information provided by Podratzki (2003) and Klatt (2003) about the material reintegration rates, we set the recyclability fraction µm,t to 50% of a computer’s weight. This ﬁgure encompasses all metallic materials which could be reintegrated as new after separation. Sarkis (2003) and Klatt (2003) also provide reliable ﬁgures with respect to the capital required to setup recovery centers able to process two million computers yearly.5 The investments necessary to open a recycling plant with a shredder amount for 20,000,000 Euros and induce yearly overhead of one tenth of the initial investment (κFt R = 2,000,000). The refurbishing process, which is more personal-intensive, requires lower investments (inv M = 5,000,000 Euros) to process and sort the incoming ﬂow. Few data is available on the overhead of this size, however, the demanufacturing line overhead κFt M would at least amount for 500,000 Euros yearly and should be close to the overhead 5

Given the turnover of the manufacturer investigated and the return distributions used, the capacity of the asset center does not represent a bottleneck for the closed-loop.

126

7 Computer Industry

of Compaq’s former plant (Sarkis, 2003). To assess the return on investment of the closed-loop supply chain, the investigation underlies an average cost of capital wacc of 10% per year.

7.3 Optimization Results In order to understand to what extent closed-loop supply chains are attractive for Original Equipment Manufacturers, we compute in the following ten scenarios (see ﬁg. 7.2) ﬁrst investigating the drivers leading to an internal reintegration of the recycled materials. Scenario ini depicts the situation of a German computer manufacturer prior to the introduction of the WEEE while scenario reg assesses the ﬁnancial burden to comply with the legislation. Scenario reg depicts a manufacturer outsourcing the resource recovery process, scenario rec computes the discounted cash-ﬂow in the case of an internal reintegration of the recovered materials. In order to understand the impact of the materials’ price level on the attractiveness of recycling-based closed-loops, scenario sca underlies a yearly increase of 2% for the procurement of virgin materials. Finally, scenario scm assesses the impact of a 100% material recycling rate with the contingencies of scenario sca. Consequently to the introduction of the WEEE, we analyze in a second step the proﬁtability yields of product refurbishing programs. Scenario rem computes the case by which a manufacturer might refurbish his products with the possibility to internally reintegrate the residual materials while scenario reo does not allow internal recycling activities. Comparing both scenarios will be useful to determine to what extent recycling aﬀects the proﬁtability of remanufacturing. Finally, we investigate the impact of measures intended to improve the attractiveness of computer refurbishing. One potential way to achieve this is to increase the residual value of used computers. In this context, scenario lam assumes a yearly value loss pattern of 40% instead of 50%. Like in the tire’s case, a preponement of the reverse ﬂows might increase the remanufacturing potential of computers (sc. pre). We assume thereby that the distribution of the reverse ﬂows follows a normal distribution with a mean of six years instead of eight.6 To improve the proﬁtability of product reintegration, the lawmaker is also able to penalize the product returns which cannot be refurbished. Scenario dis computes the case where the disposal fee accounts for 15 Euros per computer not remanufactured. 6

For the products sold only. The leased products still have a leasing duration of three years.

7.3 Optimization Results

Product stewardship: Impact assessment

sc. ini initial situation, no financial burden

sc. scm own recycling with yearly increases of material prices (2%) and 100% recycling rates

Remanufacturing: Impact assessment

sc. reg current situation: WEEE without own recycling

sc. reo enable remanufacturing without recycling

sc. rec WEEE with own recycling

sc. rem enable remanufacturing

sc. sca own recycling with yearly increases of material prices (2%)

127

sc. pre prepone reverse flow distribution

sc. lam improve residual value pattern

sc. dis disposal penalty of 15 / computer

Fig. 7.2. Computer CLSC assessment: Scenarios overview

7.3.1 Impact Assessment of the European Product Stewardship Laws According to the results of scenario reg (see table 7.3), the Waste Electrical and Electronic Waste directive (WEEE) induces an additional burden with a net present value of 108 M Euros7 or 2.7 Euros per computer sold. This ﬁgure is comparable with the estimations of a major computer manufacturer for Germany (Kriebel, 2004). The main part of this burden is assigned to the logistics eﬀorts required to guide the used products to a point of recovery (2.08 Euros). As targeted by the takeback legislation, the setup of an own recycling pipeline improves the ﬁnancial situation of the manufacturer investigated (sc. rec). The net present value (NPV) of an immediate investment in a material-based closed-loop yields N P Vrec − N P Vreg = 67 M Euros if the recyclable fraction amounts for 50% of the materials’ weight. After the launch of the closed-loop, the OEM saves 82 M Euros on the procurement of virgin material (2.07 Euros per computer). The savings on disposal costs are compensated by the processing costs to reclaim materials. The manufacturer carries additional overhead and depreciation costs for the recovery plant which amount for approximatively 15 M Euros (0.39 Euro). This internal recovery strategy might bring additional beneﬁts when the material prices soar by 2% yearly during the next twenty years 7

M = million

128

7 Computer Industry

(sc. sca). The discounted procurement costs would increase by 0.81 Euro up to 7.25 Euros per computer. Nevertheless, the manufacturer would still be better oﬀ than in the current situation without material price increases (+ 34 M Euros). As a result, the manufacturer has an incentive to improve the recyclability of his products under such cost settings. This presumption is conﬁrmed in scenario scm which assesses the impact of a material reintegration rate of 100% with the premises of scenario sca. The immediate possibility to fully reintegrate the materials initially used has a positive net present value of N P Vscm − N P Vsca = 102 M Euros or 5 M Euros per year. This positive result stems mostly from a 35% decrease of the procurement costs for new materials (from 7.25 to 4.72 Euros per computer). Table 7.3. WEEE impact: Optimization results Scenarios (ﬁnancial ﬁgures in EUR) ini reg rec sca scm Total NPV (in million) 1,758 1,650 1,717 1,684 NPV / unit sold 43,94 41,24 42,92 42,11 sales 428.11 assembly costs 8.51 procurement new parts 367.15 reverse logistics costs 2.08 reassembly costs proc. new materials 8.51 6.44 7.25 recovered materials 0.31 0.31 overhead disassembly overhead recycling 0.39 disposal costs 0.62 0.31 0.31 % of sales PL 7.5% NL 22.5% P 17.5% N 52.5%

1,786 44,64

4.72 0.62

0

The results obtained in this case study support the insights obtained from the strategic model of chapter 3: The legislation plays a role in making recycling more attractive for original

7.3 Optimization Results

129

equipement manufacturers. While the launch of a recycling program turns out to be a wrong investment without the existence of the legislation (N P Vrec − N P Vini = −108 M Euros), it is proﬁtable when the WEEE is already in place since N P Vrec − N P Vreg = +67 M Euros. Furthermore, manufacturers’ eﬀorts to increase the recyclable fraction of their computers (sc. scm) pays oﬀ in the case where the raw material prices keep on rising constantly in the near future. Nevertheless, without a proprietary reverse chain, OEMs might not be able to reclaim their eﬀorts as the market prices for recycled materials are correlated to the prices of virgin resources. Without closed-loop, the recyclability improvements performed by manufacturers would only beneﬁt to waste processors who would increase their revenues. 7.3.2 Impact Assessment of Computer Refurbishing Concerning remanufacturing, the initial research question is to ﬁnd out whether desktop refurbishing would be proﬁtable without the existence of own recycling channels8 (sc. reo). We notice a cash ﬂow increase of 105 M Euros from the current situation (sc. reg), or 2.6 Euros per computer. This improvement is mainly due to the fact that the OEM is able to save 755 M Euros on components’ procurement while the turnover goes down by 624 M Euros. The apparent cannibalization effect between the N L product line and the refurbished line stems from the low relative proﬁtability of bottom-line computer leasing compared to refurbished desktops. However, the refurbished product line is, according to the results, completely dependent on the lease returns as the sales of P R3 show. As recycling and remanufacturing are mutually exclusive pipelines, we evoked in chapter 5 the potential existence of an internal competition between both recovery paths. In other words, the ﬁnancial attractiveness of material recovery might harm product refurbishing by guiding remanufacturable products into the recycling pipeline. A comparison between scenario reo and scenario rem shows us that this presumption does not hold in the computer’s case. In fact, both recovery paths support each other since the market shares remain unchanged but the OEM’s closed-loop present value grows by N P Vrem − N P Vreo = 53 M Euros or 1.32 Euro per computer sold.

8

Which is currently the case for the great majority of the computer manufacturers.

130

7 Computer Industry Table 7.4. Remanufacturing impact: Optimization results Scenarios (ﬁnancial ﬁgures in EUR) rem reo pre lam dis Total NPV (in million) 1,808 1,755 2,076 1,662 1,511 NPV / unit sold 45.21 43.89 51.92 41.54 37.78 sales 412.49 405.1 400.72 409.78 assembly costs 8.51 procurement new parts 348.27 333.72 340.02 345.14 reverse logistics costs 1.98 2.27 1.94 1.95 reassembly costs 0.78 1.39 1.07 1.04 proc. new materials 6.28 7.99 5.79 6.22 6.3 recovered materials 0.26 0.27 0.24 overhead disassembly 0.56 overhead recycling 0.39 0.39 disposal costs 0.26 0.51 0.27 0.24 7.25 % of sales PL 7.5% 6.2% 7.5% NL 14% 8.16% 12.03% 12.27% P 17.5% N 52.5% P R1 0.28% 1.25% 0.28% P R2 0.37% 1.22% 0.37% P R3 6.38% N R1 0.63% 2.78% 0.67% 0.65% N R2 0.84% 2.71% 0.88% 0.86% N R3 0% 0% 3.19% 1.69%

As expected, the cannibalization eﬀects are conﬁrmed by scenario lam assuming that the yearly loss amounts to 40% of a product’s value instead of 50%. Nevertheless, an interesting eﬀect occurs: By improving the residual value of a product, manufacturers will automatically reduce the contractual price for new leased items. As an example, instead of charging 900 · (1 − 0.53 ) = 787.5 Euros for an average computer N L, the manufacturer would charge 900 · (1 − 0.63 ) = 784 Euros. This change in the value evolution pattern of N L and refurbished products is in favour of N R3 which is then proﬁtable to remanufacture.

7.3 Optimization Results

131

The lawmaker might also increase the attractiveness of refurbishing by charging a 15 Euros fee in the case a returned computer cannot be refurbished (sc. dis). According to the optimization results, we observe that this measure would solely improve the cost margins for the products’ bottom line N R3 . Thanks to the recovery programs, the real disposal burden accounts for 7.25 Euros per computer. As this case study reveals, early returns, which are mostly linked to lease contracts, are proﬁtable to refurbish as soon as their current service life does not surpass three years. In fact, without product leasing, only a small fraction of the product returns (about 5% if all products are returned to the OEM)9 would be returned early enough. Considering the implementation diﬃculties of asset recovery programs, especially due to the internal competition between new and recovered products (see section 3.3.2), the marginal business relevance of refurbishing might work against the introduction of OEM-owned closed-loop supply chains. In this case, the additional management complexity may outweigh the potential beneﬁts of asset recovery programs. Generally speaking, the longer a product remains on the market, the lower is its business relevance. Apart from psycho-sociological reasons, table 7.5, which depicts the average service life of electrical and electronic appliances in Germany, gives advice about the very marginal business relevance of refurbishing for these products due to their long market sojourn. Table 7.5. Average service life of WEEE appliances (Hauser and R¨ottchen, 1995, Cooper, 1994)

appliance years ovens televisions PCs

18 8 8

appliance

years

wash. machines microwaves refrigerators

15 7 16

Kostecki (1998) explains the extended service life of these appliances with help of an opportunity costs’ analysis of product replacement. The author states that buyers will replace their appliances as soon as the 9

Given ρS t the return probability of sold products after t years of use, 0.0478

3 0

ρS t dt =

132

7 Computer Industry

expected discounted utilization cost10 diﬀerential between used and old new − κU rises above the procurement costs for a new new items κU t t product σN plus the take back costs κρ + κO . n

new κUold − κU t E t (1 + r)t t=0

≥ σ N + κρ + κO

(7.1)

This problem formulation explains why products with a high retail price and technically mature (e.g. houses, washing machines, cars) are kept longer by customers since the opportunity costs of use appear to be too low to justify a replacement. For the same reason, the willingness of second-hand buyers to pay for used items tends to zero when the used products are technically and functionally obsolete, i.e. when the expected opportunities costs are high. Therefore, in the absence of return incentives, rational ﬁnal users will return their end-of-life products to an asset recovery center when there is no other proﬁtable alternative. Hence, end-of-life reverse ﬂows of sold products are per se more a burden than a business opportunity. In this context, leasing appears to be the only means to reclaim enough valuable cores for remanufacturing. However, despite the enabling role of leasing for computer refurbishing, we observed in this case study that the development of product refurbishing is made at the expense of leased products. Given this, the introduction of a long-range remanufacturing program should be assessed with respect to its impact on the leased sales. The strategic planning model developed in chapter 5 and applied here seems to be an adequate means to cope with the complexity of decision-making in closed-loop supply chains.

10

κU t encompasses energy and maintenance costs.

8 Conclusion and Outlook

Thesis Overview Reverse logistics and its related ﬁelds sustainable supply chain management and closed-loop supply chains are now accepted as a main research ﬁeld supported by an extensive variety of mathematical decision models. Nevertheless, until now, the literature has not been speciﬁcally concerned with the identiﬁcation of best responses to the reverse ﬂow challenges faced by Original Equipment Manufacturers. While the current European take-back legislation targets an internalization of the recycling costs on the manufacturers’ side, we showed that these are still reluctant to implement recovery programs in which the reverse ﬂows would be reintegrated into the initial supply chain. In our opinion, this lack of interest is only partly due to the expected low profitability of resource reintegration. In fact, very few decision frameworks enable manufacturers to identify and quantify the strategic opportunity oﬀered by product returns. The monograph’s initial research objective was to ﬁll this gap and to structure the scattered insights of the reverse logistics research. To achieve this goal, we have adopted a perspective consisting of answering a sequence of questions managers would face during their investigations. Chapter 2 ﬁrst assessed the contribution of closed-loop supply chains to the achievement of a competitive strategy. We showed that product and material recovery provide a wide range of possibilities to support either cost-leadership or diﬀerentiation-based strategies. On the basis of these insights, we identiﬁed the critical processes which should be embedded into a company’s current value chain. Logically, the implementation of resource reintegration at the value chain level diﬀers slightly depending on which competitive strategy a manufacturer follows. Hence, closed-loops underly the same constraints as other com-

134

8 Conclusion and Outlook

petitive drivers: The potential of asset recovery can only be unleashed if the measures taken to support it are coherent with the current organization of a company’s value chain. In a second step, we developed in chapter 3 a mathematical model to quantify the ﬁnancial impact of the generic asset recovery strategies: Reuse, remanufacturing and recycling. We found out that the legislation has a positive impact on the attractiveness of recycling but may not be suﬃcient to convince OEMs committing to a closed-loop, especially when the disposal costs are low. With help of the model, the market patterns favorable to the development of reuse and refurbishing have been also identiﬁed. Generally speaking, if the price rebate granted for recovered products is lower than the procurement costs for the replacement parts, then remanufacturing is proﬁtable for an OEM although he would sell fewer new products. This situation especially applies when the product’s obsolescence eﬀects are limited over time. Despite its positive impact on a company’s proﬁts, we observed that reuse and remanufacturing might face acceptance problems within the company because of cannibalization eﬀects. Thus, the sales division in charge of refurbished products would make proﬁts at the expense of the division selling new products. In order to solve this problem, we developed a transfer pricing scheme to align the incentives in both sales divisions. We analyzed the prerequisites for the application of such an internal pricing policy and found out that it is possible to set a satisfactory transfer price for both divisions when reuse or remanufacturing is proﬁtable for the company as a whole. Given the internal implementation diﬃculties of product recovery strategies, it is not surprising to see that many remanufacturing programs have been initially launched by independent competitors reclaiming a manufacturer’s end-of-cycle items. In this situation, Original Equipment Manufacturers have to decide how to react to this source of cannibalization. While some producers now use leasing as a means to preempt the entry of free-riders, some others have chosen to compete against free-riders on the procurement of used cores. Furthermore, while manufacturers of toner cartridges prefer to destroy the reclaimed products instead of refurbishing them, truck tire OEMs have chosen to retread their own products. To determine the right asset procurement and recovery strategy, we extended in chapter 4 the closed-loop model of the previous chapter to this issue. We ﬁrst found out that a producer should rather destroy the reclaimed cores when the adjusted marginal income is lower than the average income per use phase, i.e. when the marginal revenues of refurbishing are signiﬁcantly lower than the mar-

8 Conclusion and Outlook

135

gins for new items. In this context, we pointed out that manufacturers of cartridges should rather start refurbishing activities instead of shedding the returned cores. We further identiﬁed ex-ante disposal systems as an eﬀective means for manufacturers to prevent the entry of independent refurbishers but, in accordance with the results, we concluded with the recommendation not to increase the complexity of disassembly operations as both OEM and refurbishers would be hit by such measures. The computations pointed out indeed, that a manufacturer would harm his own proﬁt function by increasing the refurbishing costs. While chapters 2 to 4 provided guidance into whether to start a resource recovery program or not, managers have also the task to concretize strategic objectives. The decision-making process, also called planning process, has been extensively addressed in chapter 5. Our investigations led to the conclusion that the scope of strategic closedloop planning does not diﬀer from the classical strategic network planning tasks. Starting from the observation that OEMs committed in remanufacturing operations have followed a sequence of development milestones, we presented a generic strategic network planning model, targeted at assessing the proﬁtability of remanufacturing for a given set of products with respect to technical, qualitative and psychological obsolescence. For this purpose, we embedded three key parameters into the planning model: The market structure, the return ﬂow distribution as well as the reintegration potential. This decision-making framework has then been extended to recycling and investment issues to address the problems faced by Original Equipment Manufacturers seeking to reduce the side costs of product refurbishing. Another way to improve the cost eﬃciency of closed-loop supply chains might consist in moving the labor-intensive activities to low cost countries. For the sake of practical relevance, we therefore extended the generic strategic planning model to the location planning issues. The second part of the monograph has been dedicated to practical applications implementing the theoretical concepts developed previously. First, we assessed in chapter 6 the remanufacturing potential of tires with help of the generic strategic planning model. The observed discrepancies between the reintegration potential of car and truck tires enabled us to isolate a product characteristic explaining the high relevance of tire or copier remanufacturing as well as the low importance of car tire or mobile phone refurbishing. To summarize, products with a functional nature, i.e. which are bought for their core utility rather than for psycho-sociological reasons, are candidates for high-range refurbishing programs. The sensitivity analysis performed in the tire case

136

8 Conclusion and Outlook

study underlined that the modiﬁcation of one key parameter would not automatically improve the proﬁtability of tire retreading. In fact, a preponement of the reverse ﬂow distribution or strengthening of a tire casing would only lead to additional sales of refurbished items if enough customers are willing to buy retreaded tires. As a conclusion, the product selection process prior to the setup of a closed-loop requires a comprehensive approach encompassing the interactions between technical and ﬁnancial parameters. After having pointed out the environmental burden caused by computers, we investigated in chapter 7 whether computer manufacturers should invest into a product or a material-based closed-loop supply chain. In accordance with the insights of chapter 3, we found out that the existence of the WEEE directive is a prerequisite to generate a positive discounted cash ﬂow from the setup of a recycling center. Given the expected return on investment of recycling, computer manufacturers should at least take this alternative into consideration although material recovery does not belong to their strategic portfolio so far. In addition to these insights, the strategic planning model contributed to quantify the impact of remanufacturing on a manufacturer’s global earnings. According to the optimization results, the refurbishing of recent computers turns out to be as well a very attractive alternative which, once again, cannibalizes the supply of a manufacturer’s bottom product line. Moreover, the development of refurbishing is tied to the availability of leasing returns and is obviously more proﬁtable than the operational leasing of average computers. Consequently, increasing the turnover through high-end leased items is strategically sounder than to support the lease contracts of low-end computers. We ﬁnally observed a complementarity between remanufacturing and recycling although both recovery alternatives might, a priori, exclude themselves. In the computer’s case, recycling contributes to improve the proﬁtability of remanufacturing operations so that the setup of a material recovery center should be the logical sequel of a computer refurbishing program. Future Research The present monograph should be interpreted as a scientiﬁc invitation to OEMs to investigate extensively the opportunities provided by closed-loops. Nevertheless, as Guide and van Wassenhove (2005) notice, manufacturers need to be aware of their return ﬂow issues before starting to deal with them. As we pointed out in chapter 2, most established information systems do not cover product returns so that decision-makers are not able to quantify the ﬁnancial burden related

8 Conclusion and Outlook

137

to the management of the reverse ﬂows in an accurate way. Consequently, one of the most critical future research challenge will consist in ﬁnding ways to extend current information systems and to develop eﬃcient business workﬂows supporting closed-loop supply chain management. To the knowledge of the author, very few publications (e.g. Corsten and Reis, 1994, Rautenstrauch, 1997, or Chouinard et al., 2005) provide preliminary insights into this problem. The existing environmental legislation and the unexpected reaction of Original Equipment Manufacturers to the regulatory constraints have been one of the main initiators of this study. In this context, we have limited our research to producers of complex goods and, thus, have left the chemical industry outside of this monograph. With respect to the upcoming increase of the oil price, however, the access to plastics from end-of-life appliances will surely become a critical competitive advantage for chemical companies in a near future. While most of the insights drawn in this monograph are generalizable to the chemical or paper industry, the methods to assess potential closed-loops in these industries would have also to take the speciﬁcities concerning the material composition constraints or the several markets for by-products into account. While the present monograph has mostly relied on linear programming techniques, real options analysis1 could provide another means to take decisions related to the setup of closed-loop supply chains. In fact, due to the high parametric uncertainty underlying the technical and ﬁnancial proﬁtability drivers, considering closed-loop supply chains as an option to reclaim value in the future might provide additional arguments for decision-makers to commit to sustainable supply chain management. The connection between real options analysis and investments in recovery assets has not been investigated in the literature so far. Considering the remaining research questions and the upcoming environmental challenges, we hope that the present monograph has provided enough insights for companies to consider reverse ﬂows as a tremendous business opportunity worth to be investigated.

1

We refer to Copeland and Antikarov (2004) for a comprehensive introduction to real options analysis and its beneﬁts compared to classical net present value assessments.

References

1. Allgemeiner Deutscher Automobilclub e.V. (2002) Reifenpreisliste September 2002. ADAC Motorwelt 10/2002. 2. Ansoﬀ I (1975) Managing Strategic Surprise by Response to Weak Signals. California Management Review, 18(2):21–33. 3. Anthony RN (1965) Planning and Control Systems: A Framework for Analysis. Studies in Management Control, Harvard University. 4. Apple Inc. (2005) Apple Store Second Hand. Link: http://promo.euro. apple.com/promo/refurb/de/. Last visited 30th September 2005. 5. Atasu A, Sarvary M and LN van Wassenhove (2005) Remanufacturing as a Marketing Strategy. Working Paper 05/63/TOM/MKT, INSEAD. 6. Ayres RU, Ferrer G and T Van Leynseele (1997) Eco-Eﬃciency, Asset Recovery and Remanufacturing. European Management Journal, 15(5):557–574. ¨ 7. Behrend S, Kreibich R and S Lundie (1997) Okobilanzierung komplexer Elektronikprodukte. Springer, Berlin. 8. Bellmann K (1990) Langlebige Gebrauchsg¨ uter. DUV, Wiesbaden. 9. Beullens P, van Oudheusen D and LN van Wassenhove (2004) Collection and Vehicle Routing Issues in Reverse Logistics. In: Dekker R, Fleischmann M, Inderfurth K and LN van Wassenhove (eds.) Reverse Logistics – Quantitative Models for Closed-Loop Supply Chains. Springer, Berlin, 95–134. 10. Blackburn J, Guide Jr VDR, Souza G and LN van Wassenhove (2004) Reverse Supply Chains for Commercial Returns. California Management Review, 46(2):6–22. 11. Bloemhof-Ruwaard J, van Wassenhove LN, Gabel HL and PM Weaver (1996) An Environmental Life-Cycle Optimization for the European Pulp and Paper Industry. European Journal of Operational Research, 88:490–503. 12. Braunmiller U and G W¨ ohrle (1999) Materialrecycling. IRB, Stuttgart.

140

References

13. Brucks M, Zeithaml VA and G Naylor (2000) Price and Brand Name as Indicators of Quality Dimensions for Consumer Durables. Journal of the Academy of Marketing Science, 28(3):359–374. 14. Bundesverband Reifenhandel und Vulkaniseur Handwerk e.V. (2000) Statistik, Markanteile der Reifenhersteller im Reifenersatzgesch¨ aft 1999. 15. Bundesverband Reifenhandel und Vulkaniseur Handwerk e.V. (2001) Statistik: Der Markt f¨ ur runderneuerte Reifen in Deutschland 2000. 16. B¨ urgermeister M, Obrocki S and C Calvez (2003) Wiederaufbereitung von Tonerkartuschen. Optimization Technical Report, University of Augsburg. 17. Carter CR and LM Ellram (1998) Reverse Logistics: A Review of the Literature and Framework for Further Investigations. Journal of Business Logistics, 19(1):85–102. 18. Chopra S and P Meindl (2001) Supply Chain Management. PrenticeHall, New Jersey. 19. Chouinard M, d’Amours S and D Ait-Kadi (2005) Integration of Reverse Logistics Activities within a Supply Chain Information System. Computers in Industry. 56(1):105–124. 20. Christopher M (1998) Logistics and Supply Chain Management. Prentice-Hall, New-Jersey. 21. Coenenberg AG, Cantner J and C Fink (2003) Kostenrechnung und Kostenanalyse. 5th ed., Sch¨ aﬀer-Poeschl, Stuttgart. 22. Cole D (2003) Energy Consumption and Personal Computers. In: Kuehr R and E Williams (eds) Computers and the Environment: Understanding and Managing their Impacts. Kluwer, Dordrecht, 131–160. 23. Cooper T (1994) Beyond Recycling: The Longer Life Option. New Economics Foundation, London. 24. Continental Tire North America Inc (2003) Form SS-HDR 01/03. 25. Copeland T and V Antikarov (2003) Real Options: A Practioner’s Guide. 2nd ed., Texere, New York. 26. Corbett CJ and LN van Wassenhove (1993) The Green Fee: Internalizing And Operationalizing Environmental Issues. California Management Review, 36(1):116–135. 27. Corsten D and M Reiss (1991) Recycling in PPS-Systemen. Die Betriebswirtschaft, 54:615–627. 28. Davey S, Guide Jr VDR, Neeraj K and LN van Wassenhove (2005) Commercial Returns of Printers: The HP Case. In: Flapper SDP, van Nunen JAEE and LN van Wassenhove (eds) Managing Closed-Loop Supply Chains. Springer, Berlin, 87–96. 29. Dawar N and A Chattopadhyay (2002) Rethinking Marketing Programs for Emerging Markets. Long Range Planning, 35:457–474. 30. de Brito MP and R de Koster (2004) Product Returns: Handling and Warehousing Issues. In: Dekker R, Fleischmann M, Inderfurth K and LN van Wassenhove (eds.) Reverse Logistics – Quantitative Models for Closed-Loop Supply Chains. Springer, Berlin, 135–156.

References

141

31. de Brito MP, Dekker R (2004) A Framework for Reverse Logistics. In: Dekker R, Fleischmann M, Inderfurth K and LN van Wassenhove (eds.) Reverse Logistics – Quantitative Models for Closed-Loop Supply Chains. Springer, Berlin, 3–28. 32. de Brito MP, Flapper SDP and R Dekker (2003) Reverse Logistics: A Review of the Case Studies. Report Series Research in Management ERS-2003-012-LIS, Erasmus University Rotterdam. 33. Deierlein R (1988) Fleet Feeback From Buyers and Speciﬁers – Tire Management. Fleet Equipment, 11:43–47. 34. Deierlein R (1997) Retreading Has Finally Been Accepted. Beverage World, 97(6):119. 35. Debo L, Toktay LB and LN van Wassenhove (2005) Market Segmentation and Production Technology Selection for Remanufacturable Products. Working Paper 05/01/TOM, INSEAD, Fontainebleau. 36. Debo L, Savaskan C and LN van Wassenhove (2004) Coordination in Closed-Loop Supply Chains. In: Dekker R, Fleischmann M, Inderfurth K and LN van Wassenhove (eds.) Reverse Logistics – Quantitative Models for Closed-Loop Supply Chains. Springer, Berlin, 295–312. 37. Dekker R, Fleischmann M, Inderfurth K and LN van Wassenhove (2004) Quantitative Models for Reverse Logistics Decision Making. In: Dekker R, Fleischmann M, Inderfurth K and LN van Wassenhove (eds.) Reverse Logistics – Quantitative Models for Closed-Loop Supply Chains. Springer, Berlin, 29–41. 38. Domschke W and A Scholl (2003) Grundlagen der Betriebswirtschaftslehre. Springer, Berlin. 39. Driesch HM, van Oyen HE and SDP Flapper (2005) Recovery of Car Engines: The Mercedes-Benz Case. In: Flapper SDP, van Nunen JAEE and LN van Wassenhove (eds) Managing Closed-Loop Supply Chains. Springer, Berlin, 157–168. 40. Dutz E and C Femerling (1994) Prozessmanagement in der Entsorgung – Ans¨atze und Verfahren. Die Betriebswirtschaft, 54(2):221–245. 41. Ebers M and W Gotsch (1999) Institutionen¨ okonomische Theorien der Organisation. In: Kieser A (ed.) Organisationstheorien. Stuttgart, Kohlhammer, 199–252. 42. European Commission (1998) Directive 98/71/EC of the European Parliament and of the Council of 13 October 1998 on the Legal Protection of Designs. Oﬃcial Journal of the European Community, L289:28-35, October, 28th 1998. 43. Fandel G and M Stammen (2004) A General Model for Extended Strategic Supply Chain Management with Emphasis on Product Life Cycles Including Development and Recycling. International Journal of Production Economics, 89(3): 293–308. 44. Ferber S (2005) Strategische Kapazit¨ ats- uns Investitionsplanung in der globalen Supply Chain eines Automobilherstellers. Shaker, Aachen.

142

References

45. Ferrer G (1997) The Economics of Tire Remanufacturing. Resources, Conservation and Recycling, 19:221-255. 46. Ferrer G (1997) The Economics of PC Manufacturing. Resources, Conservation, Recycling, 21:79–108. 47. Ferrer G (2001) On the Widget Remanufacturing Operation. European Journal of Operational Research, 135:373–393. 48. Flapper SDP (2003) Recovery Strategies. In: Guide Jr VDR and LN van Wassenhove (eds) Business Aspects of Closed-Loop Supply Chains: Exploring the Issues. Carnegy Mellon University Press, Pittsburgh, 71– 92. 49. Flapper SDP, van Nunen JAEE and LN van Wassenhove (2005) Managing Closed-Loop Supply Chains. Springer, Berlin. 50. Fleischmann B, Meyr H and M Wagner (2005) Advanced Planning. In: Stadtler H and C Kilger (eds) Supply Chain Management and Advanced Planning. 3rd ed., Springer, Berlin, 81–107. 51. Fleischmann B, Ferber S and P Henrich (2005) Strategic Planning of BMW’s Global Production Network. Interfaces, accepted for publication. 52. Fleischmann M (2001) Quantitative Models for Reverse Logistics. Springer, Berlin. 53. Fleischmann M, Beullens P, Bloemhof-Ruwaard JM and LN van Wassenhove (2001) The Impact of Product Recovery on Logistics Network Design. Production and Operations Management, 10(2):156–173. 54. Fleischmann M, van Nunen JAEE and B Gr¨ ave (2003) Integrating Closed-Loop Supply Chains and Spare Parts Management. Interfaces, 33(6):44–56. ¨ 55. Fujitsu-Siemens Computers (2002) Okobilanz Wiedervermarktung. 56. Furubotn EG and S Pejovich (1972) Property Rights and Economic Theory: A Survey of Recent Literature. Journal of Economic Literature, 10:1137–1162. 57. Geissler T (2004) Personal Interview about the Market Potential of Refurbished Cartridges, 16th June 2004. 58. Giarini O and W Stahel (1989) The Limits to Certainty. Kluwer, Dordrecht. 59. Geyer R. and LN van Wassenhove (2002) Remanufacturing Products with Limited Component Durability and Finite Life Cycles. Insead Working Paper. 60. Goetschalckx M and B Fleischmann (2005) Strategic Network Planning. In: Stadler H and J Kilger (eds) Supply Chain Planning and Advanced Planning Systems. 3rd ed., Springer, Berlin, 117–137. 61. Goh TN and N Varaprasad (1986) A Statistical Methodology for the Analysis of the Life-Cycle of Reusable Containers. IIE Transactions, 18:42–47. 62. Goodyear Inc. (2003) Goodyear Retreads Handle Tough Terrain.

References

143

63. Grant D (1999) Recycling and Market Power: A More General Model and Re-Evaluation of the Evidence. International Journal of Industrial Organization, 17:59–80. 64. Grote A (1994) Gr¨ une Rechnung. C’t, 94(12):94–101. 65. Guide Jr VDR (2000) Production Planning and Control for Remanufacturing: Industry Practice and Research Needs. Journal of Operations Management, 18:467–483. 66. Guide Jr VDR and LN van Wassenhove (2001) Managing Product Returns for Remanufacturing. Production and Operations Management 10(2):142–154. 67. Guide Jr VDR and LN van Wassenhove (2003) Managing Product Returns for Remanufacturing. In: Guide Jr VDR and LN van Wassenhove (eds) Business Aspects of Closed-Loop Supply Chains: Exploring the Issues. Carnegy Mellon University Press, Pittsburgh, 355–379. 68. Guide Jr VDR, Neeraj K, Newman C and LN van Wassenhove (2005) Cellular Telephone Reuse: The ReCellular Inc. Case. In: Flapper SDP, van Nunen JAEE and LN van Wassenhove (eds) Managing Closed-Loop Supply Chains. Springer, Berlin, 151–156. 69. Gesamtverband Autoteile-Handel e.V. (2004) EU-Designschutz f¨ ur Autoersatzteile: Zahlt der Autofahrer drauf? NRW-FDP und GVA fordern Verbraucherschutz und Wettbewerb. GVA e.V., May 2004. 70. Gutenberg E (1983) Grundlagen der Betriebswirtschaftslehre. Band 1: Die Produktion. 23rd ed., Springer, Berlin. 71. Hall RA (2001) Product Leasing Increases Chances of Reuse. Waste Age, 32(6):6–8. 72. Prahalad CK and G Hamel (1990) The Core Competencies of the Corporation. Harvard Business Review, 68(3):79–91. 73. Hansen U (1999) Produktkreisl¨ aufe. Schl¨ ussel zum nachhaltigen Wirtschaften. Fraunhofer IRB, Stuttgart. 74. Harvard Business School (2003) Managing Change and Transition. Harvard Business School Press, Boston. 75. Hauser H and P R¨ ottchen (1995) Elektro- und Elektronikger¨ ate. In: Rinschede A and KH Wehking (eds) Entsorgungslogistik, Bd. 3: Kreislaufwirtschaft. Erich Schmidt Verlag, Berlin, 139–160. 76. Hax AC and HC Meal (1975) Hierarchical Integration of Production Planning and Scheduling. In Geisler MA (ed.) Studies in Management Sciences, Vol. 1: Logistics. Elsevier, New York. 77. Herold M and G Kovacs (2005) Creating Competitive Advantage with End-of-Use Products. Logistik Management, 7(1):42–56. 78. Hewlett Packard Inc. (2004) Das Umweltschutzprogramm fr Original HP Zubeh¨ or. Link: http://www.hp.com/ereturns/, last visited: 30th September 2005. 79. Hollander A and P Lasserre (1998) Monopoly and the Preemption of Competitive Recycling. International Journal of Industrial Organization, 6:489–497.

144

References

80. Institut f¨ ur Energie und Umweltforschung Heidelberg (1999) ¨ Okologische Bilanzen in der Abfallwirtschaft – Fallbeispiel Verwertung von Altreifen. Umweltbundesamt, Berlin. 81. Ihde GB, Merkel H and R Henning (1999) Ersatzteillogistik. 3rd edition, Huss Verlag, Munich. 82. Inderfurth K (2002) Product Recovery Behavior in a Closed-Loop Supply Chain. FEMM working paper 6/2002, University of Magdeburg. 83. Inderfurth K, Flapper SDP, Lambert AJD, Pappis C and TG Voutsinas (2004) Production Planning for Product Recovery Management. In: Dekker R, Fleischmann M, Inderfurth K and LN van Wassenhove (eds.) Reverse Logistics – Quantitative Models for Closed-Loop Supply Chains. Springer, Berlin, 275–294. 84. InkSolution Ltd. (2005) Average Procurement Costs for Standard Cartridge Parts. Link: http://www.inksolution.net, last visited: 30th September 2005. 85. Jayaraman V, Guide Jr VDR and R Srivastava (1999) A Closed-Loop Logistics Model for Remanufacturing. Journal of the Operational Research Society, 50:497–508. 86. Johnson HT and RS Kaplan (1988) Relevance Lost – The Rise and Fall of Management Accounting. Harvard Business School Press, Boston. 87. Kelle P and EA Silver (1989) Forecasting the Returns of Returnable Containers. Journal of Operations Management, 8(1):17–35. 88. W Kerr and C Ryan (2001) Eco-Eﬃciency Gains from Remanufacturing – a Case Study of Fuji Xerox. Journal of Cleaner Production, 9(1):75–81. 89. Kircher T (2004) Analyse des Altfahrzeugaufkommens zur Erstellung von R¨ ucklaufprognosen im Rahmen der EU-Altfahrzeugrichtlinie. Master Thesis, University of Augsburg. 90. Klatt S (2003). Recycling Personal Computers. In: Kuehr R and E Williams (eds) Computers and the Environment: Understanding and Managing their Impacts. Kluwer, Dordrecht, 211–230. 91. Klausner M and C Hendrickson (2000) Reverse-Logistics Strategy for Product Take-Back. Interfaces 30(3):156–165. 92. Kopicki RJ, Berg MJ, Legg L and C Maggioni (1993) Reuse and Recycling: Reverse Logistics Opportunities. Council of Logistics Management, Oak Brook. 93. Kostecki M (1998) The Durable Use of Consumer Products: New Options for Business and Consumption. Kluwer, Dordrecht. 94. Kotler P (1999) Grundlagen des Marketing. 2nd ed., Prentice-Hall, New Jersey. 95. Kotler P (2003) Marketing Management. 11th ed., Prentice-Hall, New Jersey. 96. Kriebel S (2004) Organisatorische und ﬁnanzielle Herausforderungen an die Hersteller aus der Produktverantwortung durch die WEEERichtlinie. Master Thesis, University of Augsburg.

References

145

97. Krikke HR, van Harten A and PC Schuur (1998) On a Medium Term Product Recovery and Disposal Strategy for Durable Assembly Products. International Journal of Production Research, 36(1):111–139. 98. Krikke HR, van Harten A and PC Schuur (1999) Business Case Oce: Reverse Logistics Network Redesign for Copiers. OR Spektrum, 21(3):381– 409. 99. Krikke HR, le Blanc, HM and S van de Velde (2004) Product Modularity and the Design of Closed-Loop Supply Chains. California Management Review, 46(2):23–39. 100. Kreislaufwirtschafts- und Abfallgesetz – KrW-/AbfG (2001). In: Umweltrecht, 295–329. 13th ed., DTV, Frankfurt. 101. Kroon L and G Vrijens (1995) Reusable Containers: An Example of Reverse Logistics Management. International Journal of Physical Distribution and Logistics Management, 25(2):56–68. 102. Kuik R, van Nunen JAEE, Gerrits J and MHP Hogenboom (2005) The Repair of Electronic Equipment: The OMRON Case. In: Flapper SDP, van Nunen JAEE and LN van Wassenhove (eds) Managing Closed-Loop Supply Chains. Springer, Berlin, 109–118. 103. Leavitt HJ (1954) A Note on some Experimental Findings about the Meaning of Price. Journal of Business, 27:205–210. 104. Lebreton B and A Tuma (2002) Ein Ansatz zur Bewertung von Remanufacturingstrategien. In: Leopold-Wildburger U, Rendl F and G W¨ ascher (eds) Operations Research Proceedings 2002. Springer, Berlin, 284–289. 105. Lebreton B and A Tuma (2003) Evaluating Component Recycling Strategies. In: Voß S (ed.) Logistik Management: Prozesse, Systeme, Ausbildung. Physica, Heidelberg, 333–347. 106. Lebreton B, Siestrup G and A Tuma (2004) Fallbeispiel Kreislaufwirtschaft in der Lkw-Reifen-Industrie. Working Paper, University of Augsburg. 107. Lebreton B and A Tuma (2005) A Quantitative Approach to Assessing the Proﬁtability of Car and Truck Tire Remanufacturing. International Journal of Production Economics, in press. 108. Lund RT (1977) Making Products Live Longer. Technology Review, 80(1):49–55. 109. Majumder P and H Groenevelt (2001) Competition in Remanufacturing. Production and Operations Management, 10(2):125–141. 110. Marx-Gomez JG (2001) Rechnergest¨ utzte Ans¨atze zur R¨ uckﬂussprognose von Altprodukten zum Remanufacturing. Shaker, Aachen. 111. Maslennikova I and D Foley (2000) Xerox’s Approach to Sustainability. Interfaces, 30(3):226–233. 112. Maslow, AH (1970) Motivation and Personality. 2nd ed., Prentice-Hall, New Jersey.

146

References

113. Matthews HS and DH Matthews (2003) Information Technology Products and the Environment. In: Kuehr R. and E Williams (eds) Computers and the Environment: Understanding and Managing their Impacts. Kluwer, Dordrecht, 17–40. 114. Meadows, DH, Meadows, DL and J Randers (1972) The Limits to Growth. Universe Books, New York. 115. Rohde J, Meyr H and M Wagner (2000) Die Supply Chain Planning Matrix. PPS–Management, 5(1):10–15. 116. Michelin SA (2004) Pneu Laurent Brand Presentation. Link: www.pneulaurent.com/pnltpor/home/home.jsp, last visited: 30th September 2005. 117. Miller T (2000) Hierarchical Operations and Supply Chain Planning. Springer, Berlin. 118. Navin-Chandra D (1994) The Recovery Problem in Product Design. Journal of Engineering Design, 5(1):65–86. ¨ 119. Oetzel U (1997) Okologie und Marketing. In: Stengel M and K W¨ ustner (eds) Einf¨ uhrung in die Umwelt¨ okonomie. Vahlen, Munich, 161–180. 120. Packard V (1960) The Wastemakers. Penguin, New York. 121. Penev KD and AJ de Ron (1995) Disassembly and Recycling of Electronic Consumer Products: An Overview. Technovation, 15(6):363–374. 122. Pepi J (1998) Chelsea Center for Recycling and Economic Development: Plastics Conversion Project. Technical Report 7, August 1998, University of Massachusetts Amherst. 123. Phillips RL (2005) Pricing and Revenue Optimization. Stanford University Books, Stanford. 124. Podratzki H (2003) Environmental Management at Fujitsu Siemens Computers. In: Kuehr R and E Williams (eds.): Computers and the Environment: Understanding and Managing their Impacts. Kluwer, Dordrecht, 140–161. 125. Porter M (1980) Competitive Strategy: Techniques for Analyzing Industries and Competitors. The Free Press, New York. 126. Porter M (1985) Competitive Advantage: Creating and Sustaining Superior Performance. The Free Press, New York. 127. Porter M and C van der Linde (1995) Green and Competitive: Ending the Stalemate. Harvard Business Review, 73(5):120–133. 128. Porter M (1996) What is Strategy? Harvard Business Review, 74(6):61– 78. 129. Rao AR and KB Monroe (1989) The Eﬀect of Price, Brand Name and Store Name on Buyers’ Perceptions of Product Quality: An Integrative Review. Journal of Marketing Research, 26(3):351–358. 130. Rautenstrauch C (1997) Fachkonzept f¨ ur ein integriertes Produktions, Recyclingplanungs- und -steuerungssystem (PRPS-System). Springer, Berlin.

References

147

131. Richter K (1996) Modellierung von kombinierten Wiederverwendungsund Entsorgungsprozessen. In: Wildemann H (ed.) Produktions- und Zuliefernetzwerke. TCW, Munich, 279–292. 132. Ries A and J Trout (1993) The 22 Immutable Laws of Marketing. Harper Business, New York. 133. Rinschede A and KH Wehking (1995) Entsorgungslogistik III: Kreislaufwirtschaft. Erich Schmidt Verlag, Berlin. 134. Robotis A, Bhattacharya S and LN van Wassenhove (2004) The Eﬀect of Remanufacturing on Procurement Decisions for Resellers in Secondary Markets. European Journal of Operational Research, 163:688705. 135. Rogers D and RS Tibben-Lembke (1999) Going Backwards: Reverse Logistics Trends and Practices. Reverse Logistics Executive Council, Reno. 136. Rohde J, Meyr H and M Wagner (2000) Die Supply Chain Planning Matrix. PPS-Management, 5(1):10–15. 137. Romm JJ (1999) Cool Companies: How the Best Businesses Boost Profits and Productivity by Cutting Greenhouse Gas Emissions. Earthscan, London. 138. Savaskan RC, Bhattacharya S and LN van Wassenhove (1999) Channel Choice and Coordination in a Remanufacturing Environment. INSEAD Working Paper 99/14/TM, INSEAD, Fontainebleau. 139. Saied M and GT Velasquez (2003) PCs and Consumers – A Look at Green Demand, Use and Disposal. In: Kuehr R and E Williams (eds) Computers and the Environment: Understanding and Managing their Impacts. Kluwer, Dordrecht, 161–182. 140. Sarkis J (2003) Operations of a Computer Equipment Resource Recovery Facility. In: Kuehr R and E Williams (eds.) Computers and the Environment: Understanding and Managing their Impacts, Kluwer, Dordrecht, 231–257. 141. Schmidt-Bleek F (1998) Das MIPS-Konzept – Faktor 10. Droemer, Munich. 142. Schneeweiss C (2005) Distributed Decision Making. 2nd ed., Springer, Berlin. 143. Schultmann F, Fr¨ ohling M and O Rentz (2002) Demontageplanung und -steuerung mit Enterprise-Resource- und Advanced-Planning-Systemen. Wirtschaftsinformatik 44(6):557–565. 144. Schulz S (2004) Empirische Analyse der Kundenerwartung bei der Altfahrzeugr¨ uckgabe zur Ermittlung der Potentiale fr die BMW Group und die Demontagebetriebe. Master Thesis, University of Augsburg. 145. Seitz M (2005) Wiederaufarbeitung Scheibenbremse: Bewertung einer komponentenbasierten Recyclingstrategie mittels LP-Modellierung. Master Thesis, University of Augsburg. 146. Seitz M, Dlugosch M, Kring F and W Zinsmeister (2003) Ist eine Schliessung der Kreisl¨ aufe bei Computern durch Remanufacturing m¨ oglich? Optimization Technical Report, University of Augsburg.

148

References

147. Siestrup G (1999) Produktkreislaufsysteme: Ein Ansatz zur betriebswirtschaftlichen Bewertung produktintegrierter Umweltschutzstrategien in kreislaufwirtschaftsorientierten Produktionsnetzwerken. Erich Schmidt Verlag, Berlin. 148. Siestrup G and D Haasis (1997) Strategische Planung von Produktkreislaufsystemen. Zeitschrift fr Planung, 8:149–167. 149. Souren R (2002) Konsumg¨ uterverpackungen in der Kreislaufwirtschaft – Stoﬀstr¨ ome, Transformationsprozesse, Transaktionsbeziehungen. DUV, Wiesbaden. 150. Souza GC, Guide Jr VDR, van Wassenhove LN and JD Blackburn (2003) Time Value of Commercial Product Returns. Working Paper 03/48/TM, INSEAD. 151. Spengler T (1994) Industrielle Demontage- und Recyclingkonzepte. Betriebswirtschaftliche Planungsmodelle zur o¨konomisch eﬃzienten Umsetzung abfallrechtlicher R¨ ucknahme- und Verwertungspﬂichten. Erich Schmidt Verlag, Berlin. 152. Spengler T, P¨ uchert H, Penkuhn T and O Rentz (1997) Environmental Integrated Production and Recycling Management. European Journal of Operational Research, 97:308–326. 153. Spengler T and M Schr¨ oter (2003) Strategic Management of Spare Parts in Closed-Loop Supply Chains – a System Dynamics Approach. Interfaces, 33(6):7–17. 154. Spengler T and M Schr¨ oter (2005) Konzeption eines System Dynamics Modells zur strategischen Planung von Closed-Loop Supply Chains – dargestellt am Beispiel der Ersatzteilversorgung. Zeitschrift f¨ ur Betriebswirtschaft, Sp. issue 05(3):1–30. 155. Spengler T, St¨ olting W and M Ploog (2003) Recovery Planning in Closed Loop Supply Chains. In: Dyckoﬀ H, Lackes, R and J Reese (eds) Supply Chain Management and Reverse Logisitics. Springer, Berlin, 61– 89. 156. Stahel WR (1986) Product Life as a Variable: The Notion of Utilization. Science and Public Policy, 13(4):185–193. 157. Stahlknecht P and U Hasenkamp (2002) Einf¨ uhrung in die Wirtschaftsinformatik. 10th ed., Springer, Berlin. 158. Steinhilper R (1998) Remanufacturing, The Ultimate Form of Recycling. Fraunhofer IRB, Stuttgart. 159. Steven M (1994) Hierarchische Produktionsplanung. Physica, Heidelberg. 160. St¨ olzle W (1996) Grundz¨ uge des Outsourcing von Entsorgungsleistungen. Zeitschrift f¨ ur Betriebswirtschaftslehre, Sp. issues 96(2):121–145. 161. Stock JR (1992) Reverse Logistics. Council of Logistics Management, Oak Brook. 162. Teunter RH (2001) A Reverse Logistics Valuation Method for Inventory Control. International Journal of Production Research 39(9):2023–2035.

References

149

163. Teunter RH and L Fortuin (1999) End-of-Life Service. International Journal of Production Economics, 59(3):487–498. 164. Teunter RH and E van der Laan (2004) Valuation of Inventories in Product Recovery Systems. In: Dekker R, Fleischmann M, Inderfurth K and LN van Wassenhove (eds.) Reverse Logistics – Quantitative Models for Closed-Loop Supply Chains. Springer, Berlin, 275–293. 165. Thierry MC, Salomon M, van Nunen J and LN van Wassenhove (1995) Strategic Issues in Product Recovery Management. California Management Review, 37(2):114–135. 166. Thierry MC (1997) An Analysis of the Impact of Product Recovery Management on Manufacturing Companies. PhD Thesis, Erasmus University Rotterdam. 167. Tibben-Lembke RS (1998) The Impact of Reverse Logistics on the Total Costs of Ownership. Journal of Marketing: Theory and Practice, 6(4):51–60. 168. Toktay LB, van der Laan E and M de Brito (2004) Managing Product Returns: The Role of Forecasting. In: Dekker R, Fleischmann M, Inderfurth K and LN van Wassenhove (eds.) Reverse Logistics – Quantitative Models for Closed-Loop Supply Chains. Springer, Berlin, 45–64. 169. Toktay LB and ME Ferguson (2005) The Eﬀect of Competition on Recovery Strategies. Working Paper INSEAD / Georgia Tech University. 170. Toktay LB and D Wei (2005) Cost Allocation in Manufacturing– Remanufacturing Operations. Working Paper, INSEAD. 171. The World Commission on Environment and Development (1987) Our Common Future (a.k.a. Brundtland Report). United Nations, New York. 172. van Wassenhove LN and VDR Guide Jr (2005) The Evolution of ClosedLoop Supply Chain Research. Keynote CLSC Conference, Vanderbilt University, Nashville. 173. von Weizs¨acker EU, Hunter AB and L Hunter-Lovins (1995) Faktor Vier Doppelter Wohlstand – halbierter Naturverbrauch: Der neue Bericht an den Club of Rome. Droemer Knaur Verlag, Munich. 174. Verband der Automobilindustrie (2002) Autojahresbericht 2002. VDA, Frankfurt. 175. Verband Deutscher Ingenieure (2002) Recycling-Oriented Product Development – VDI 2243. VDI, D¨ usseldorf. 176. Westdeutscher Rundfunk (2003) Dosenpfand: Millionen landen im M¨ ull. Link: http://www.wdr.de, last visited: 29th September 2005. 177. European Community (2002) Directive 2002/96/EC on Waste of Electric and Electronic Equipment (WEEE). Oﬃcial Journal of the European Communities, 2002/C 110 E/01. 178. White CD, Masanet E, Meisner-Rosen C and SL Bechman (2003) Product Recovery with Some Byte: An Overview of Management Challenges and Environmental Consequences in Reverse Manufacturing for the Computer Industry. Journal of Cleaner Production, 11:445–458.

150

References

179. Williams E (2003) Environmental Impacts in the Production of Personal Computers. In: Kuehr R and E Williams (eds) Computers and the Environment: Understanding and Managing their Impacts. Kluwer, Dordrecht, 41–72. 180. Williams E and R Kuehr (2003) Todays Markets for Used PCs – And Ways to Enhance Them. In: Kuehr R and E Williams (eds) Computers and the Environment: Understanding and Managing Their Impacts. Kluwer, Dordrecht, 197–210.

List of Figures

1.1

Asset recovery processes: Overview (modiﬁed from White et al. 2003) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5

2.1 2.2 2.3 2.4

Kotler’s product levels (1999) . . . . . . . . . . . . . . . . . . . . . . . . Generic competitive strategies (Porter, 1980) . . . . . . . . . . Value chain according to Porter (1985) . . . . . . . . . . . . . . . . Cost leadership with asset recovery: Value chain . . . . . . . .

14 15 17 23

3.1

Closed-loop accounting model . . . . . . . . . . . . . . . . . . . . . . . . 34

4.1

Proﬁtability yields of core destruction with respect to τ and λ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

5.1

Supply chain planning matrix (modiﬁed from Rohde et al., 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Steps towards closed-loop supply chains: An OEM’s perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Strategic network planning dimensions in closed-loops . . . 5.4 Generic life cycle phases and remanufacturing key factors 5.5 The role of material markets in proprietary closed-loops . 5.6 Generic return distribution . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7 Wearing pattern of durable goods, Lund (1977) . . . . . . . . 5.8 Impact of wearing and technological evolution on the reintegration probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.9 Strategic network planning model: Overview . . . . . . . . . . . 5.10 Recovery path selection: Extended model ﬂows . . . . . . . . . 5.11 Depreciation modelling framework (modiﬁed from Ferber, 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.12 Strategic closed-loop location planning . . . . . . . . . . . . . . . .

68 70 70 72 73 78 79 81 82 88 90 92

152

List of Figures

6.1 6.2 6.3 6.4 6.5 6.6 7.1 7.2

Average material composition of a tire (IFEU, 1999) . . . . 102 Allocation of scrap tires to recovery alternatives (IFEU, 1999) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Tire structure, Continental (2003) . . . . . . . . . . . . . . . . . . . . 104 Optimization results: Contribution margin per tire sold . 111 Tire remanufacturing pipeline . . . . . . . . . . . . . . . . . . . . . . . . 114 Present remanufacturing rate for functional and psycho-sociological goods – (functional goods underlined) 117 Investigation scope: selected PC modules, Stahlknecht et al. (2002) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Computer CLSC assessment: Scenarios overview . . . . . . . . 127

List of Tables

1.1

European and German asset recovery legislation . . . . . . . .

2.1

Energy consumption and CO2 emissions in a computer’s life-cycle, Grote (1994) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Use pattern estimation for personal computers, Williams and Kuehr (2003) . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Closed-loop supply chains and competitive strategy: Case overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

2.2 2.3 3.1 3.2 3.3 4.1 4.2 4.3 4.4

8

Technical parameter settings of the investigated scenarios 39 Leasing and refurbishing of power computers: Aggregated data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Proﬁtability of computer lease remanufacturing by age: First estimations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Competition matrix for laser printers . . . . . . . . . . . . . . . . . Competition matrix for brake systems . . . . . . . . . . . . . . . . . OEM’s competitive options against external remanufacturers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameter settings for toner cartridges . . . . . . . . . . . . . . . .

53 53 55 62

5.1

Intergenerational parts’ compatibility between products γc ,c,p . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

6.1

Resource consumption for the production of a car tire, see Ferrer (1997) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Tire remanufacturing rates, IFEU (1999), Ferrer (1997) . 105 Demand segmentation of tire markets in Germany (BRV, 2000 and 2001) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

6.2 6.3

154

List of Tables

6.4 6.5 6.6

Product market share upper bounds in given segments . . 107 Average retail prices in Germany in Euros (ADAC, 2002) 108 Products’ market share among scenarios in % . . . . . . . . . . 112

7.1 7.2

max (initial scenario) . . . . . . 122 Market share upper bounds βp,q Average component prices (1997-2003), in Euro (Seitz et al., 2003) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 WEEE impact: Optimization results . . . . . . . . . . . . . . . . . . 128 Remanufacturing impact: Optimization results . . . . . . . . . 130 Average service life of WEEE appliances (Hauser and R¨ottchen, 1995, Cooper, 1994) . . . . . . . . . . . . . . . . . . . . . . . 131

7.3 7.4 7.5

Lecture Notes in Economics and Mathematical Systems For information about Vols. 1–495 please contact your bookseller or Springer-Verlag Vol. 496: P. Weiß, Unemployment in Open Economies. XII, 226 pages. 2001.

Vol. 517: H. I. Meinhardt, Cooperative Decision Making in Common Pool Situations. VIII, 205 pages. 2002.

Vol. 497: J. Inkmann, Conditional Moment Estimation of Nonlinear Equation Systems. VIII, 214 pages. 2001.

Vol. 518: S. Napel, Bilateral Bargaining. VIII, 188 pages. 2002.

Vol. 498: M. Reutter, A Macroeconomic Model of West German Unemployment. X, 125 pages. 2001.

Vol. 519: A. Klose, G. Speranza, L. N. Van Wassenhove (Eds.), Quantitative Approaches to Distribution Logistics and Supply Chain Management. XIII, 421 pages. 2002.

Vol. 499: A. Casajus, Focal Points in Framed Games. XI, 131 pages. 2001. Vol. 500: F. Nardini, Technical Progress and Economic Growth. XVII, 191 pages. 2001. Vol. 501: M. Fleischmann, Quantitative Models for Reverse Logistics. XI, 181 pages. 2001. Vol. 502: N. Hadjisavvas, J. E. Martínez-Legaz, J.-P. Penot (Eds.), Generalized Convexity and Generalized Monotonicity. IX, 410 pages. 2001. Vol. 503: A. Kirman, J.-B. Zimmermann (Eds.), Economics with Heterogenous Interacting Agents. VII, 343 pages. 2001. Vol. 504: P.-Y. Moix (Ed.), The Measurement of Market Risk. XI, 272 pages. 2001. Vol. 505: S. Voß, J. R. Daduna (Eds.), Computer-Aided Scheduling of Public Transport. XI, 466 pages. 2001. Vol. 506: B. P. Kellerhals, Financial Pricing Models in Continuous Time and Kalman Filtering. XIV, 247 pages. 2001. Vol. 507: M. Koksalan, S. Zionts, Multiple Criteria Decision Making in the New Millenium. XII, 481 pages. 2001. Vol. 508: K. Neumann, C. Schwindt, J. Zimmermann, Project Scheduling with Time Windows and Scarce Resources. XI, 335 pages. 2002. Vol. 509: D. Hornung, Investment, R&D, and Long-Run Growth. XVI, 194 pages. 2002. Vol. 510: A. S. Tangian, Constructing and Applying Objective Functions. XII, 582 pages. 2002. Vol. 511: M. Külpmann, Stock Market Overreaction and Fundamental Valuation. IX, 198 pages. 2002. Vol. 512: W.-B. Zhang, An Economic Theory of Cities.XI, 220 pages. 2002. Vol. 513: K. Marti, Stochastic Optimization Techniques. VIII, 364 pages. 2002.

Vol. 520: B. Glaser, Efﬁciency versus Sustainability in Dynamic Decision Making. IX, 252 pages. 2002. Vol. 521: R. Cowan, N. Jonard (Eds.), Heterogenous Agents, Interactions and Economic Performance. XIV, 339 pages. 2003. Vol. 522: C. Neff, Corporate Finance, Innovation, and Strategic Competition. IX, 218 pages. 2003. Vol. 523: W.-B. Zhang, A Theory of Interregional Dynamics. XI, 231 pages. 2003. Vol. 524: M. Frölich, Programme Evaluation and Treatment Choise. VIII, 191 pages. 2003. Vol. 525: S. Spinler, Capacity Reservation for Capital- Intensive Technologies. XVI, 139 pages. 2003. Vol. 526: C. F. Daganzo, A Theory of Supply Chains. VIII, 123 pages. 2003. Vol. 527: C. E. Metz, Information Dissemination in Currency Crises. XI, 231 pages. 2003. Vol. 528: R. Stolletz, Performance Analysis and Optimization of Inbound Call Centers. X, 219 pages. 2003. Vol. 529: W. Krabs, S. W. Pickl, Analysis, Controllability and Optimization of Time-Discrete Systems and Dynamical Games. XII, 187 pages. 2003. Vol. 530: R. Wapler, Unemployment, Market Structure and Growth. XXVII, 207 pages. 2003. Vol. 531: M. Gallegati, A. Kirman, M. Marsili (Eds.), The Complex Dynamics of Economic Interaction. XV, 402 pages, 2004. Vol. 532: K. Marti, Y. Ermoliev, G. Pﬂug (Eds.), Dynamic Stochastic Optimization. VIII, 336 pages. 2004. Vol. 533: G. Dudek, Collaborative Planning in Supply Chains. X, 234 pages. 2004. Vol. 534: M. Runkel, Environmental and Resource Policy for Consumer Durables. X, 197 pages. 2004.

Vol. 514: S. Wang, Y. Xia, Portfolio and Asset Pricing. XII, 200 pages. 2002.

Vol. 535: X. Gandibleux, M. Sevaux, K. Sörensen, V. Tkindt (Eds.), Metaheuristics for Multiobjective Optimisation. IX, 249 pages. 2004.

Vol. 515: G. Heisig, Planning Stability in Material Requirements Planning System. XII, 264 pages. 2002.

Vol. 536: R. Brüggemann, Model Reduction Methods for Vector Autoregressive Processes. X, 218 pages. 2004.

Vol. 516: B. Schmid, Pricing Credit Linked Financial Instruments. X, 246 pages. 2002.

Vol. 537: A. Esser, Pricing in (In)Complete Markets. XI, 122 pages, 2004.

Vol. 538: S. Kokot, The Econometrics of Sequential Trade Models. XI, 193 pages. 2004.

Vol. 565: W. Lemke, Term Structure Modeling and Estimation in a State Space Framework. IX, 224 pages. 2006.

Vol. 539: N. Hautsch, Modelling Irregularly Spaced Financial Data. XII, 291 pages. 2004.

Vol. 566: M. Genser, A Structural Framework for the Pricing of Corporate Securities. XIX, 176 pages. 2006.

Vol. 540: H. Kraft, Optimal Portfolios with Stochastic Interest Rates and Defaultable Assets. X, 173 pages. 2004.

Vol. 567: A. Namatame, T. Kaizouji, Y. Aruga (Eds.), The Complex Networks of Economic Interactions. XI, 343 pages. 2006.

Vol. 541: G.-y. Chen, X. Huang, X. Yang, Vector Optimization. X, 306 pages. 2005. Vol. 542: J. Lingens, Union Wage Bargaining and Economic Growth. XIII, 199 pages. 2004.

Vol. 568: M. Caliendo, Microeconometric Evaluation of Labour Market Policies. XVII, 258 pages. 2006.

Vol. 543: C. Benkert, Default Risk in Bond and Credit Derivatives Markets. IX, 135 pages. 2004.

Vol. 569: L. Neubecker, Strategic Competition in Oligopolies with Fluctuating Demand. IX, 233 pages. 2006.

Vol. 544: B. Fleischmann, A. Klose, Distribution Logistics. X, 284 pages. 2004.

Vol. 570: J. Woo, The Political Economy of Fiscal Policy. X, 169 pages. 2006.

Vol. 545: R. Hafner, Stochastic Implied Volatility. XI, 229 pages. 2004.

Vol. 571: T. Herwig, Market-Conform Valuation of Options. VIII, 104 pages. 2006.

Vol. 546: D. Quadt, Lot-Sizing and Scheduling for Flexible Flow Lines. XVIII, 227 pages. 2004.

Vol. 572: M. F. Jäkel, Pensionomics. XII, 316 pages. 2006

Vol. 547: M. Wildi, Signal Extraction. XI, 279 pages. 2005.

Vol. 573: J. Emami Namini, International Trade and Multinational Activity, X, 159 pages, 2006.

Vol. 548: D. Kuhn, Generalized Bounds for Convex Multistage Stochastic Programs. XI, 190 pages. 2005.

Vol. 574: R. Kleber, Dynamic Inventory Management in Reverse Logisnes, XII, 181 pages, 2006.

Vol. 549: G. N. Krieg, Kanban-Controlled Manufacturing Systems. IX, 236 pages. 2005.

Vol. 575: R. Hellermann, Capacity Options for Revenue Management, XV, 199 pages, 2006.

Vol. 550: T. Lux, S. Reitz, E. Samanidou, Nonlinear Dynamics and Heterogeneous Interacting Agents. XIII, 327 pages. 2005.

Vol. 576: J. Zajac, Economics Dynamics, Information and Equilibnum, X, 284 pages, 2006.

Vol. 551: J. Leskow, M. Puchet Anyul, L. F. Punzo, New Tools of Economic Dynamics. XIX, 392 pages. 2005. Vol. 552: C. Suerie, Time Continuity in Discrete Time Models. XVIII, 229 pages. 2005. Vol. 553: B. Mönch, Strategic Trading in Illiquid Markets. XIII, 116 pages. 2005. Vol. 554: R. Foellmi, Consumption Structure and Macroeconomics. IX, 152 pages. 2005. Vol. 555: J. Wenzelburger, Learning in Economic Systems with Expectations Feedback (planned) 2005. Vol. 556: R. Branzei, D. Dimitrov, S. Tijs, Models in Cooperative Game Theory. VIII, 135 pages. 2005. Vol. 557: S. Barbaro, Equity and Efﬁciency Considerations of Public Higer Education. XII, 128 pages. 2005. Vol. 558: M. Faliva, M. G. Zoia, Topics in Dynamic Model Analysis. X, 144 pages. 2005. Vol. 559: M. Schulmerich, Real Options Valuation. XVI, 357 pages. 2005.

Vol. 577: K. Rudolph, Bargaining Power Effects in Financial Contracting, XVIII, 330 pages, 2006. Vol. 578: J. Kühn, Optimal Risk-Return Trade-Offs of Commercial Banks, IX, 149 pages, 2006. Vol. 579: D. Sondermann, Introduction to Stochastic Calculus for Finance, X, 136 pages, 2006. Vol. 580: S. Seifert, Posted Price Offers in Internet Auction Markets, IX, 186 pages, 2006. Vol. 581: K. Marti; Y. Ermoliev; M. Makowsk; G. Pﬂug (Eds.), Coping with Uncertainty, XIII, 330 pages, 2006. Vol. 582: J. Andritzky, Sovereign Default Risks Valuation: Implications of Debt Crises and Bond Restructurings. VIII, 251 pages, 2006. Vol. 583: I.V. Konnov, D.T. Luc, A.M. Rubinov (Eds.), Generalized Convexity and Related Topics, IX, 469 pages, 2006. Vol. 584: C. Bruun, Adances in Artiﬁcial Economics: The Economy as a Complex Dynamic System. XVI, 296 pages, 2006.

Vol. 560: A. von Schemde, Index and Stability in Bimatrix Games. X, 151 pages. 2005.

Vol. 585: R. Pope, J. Leitner, U. Leopold-Wildburger, The Knowledge Ahead Approach to Risk, XVI, 218 pages, 2007.

Vol. 561: H. Bobzin, Principles of Network Economics. XX, 390 pages. 2006.

Vol. 586: B. Lebreton, Strategic Closed-Loop Supply Chain Management. X, 150 pages, 2007.

Vol. 562: T. Langenberg, Standardization and Expectations. IX, 132 pages. 2006.

Vol. 587: P.N. Baecker, Real Options and Intellectual Property: Capital Budgeting Under Imperfect Patent Protection. X, 276 pages, 2007 (planned).

Vol. 563: A. Seeger (Ed.), Recent Advances in Optimization. XI, 455 pages. 2006. Vol. 564: P. Mathieu, B. Beauﬁls, O. Brandouy (Eds.), Artiﬁcial Economics. XIII, 237 pages. 2005.

Vol. 588: D. Grundel, M. Robert, P. Panos, O. Prokopyev (Eds.), Cooperative Systems: Control and Optimization. IX, 401 pages, 2007 (planned).

Strategic Supply Chain Management

Supply Chain Coordination Mechanisms: New Approaches for Collaborative Planning (Lecture Notes in Economics and Mathematical Systems)

Standardization and Expectations (Lecture Notes in Economics and Mathematical Systems)

Strategic Supply Chain Management in Process Industries: An Application to Specialty Chemicals Production Network Design (Lecture Notes in Economics and Mathematical Systems)

Strategic Closed-Loop Supply Chain Management (Lecture Notes in Economics and Mathematical Systems)

Strategic Supply Chain Management

Supply Chain Coordination Mechanisms: New Approaches for Collaborative Planning (Lecture Notes in Economics and Mathematical Systems)

Standardization and Expectations (Lecture Notes in Economics and Mathematical Systems)

Bond Portfolio Optimization (Lecture Notes in Economics and Mathematical Systems)

Cooperative Lot Sizing Games in Supply Chains (Lecture Notes in Economics and Mathematical Systems 644)

Total Supply Chain Management

Supply Chain Logistics Management

Supply Chain Logistics Management

Supply Chain Project Management

Supply Chain Logistics Management

Purchasing and Supply Chain Management

Strategic Competition in Oligopolies with Fluctuating Demand (Lecture Notes in Economics and Mathematical Systems)

Strategic Supply Chain Management in Process Industries: An Application to Specialty Chemicals Production Network Design (Lecture Notes in Economics and Mathematical Systems)

Supply Chain Logistics Management

Supply chain management

Supply Chain Management

Supply Chain Logistics Management

Purchasing and Supply Chain Management

Purchasing and Supply Chain Management

Supply Chain Management

Food Supply Chain Management

Supply Chain Management

Supply Chain Coordination Mechanisms: New Approaches for Collaborative Planning (Lecture Notes in Economics and Mathematical Systems 628)

Artificial Economics: The Generative Method in Economics (Lecture Notes in Economics and Mathematical Systems)

Supply Chain Design and Management: Strategic and Tactical Perspectives

Production and Inventory Management with Substitutions (Lecture Notes in Economics and Mathematical Systems)

Handbook of Supply Chain Management

Handbook of Supply Chain Management

Global Integrated Supply Chain Systems

Essentials of Supply Chain Management

Strategic Closed-Loop Supply Chain Management (Lecture Notes in Economics and Mathematical Systems)

Strategic Supply Chain Management

Supply Chain Coordination Mechanisms: New Approaches for Collaborative Planning (Lecture Notes in Economics and Mathematical Systems)

Standardization and Expectations (Lecture Notes in Economics and Mathematical Systems)

Bond Portfolio Optimization (Lecture Notes in Economics and Mathematical Systems)

Cooperative Lot Sizing Games in Supply Chains (Lecture Notes in Economics and Mathematical Systems 644)

Total Supply Chain Management

Supply Chain Logistics Management

Supply Chain Logistics Management

Supply Chain Project Management

Supply Chain Logistics Management

Purchasing and Supply Chain Management

Strategic Competition in Oligopolies with Fluctuating Demand (Lecture Notes in Economics and Mathematical Systems)

Strategic Supply Chain Management in Process Industries: An Application to Specialty Chemicals Production Network Design (Lecture Notes in Economics and Mathematical Systems)

Supply Chain Logistics Management

Supply chain management

Supply Chain Management

Supply Chain Logistics Management

Purchasing and Supply Chain Management

Purchasing and Supply Chain Management

Supply Chain Management

Food Supply Chain Management

Supply Chain Management

Supply Chain Coordination Mechanisms: New Approaches for Collaborative Planning (Lecture Notes in Economics and Mathematical Systems 628)

Artificial Economics: The Generative Method in Economics (Lecture Notes in Economics and Mathematical Systems)

Supply Chain Design and Management: Strategic and Tactical Perspectives

Production and Inventory Management with Substitutions (Lecture Notes in Economics and Mathematical Systems)

Handbook of Supply Chain Management

Handbook of Supply Chain Management

Global Integrated Supply Chain Systems

Essentials of Supply Chain Management

Recommend Documents