MIKADO : a decision support tool for pollution reduction in aluminium pressure die casting

MIKADO: a Decision Support Tool for Pollution Reduction in Aluminium Pressure Die Casting Belmira Neto Promotor: Prof. Dr. L. Hordijk Hoogleraar Milieusysteemanalyse, Wageningen Universiteit Co-promotoren: Dr. C. Kroeze Universitair Hoofddocent bij de leerstoelgroep Milieusysteemanalyse, Wageningen Universiteit Prof. Dr. C. A.V. Costa Professor of Chemical Engineering, Faculty of Engineering University of Porto, Portugal Samenstelling promotiecommissie: Prof. Dr. Ir. W. H. Rulkens, Wageningen Universiteit Prof. Dr. H.C. Moll, Rijksuniversiteit Groningen Prof. Dr. L.F. Malheiros, Porto University, Portugal Prof. Dr. Ir. C.S.A. van Koppen, Wageningen Universiteit / Universiteit Utrecht Dit onderzoek is uitgevoerd binnen de onderzoekschool SENSE. MIKADO: a Decision Support Tool for Pollution Reduction in Aluminium Pressure Die Casting Belmira Neto Proefschrift ter verkrijging van de graad van doctor op gezag van de rector magnificus van Wageningen Universiteit, Prof. Dr. M. J. Kropff, in het openbaar te verdedigen op maandag 2 april 2007 des namiddags te half twee in de Aula Neto, B., 2007 MIKADO: a Decision Support Tool for Pollution Reduction in Aluminium Pressure Die Casting PhD thesis Wageningen University, with summaries in English and Dutch ISBN 90-8504-619-X Acknowledgment I can still remember the unease before starting my research project. Several questions relating to the country in which to carry out the research and the contents of the project had to be dealt with. In that time many people were involved in keeping me on track by helping me out to apply for a post-graduate research scholarship, to document myself with appropriate literature and in deciding where to carry out my research. To all of them I express here my gratitude. In early 2002, I started the work presented in this thesis. I arrived in February that year at the Environmental Systems Analysis Group at the University of Wageningen and realised that all the ideas I had for the work seemed possible to put into practice and made sense also to others. My promotion team were primarily responsible for me to put into practice what I was aiming to do. I express here my gratitude to my promotor – Leen Hordijk. It is certainly a challenge to be guided by the Director of the International Institute for Applied Systems Analysis (IIASA) in Austria. But even with a busy schedule, when needed he took over the daily supervision and was available to meet me regularly during my stay at IIASA (Oct-Dec. 2003). It was an excellent experience to work with Leen. His very critical comments but equally sound advice were vital many times throughout the research. I also would like to thank my copromotor - Carlos Costa. I did not have many chances to discuss with him. Nevertheless, in our discussions his advice was always very constructive. Lastly, my daily supervisor – Carolien Kroeze. She played a primary role throughout the research period. She is one of the most efficient persons that I have ever met. From her I learnt how to plan my work efficiently, to define priorities and also I learnt the basis of working in science. She was the best daily supervisor one can wish and she definitely had an important role given the endurance of the research period. Many thanks also to the industrial plant managers from the case study. This thesis wouldn’t be possible without their involvement. First of all, I gratefully acknlowledge the Plant Director. He agreed to provide data from the plant to be used in the thesis. I also express thanks to the Environmental Manager, who assisted in the collection of data for the inventory and to the Production Manager who answered multiple process related questions and proposed some of the alternatives to reduce the environmental pollution in the plant. Furthermore, I would like to thank the other directors and plant workers involved in the collection of plant data and in the pleasant lunch gatherings that took place in the nice Portuguese district where the case company is located. I also thank the colleagues from the Department of Metallurgical and Materials Engineering at the Engineering Faculty of Porto University. The last period of the PhD research (from Feb. 2004 onwards) was spent back at the department, but several trips to the Netherlands took place during this last period. The Department colleagues always encouraged these trips; therefore also deserve a share of the credit for the successful accomplishment of this thesis. In addition, I thank the colleagues from the Environmental Systems Analysis (ESA) group. Together, we moved twice in between three distinct buildings at the University of Wageningen. We had together several table lunches, walks and coffee breaks. My special thanks go to Ria Cuperus. She was always efficient in whatever arrangements were needed. Thanks Ria and Olivia for arranging the conference calls. The gezelligheid in the ESA group always made me feel at home. My thanks also go to the colleagues of the former Emission and Assessment Group at TNO. It was a pleasure to share with some of them my first impressions about living in the Netherlands. They also provide me some amusement in the form of music audio files. During this time many friendships relations were made. The Netherlands is a wonderful country in also promoting multi-cultural social gatherings. In these last four years I lived in more than ten different places in at least three different European countries. So, it is quite impressive the number of people from all continents that crossed my life recently. However, a core group was kept and to name them is possible. In alphabetical order, I thank them for joining my journey at some point of the PhD: Ana, Angels, Angel, Beto, Cora, Evelien, Esther, Hans, Gijs, Goran, Joachim, John, Laurence, Mahbod, Manuela, Marian, Piet, Peter M., Peter T., Sónia, Silvia, Stasa, Tamara, Tiago, Wil and Willem. Estes últimos anos foram uma mistura excelente entre experimentar a emigração, viajar, trabalhar arduamente, leccionar e divertir-me. Foi sem dúvida uma aventura extremamente gratificante. Embora trabalhar em investigação fora do contexto nacional não seja fácil, pode sem dúvida, constituir uma excelente escola. Recomendo vivamente! A par dos altos e baixos desta aventura estiveram sempre presentes os meus pais. Obrigada aos dois pelo acompanhamento neste longo processo. Table of contents Chapter 1 Introduction 1 Chapter 2 Selecting Environmental Systems Analysis Tools: strengths and weaknesses for use in a decision support tool 13 Chapter 3 Inventory of Pollution Reduction Options for an Aluminium Pressure Die Casting Plant 31 Chapter 4 Modelling the Environmental Impact of an Aluminium Pressure Die Casting Plant and Options for Control 55 Chapter 5 Strategies to Reduce the Environmental Impact of an Aluminium Pressure Die Casting Plant: a Scenario Analysis 97 Chapter 6 Conclusions and Discussion 127 Summary 147 Samenvatting 151 Curriculum Vitae 155 - 1 - Chapter 1: Introduction 1.1. Background 1.1.1. Industry and the environment Industrial activities cause a variety of environmental problems. These problems receive attention through environmental policies aimed at limiting pollution of air, water and soil and at enhancing conservation of resources and nature. For many industrial companies, environmental performance in terms of emissions, production of waste and the use of resources is an increasing concern. Assessing this performance is not a simple task, because of the complexity of industrial processes and the complexity of the environmental issues given the variety of the compounds emitted and the variety of their environmental effects. The efforts to improve environmental performance of the industry have traditionally been driven by environmental regulations. Environmental laws, applicable to the industrial sectors, often limit the emissions of specific pollutants. Companies typically respond to these regulations by taking single actions aiming to live up to the environmental restrictions, by for instance, reducing the amount of a compound emitted. Alternatively, companies may define internal environmental policies, for instance, by implementing environmental management systems (e.g. standards as ISO 14001, 2004 or EMAS, 2001). Such pro-activeness may be implicitly driven by environmental restrictions, but requires a more integrated approach in defining ways to reduce the total environmental burden of a company. The attempts to improve the environmental performance vary between different types of industry. Industrial sectors taking the lead in this include, for instance, the pulp and paper sector (e.g. Lee and Ding, 2000; Pineda-Henson et al., 2002; Bordado and Gomes, 2003; Lopes et al., 2003; Oral et al., 2005; Lee and Rhee, 2005; Mahmood and Elliot, 2006; Gabbrielli et al., 2006), chemical industries (e.g. Eder, 2003; Alvarez et al., 2004; Smith et al., 2004; Seyler et al., 2005; Mendivil et al., 2005; Kleizen, 2006) and the metals industries. The metals industry is one of the most studied industrial sectors. A large number of environmental studies about the metals industry have been published (e.g. Proctor et al., 2000; Moors et al., 2005; Tan and Khoo, 2005a; Rebitzer and Buxmann, 2005; Moors, 2006; Norgate et al., 2007). These studies focus predominantly on the primary and secondary metals industry. Metals casting industry is an exception in this respect: the number of studies on the environmental performance of the metal casting is limited. Nevertheless, this industry is dominated by relatively small businesses supplying the largest share of casting products currently used worldwide. Within the metals casting industry a distinction can be made between ferrous and non-ferrous metals industries. The aluminium pressure die casting industry belongs to the second category and will be subject of our analysis. Chapter 1: Introduction - 8 - Step 3 (Identification of pollution reduction options) includes the construction of an inventory of reduction options, their potential to reduce emissions, and the associated costs. The reduction options to be analysed are specific for the selected plant. In this step, a general overview will be given of the pollution reduction options aiming at reducing the emissions to air, soil and water from an aluminium pressure die casting plant. The options will be investigated in terms of their potential to reduce pollution and also in terms of the costs associated with their implementation. The options to be developed for the selected plant are process specific and assumed to be implemented at the level of the plant sub-processes or sub-sub-processes. These options may include add-on techniques or be more structural, i.e., by affecting the materials consumption or changes in process operations. They focus either on the different pollutants released by the processes or on the materials/energy consumed in the processes. Step 4 (Model building) aims at exploring the consequences for the environmental impact and associated costs of individual pollution reduction options or combinations thereof. The model building is followed by an analysis of the model sensitivity to changes in model parameters. In this step a model (our DST) is developed to analyse options to reduce the environmental impact of aluminium die casting. This model takes a company perspective, so that it can be used as a decision supporting tool for environmental management. The model structure and the modelling approach are based on a study by Van Langen (2002), who describes the development of a definition language for designing processes (DESIRE). This language provides a structure and a grammar to define objects, objects’ properties and methods. Van Langen stated that his approach can be used in designing models for estimating the emissions from industrial processes. Step 5 (Model application) uses the model to explore the implementation of individual reduction options or combinations of options in well defined reduction strategies. In this step the model is explored. Three different types of analyses are made. At first, we analyse the plant’s environmental performance without implementing pollution reduction options. The analysis focuses on the relative contribution of different industrial processes levels to the environmental impact. Second, the individual pollution reduction options are analysed systematically by calculating their potential to reduce environmental problems and the cost associated with the reduction. Third, in order to analyse the situation in which a company decides to implement a number of options simultaneously, different strategies to combine reduction options are defined. These reduction strategies may, for instance, aim for reducing the largest environmental problem, or a specific activity, or a specific pollutant. Alternatively, a company may wish to combine the most cost effective options, or combining add-on techniques, or only more structural reduction options. Therefore, a range of combinations are presented, and their effects on the plant’s environmental performance analysed. The associated costs resulting from the implementation of these options are also analysed. In the final Step 6 (Evaluation of the methodological approach) the environmental systems analysis approach will be discussed in terms of iterations performed, sequence of steps and the comparison with other model studies. In this step, model uncertainties and the implication of the results of our study to the aluminium pressure die casting sector as well for other metals industry or the industry in general are also discussed. This may reveal the applicability of the approach for other industries or sectors. Chapter 1: Introduction - 9 - In this study an environmental systems analysis is performed at the plant level, using a specific combination of ESA tools. This combination aims to fulfil the current gap in decision support tools that provides companies with means to analysing options to reduce their environmental impact by defining pollution reduction options and by assessing the economic and environmental benefits of these options. 1.4. Thesis outline The thesis includes six chapters presenting the results of six steps of the environmental systems analysis procedure according to the formulated research questions. (Figure 1.1.) Figure 1.1. Schematic representation of research questions, environmental system analysis (ESA) steps and thesis chapters. This first chapter (Chapter 1) provides the general introduction, and describes the objective, the research questions addressed and the research strategy. Chapter 2 presents an overview of the different analytical tools aiming to assess the environmental performance in the industry. Thus, a selection of promising tools illustrates the need for a new DST that takes a company perspective. The literature Step 6: Evaluation of the methodological a pp roach RQ 1 RQ 2 RQ 3 RQ 4 Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Step 1: Problem definition Step 2: Evaluation and selection of existin g ESA tools Step 3: Identification of pollution reduction options Step 4: Model b uildin g Step 5: Model a pp lication ESA steps Chapters Research Questions Chapter 1: Introduction - 10 - review allows for defining the main characteristics of the decision support tool and leads to potential ESA tools useful for a DST taking a company perspective. Chapter 3 gives a general overview of pollution reduction options aiming to reduce the emissions to air, soil and water of the aluminium pressure die casting plant. The techniques are investigated in terms of their potential to reduce pollution and also in terms of the costs associated with the implementation of these options. These options focus either on the different pollutants released or on the materials/energy used in the process. Possible actions or alternatives that appear to lead to an improvement of the current situation are identified and (partially) presented in Chapter 3. This preliminary analysis of the alternatives is further explored in Chapter 5. Chapter 4 describes the model developed to analyse the pollution reduction options in order to reduce the environmental impact. This chapter describes the mathematical formulation of the model. The model is developed for and applied to the aluminium die casting plant supplying car manufacturers with aluminium die casting products. A first assessment of the environmental impact for the plant is made and results of a partial model sensitivity analysis are presented. Chapter 5 explores the model in order to analyse scenarios to reduce the environmental impact of the aluminium die casting plant. These scenarios present the modelled responses to the reduction options assumed to be implemented. The model calculates the potential to reduce emissions, and the costs associated with implementation of reduction options. First, the model results are presented for a situation in which no reduction options are assumed to be implemented (so-called zero case, reflecting the current practice in the plant). Secondly, a systematic analysis of reduction options is performed. Finally, seven types of reduction strategies are analysed by assuming to implement, simultaneously, different reduction options. These strategies are analysed with respect to their potential to reduce emissions, environmental impact and costs associated with the implementation of options. Finally, the results and methodology are discussed and conclusions drawn. Chapter 6 includes a discussion of the stepwise procedure taken and compares our decision support tool with other model studies. It discusses the model uncertainties and the implications of the results for industry. Finally, recommendations for future studies are formulated. Thus Chapter 6 not only concludes on the results for the case study, but also discusses the extent to which these results can be generalised to other industries. References Alvarez, D., Garrido, N., Sans, R., Carreras, I., 2004. Minimization-optimization of water use in the process of cleaning reactors and containers in a chemical industry. Journal of Cleaner Production 12 (7) 781-787. Backhouse, C.J., Clegg, A.J., Staikos, T., 2004. Reducing the environmental impacts of metal castings through life-cycle management. Progress in Industrial Ecology Vol. 1, Nos. 1/2/3. Bordado, J.C.M., Gomes, J.F.P., 2003. Emission and odour control in Kraft pulp mills. Journal of Cleaner Production 11 (7) 797-801. Chapter 1: Introduction - 11 - Brown, J.R., 1999. Foseco Non-Ferrous Foundryman´s Handbook. In: Brown, J.R., (Eds.), Butterworth-Heinemann, Oxford. CAEF (The European Foundry Association), 2003. Statistics from the CAEF countries members on the production of ferrous and non-ferrous alloys. Data from 1999 to 2003. In: Revista de Fundição 234 Asssociação Portuguesa de Fundição, (Eds.), Portugal (in Portuguese). The statistics from the non-ferrous metals castings industry are also available from http://www.caefeurofoundry.org/library/documents/T28-non-ferrous_37317.pdf, last acceded on 07.11.2006. Checkland, P.B., 1979. Techniques in soft systems practice. Building conceptual models. Journal of Applied Systems Analysis 6 41-49. Choi, A.C.K., Kaebernick, H., Lai, W.H., 1997. Manufacturing processes modelling for environmental impact assessment. Journal of Materials Processing Technology 70 231-238. Dalquist, S., Gutowski, T., 2004. Life cycle analysis of conventional manufacturing techniques: die casting. Available from Web.mit.edu/ebm/publications, last acceded on 8.5.2006. Eder, P., 2003. Expert inquiry on innovation options for cleaner production in the chemical industry. Journal of Cleaner Production 11 (4) 347-364. EMAS, 2001. Regulation (EC) No 761/2001 of the European parliament and of the council of 19 March 2001 allowing voluntary participation by organisations in a Community ecomanagement and audit scheme (EMAS). EIPPCB (European Integrated Pollution Prevention and Control Bureau), 2005. Draft on Reference document on Best Available Techniques in the Smitheries and Foundries Industry, European Commission. Available from http://eippcb.jrc.es/, last acceded on 10.02.2005. Findeisen, W., Quade, E. 1997. The Methodology of Systems Analysis: An Introduction and Overview, in: Miser, H., Quade, E. (eds.) Handbook of Systems Analysis, Wiley, New York. Gabbrielli, R., Medeot, C., Miconi, D., 2006. Energy saving in the tissue industry: technical and economic aspects of a case study. Journal of Cleraner Production 14 (2) 185-193. ISO, 2004. (ISO 14001). Environmental Management Systems – Requirements with Guidance for use. International Organisation for Standardisation. Kim, S.K., Cho, H., Jo, H., Han, M., Lim, S., Hur, T., 2003. Material Life Cycle Assessment for Diecasting Process. Materials Science Forum vols 462-432 3353-58. Kleizen, H.H., 2006. Conceptual chemical process design in a sustainable technological world. Journal of Cleaner Production 14 (9-11) 924-927. Lee, S.Y, Rhee, S.K., 2005. From end-of-pipe technology towards pollution preventive approach: the evolution of corporate environmentalism in Korea. Journal of Cleaner Production 13 (4) 387- 395. Lee, Y.M., Ding, C.Y., 2000. Life-cycle assessment and production policy: An application to corrugated paperboard manufacture. Journal of Environmental Management 59 157–165. Lopes, E., Dias, A., Arroja, L, Capela, I., Pereira, F., 2003. Application of life cycle assessment to the Portuguese pulp and paper industry. Journal of Cleaner Production 11 (1) 51-59. Mahmood, T., Elliott, A., 2006. A review of secondary sludge reduction technologies for the pulp and paper industry. Water Research 40 (11) 2093-2112. Mendivil, R., Fischer, U., Hungerbuhler, K., 2005. Impact of technological development, market and environmental regulations on the past and future performance of chemical processes. Journal of Cleaner Production 13 (9) 869-880. Moors, E.H.M., Mulder, K.F., Vergragt, P.J., 2005. Towards cleaner production: barriers and strategies in the base metals producing industry. Journal of Cleaner Production 13 657-668. Chapter 1: Introduction - 12 - Moors, E.H.M., 2006. Technology strategies for sustainable metals production systems: a case study of primary aluminium production in The Netherlands and Norway. Journal of Cleaner Production 14 1121-1138. Oral, J., Sikula, J., Puchyr, R., Hajny, Z., Stehlik, P., Bebar, L., 2005. Processing of waste from pulp and paper plant. Journal of Cleaner Production 13 (5) 509-515. Park, E., Enander, R., Barnett, S.M., 2002. Pollution prevention in a zinc die casting company: a 10-year case study. Journal of Cleaner Production 10 93-99. Pineda-Henson, R., Culaba, A.B., Mendoza, G.A., 2002. Evaluating Environmental Performance of Pulp and Paper Manufacturing Using Analytical Hierarchy process and Life-Cycle Assessment. Journal of Industrial Ecology Volume 6 Number 1. Pluimers, J., 2001. An environmental systems analysis of greenhouse horticulture in the Netherlands, PhD thesis Wageningen University, The Netherlands. Proctor, D.M., Fehling, K.A., Shay, E.C., Wittenborn, J.L., Green, J.J., Avent, C., Bigham, R.D., Connolly, M., Lee, B., Shepker, T.O., Zak, M.A., 2000. Physical and chemical characteristics of blast furnace, basic oxygen furnace, and electric arc furnace steel industry slags. Environmnetal Science & Technology 34 (8) 1576-1582. Rabah, M., 1999. Cost effectiveness of abatement options for emission control in Egyptian iron foundries. Waste management 19 283-292. Rebitzer, G., Buxmann K., 2005. The role and implementation of LCA within life cycle management at Alcan. Journal of Cleaner Production 13 1327-1335. Seyler, C., Hofstetter, T.B., Hungerbuhler, K., 2005. Life cycle inventory for thermal treatment of waste solvent from chemical industry: a multi-input allocation model. Journal of Cleaner production 13 (13-14) 1211-1224. Smith, R.L., Mata, T.M., Young, D.M., Cabezas, H., Costa, C.A.V., 2004. Designing environmentally friendly chemical processes with fugitive and open emissions. Journal of Cleaner Production 12 (2) 125-129 Tan, R.B.H., Khoo, H.H., 2005a. An LCA study of a primary aluminium supply chain. Journal of Cleaner Production 13 6 607-618. Tan, RB.H, Khoo, H.H., 2005b. Zinc Casting and Recycling. International Journal of LCA 10 (3) 211-18. US Department of Energy, 1998. Metalcasting industry technology roadmap. Project sponsored by Cast Metal Coalition of the American Foundrymen's Society North American Die Casting Association. Steel Founders' Society of America. US Department of Energy, 2004. Metal casting industry of the future. Fiscal Year 2004 Annual report. In: Industrial Technologies Program – Boosting the productivity and competitiveness of U.S. industry trough improvements in energy and environmental performance. U.S. Department of Energy – Energy Efficiency and Renewable Energy. Van Langen, P.H.G., 2002. The Anatomy of Design: Foundations, Models and Applications, PhD thesis, Vrije University Amsterdam, The Netherlands. Wilson, B., 1984. Systems: concepts, methodologies and application. John Wiley & Sons, Chichester, United Kingdom. - 13 - Chapter 2: Selecting Environmental Systems Analysis Tools: strengths and weaknesses for use in a decision support tool Abstract An overview of selected environmental systems analysis (ESA) tools currently used by industry is given, including tools to assess the environmental performance of a company. These tools may be useful for a decision support tool (DST) that takes a company perspective, while considering environmental and economic aspects on the decision-making process. We define criteria for a first selection of ESA tools. The criteria are related to the usefulness of a tool in an analysis that: 1) takes a company perspective; 2) includes environmental and economic aspects of decision making; 3) includes a complete coverage of the potential environmental impacts and 4) allows for an assessment of the consequences of pollution reduction strategies. Based on the purpose of our DST together with the criteria we identified twelve tools. These twelve tools are reviewed with respect to their purpose, methodology, final product, strengths, weaknesses and relevance for an environmental analysis taking a company perspective. Next, we present the characteristics of the DST to be developed. These characteristics allow for identifying the ESA tools that are a promising basis for the DST to be developed. These seven characteristics are: (i) the tool considers a gate-to-gate approach; (ii) the tool considers the processes within the company that are relevant for the assessment of the environmental impact; (iii) the tool uses company specific data easily available from the process owner; (iv) the tool considers up-to-date and company specific pollution reduction options; (v) the tool provides information on the costeffectiveness of the reduction options; (vi) the tool can be used to express the company’s environmental performance in one overall environmental indicator; and lastly (vii) the tool can be used to explore possible user-defined pollution reduction strategies. Finally, a selection of the tools that are useful for our particular DST is made. We conclude that a combination of the following seven tools is most promising: Life Cycle Analysis, Substance Flow Analysis, Multi-Criteria Analysis, Technology Assessment, Sensitivity Analysis, Scenario Analysis and Cost-Effectiveness Analysis. Chapter 2: Selecting environmental systems analysis tools - 14 - 2.1. Introduction In this chapter, different environmental systems analysis (ESA) tools assessing the environmental performance of a company will be reviewed. The tools will be discussed with respect to their usefulness, alone or in combination, in decision support tools for companies that want to analyse options to reduce their environmental impact. We aim to answer the first research question of this thesis: “What existing environmental systems analysis methods and tools can in principle be combined in a decision support tool (DST) and used to analyse the environmental performance of a plant from a company perspective?”. In the following, we will first review ESA tools currently used by industry (section 2.2). Next, the characteristics of the DST to be developed will be described (section 2.3). And finally, we will select the tools to be combined in our DST (section 2.4). 2.2. Overview of Environmental Systems Analysis Tools currently used by industry Several ESA tools exist that have been used by industry. For the purpose of the thesis in this overview, a selection of environmental systems analysis tools is discussed (selected from SETAC, 1997; Wrisberg et al., 2002, Sonnemann et al., 2004; Finnveden and Moberg, 2005; Moberg, 2006). The tools included are considered useful in an integrated analysis of possibilities to improve the environmental performance of an industrial company, while considering several environmental pollutants, and while taking a company’s perspective. The main criteria for the choices of tools are their usefulness in an analysis 1) that takes a company perspective; 2) that includes environmental and economic aspects of decision making; 3) that includes a complete coverage of the potential environmental impacts and 4) that allows for determination of the consequences of a set of alternative strategies for pollution reduction. Based on these criteria, we selected twelve ESA tools. These include Environmental Management Systems, Life Cycle Assessment, Environmental Performance Evaluation, Substance Flow Analysis, Multi-Criteria Analysis, Technology Assessment, Sensitivity Analysis, Uncertainty Analysis, Total Cost Assessment, Cost Benefit Analysis, Cost- Effectiveness Analysis and Scenario Analysis. In the following, these tools1 are discussed. A short description is presented together with a brief reference to the tools’ characteristics and the extent to which they have been applied in industry. • An Environmental Management System (EMS) specifies how an organisation can formulate an environmental policy and objectives taking legislative requirements and information about significant environmental aspects into account (UNEP/SETAC, 1 Wrisberg et al. (2002) and Sonnemann (2004) distinguish between analytical tools, procedural tools and technical elements. Here, however, we refer to all these analytical instruments as ESA tools. Chapter 2: Selecting environmental systems analysis tools - 15 - 2005). This tool has been widely implemented and has a strong policy perspective, assuring that the organisation not only meets present day environmental legal and policy requirements but will continue to do so (ISO, 2004). Using EMS, industrial companies aim at keeping the environmental burden of their processes within the limits set by environmental legislation or to minimise the impacts of their processes (Neto et al., 2003; ISO, 2005). • Life Cycle Assessment (LCA) is a tool aiming at specifying the environmental consequences of products or services over their entire lifetime (ISO, 1997; Guinée, 2002; Rebitzer et al., 2004). LCA is a tool for comparative assessments, either between different products providing similar functions or between different life cycle stages of a product in an improvement analysis (Björklund, 2000). LCA has been applied to products and functions in various sectors, predominantly in the primary and secondary sectors of industry (e.g. Berkhout and Howes, 1997; Scholl and Nisius, 1998; Frankl and Rubik, 1999; Jiménez-González et al., 2000; Lee and Ding, 2000; Zobel et al., 2002; Curran, 2004; Siegenthaler and Margni, 2005; Rebitzer and Buxmann, 2005; Tan and Khoo, 2005; Rebitzer, 2005). • Environmental Performance Evaluation (EPE) uses indicators to transform the vast quantity of information about a firm in a comprehensive and concise manner by using indicators (Olsthoorn et al., 2001; Kolk and Mauser, 2002; Barbirolli and Raggi, 2003). At a firm level the indicators of the environmental performance are mostly used to relate absolute material and energy flows to process variables providing information about an organisation’s environmental performance (ISO, 1999; Jasch, 2000). Because many firms have developed their own performance indicators, several initiatives to bring consensus on indicators have been taken (WRI, 1997; NRTEE, 1999; WBCSD, 1999; GRI, 2000). Moreover, initiatives proposing the harmonisation of environmental performance indicators are taking place (Berkhout, et al., 2001). • Substance Flow Analysis (SFA) can be used to quantify the in- and outflows, as well as a balance of one particular substance trough the material economy (SETAC, 1997). It can highlight opportunities for environmental improvement related to the substance by identifying major inflow and outflow nodes in the system (SETAC, 1997). Substance Flow Analysis (SFA) focuses on specific substances, either within a region or through its entire life cycle. Typical examples include studies of nitrogen flows or flows of a specific metal (Kytzia and Nathani, 2004; Finnveden and Moberg, 2005). • Multi-Criteria Analysis (MCA) is a tool to support decision making based on multiple criteria. MCA may assist in identifying trade-offs between different criteria and finding the best solutions (Wrisberg et al., 2002). The tool is developed for complex problems that include qualitative and/or quantitative aspects of the problem in the decision-making process (CIFOR, 1999). This tool can be used to evaluate the relative importance of all criteria involved and reflect their importance in the final decision making process (CIFOR, 1999). MCA has been applied to studies in which aggregation of environmental data is needed. Examples can be found in Pineda- Henson et al. (2002), Pun et al. (2003), Rahimi and Weidner (2004), Cziner et al. (2005), Hermann et al. (2006). Chapter 2: Selecting environmental systems analysis tools - 16 - • The purpose of Technology Assessment (TA) is to evaluate the impact of a new technology before it is implemented at a large scale. Recently the term environmental technology assessment came into use (Björklund, 2004.). TA is a tool that assesses the impact of technology, to choose from technologies, to contribute to improved technology, to identify protective measures and to show if a technology complies with laws and regulations (UN Agenda 21, 1992; Björklund, 2004). Some examples where TA can be used are, for instance, to analyse the use of end-of-pipe techniques, the substitution of unfriendly products or trough the use of technology innovation to reduce the environmental burden of industrial production (Moors et al., 2005; Assefa et al., 2005). • Sensitivity Analysis (SA) is a systematic inventory of the changes in model results as a consequence of changing the values of the parameters or the input variables used in a model. Another definition (ISO, 1997) is that it is a systematic procedure for estimating the effects of the choices made regarding methods and data on the outcome of a study. This tool can be used to analyse the sensitivity of the model results to values of model parameters and is used in model building and in presenting results from model studies (e.g. Sonesson et al., 2000; Pluimers, 2001). • Uncertainty Analysis (UA) is conducted to assess the uncertainties in the results of a study. This may be done by comparing the importance of uncertain input parameters with respect to their contributions to output uncertainty. Morgan and Henrion (1990) considered elements of effective uncertainty analysis and communication of these uncertainties is essential for quantitative policy analysis (see Morgan and Henrion for a discussion of the effectiveness of uncertainty analysis). Examples of uncertainty analysis range from estimating uncertainties in emission inventories (e.g. Van Aardenne, 2002; Frey and Zao, 2004) to the estimation of uncertainties in industrial databases (e.g. Sugiyama et al., 2005) or in model results (e.g. Pluimers, 2001; Norris and Yost, 2002; Neuman, 2003; Walker et al., 2003) and in life cycle assessments (e.g. Kaplan et al., 2005; Geisler et al., 2005). Many methods to assess uncertainties exist, ranging from qualitative assessments of uncertainties to quantitative statistical approaches. • Total Cost Assessment (TCA) describes the analysis of the full range of internal costs and savings resulting from pollution prevention projects and other environmental project undertaken by a firm (Wrisberg et al., 2002; UNEP/SETAC, 2005). The tool seeks to integrate environmental costs into a capital budgeting analysis (Beaver, 2000). Examples of studies related with developments and industrial applications of TCA are overviewed in Backman and Thun (1999). • Cost Benefit Analysis (CBA) is an economic tool used to determine whether or not the benefits of an investment or a policy outweigh its costs (Wrisberg et al., 2002). It aims at expressing all positive and negative effects of an activity in monetary units. These effects may include economic and environmental aspects (RPA, 1998). • Cost-Effectiveness Analysis (CEA) is a variant of cost benefit analysis (CBA) (Wrisberg et al., 2002) and can be used to estimate the costs per unit of avoided emission (Rabah, 1999; Pluimers 2001; Klimont et al., 2002). Cost-effectiveness analysis is a techno-economical tool that considers only the internals costs, i.e., the Chapter 2: Selecting environmental systems analysis tools - 17 - cost resulting from emission reduction technologies (Sonnemann et al., 2004). These costs are compared to the reduction of the environmental pressure as a consequence of the economic investment. Cost-effectiveness is considered to be a useful criterion for ranking alternatives (Schwarz, 1997). • Scenario Analysis (ScenA) is a tool to explore future trends. In many studies, it results in a set of answers to “What… if” type for questions illustrating the future consequences of a range of alternative decisions (Schwarz, 1997; Pesonen et al., 2000; Pallottino et al., 2005). Scenarios do not necessarily portrait what the future will look like but instead aim to stimulate ways of thinking about alternative futures. Scenario analysis is a useful tool when complexity and uncertainty are high (Wollenberg et al., 2000). Many examples of the use of scenarios analysis are available in the literature (e.g. Pluimers, 2001; Fukushima and Hirao, 2002). Each tool has its specific characteristics which are reviewed in Table 2.1. Obviously, the tools differ with respect to their purpose, methodology, final product, strengths, weaknesses and relevance for an analysis taking a company perspective. In the following, the tools are discussed with respect to each of these characteristics. In Table 2.1. the tools are first compared with respect to their purpose. The comparison shows that they all provide industry with information that is helpful for environmental decision making. Nevertheless, the tools serve different purposes. Some tools are primarily used to assess the environmental impact of human activities (e.g. EMS, LCA, EPE, SFA), while others are more focusing on the evaluation or consequences of environmental management (e.g. TA, CBA). Three tools specifically address economic consequences of decisions made (TCA, CBA, CEA). Second, the tools are compared with respect to the methodology applied. For some tools specific procedures exist (EMS, LCA, EPE, MCA and SA). For some others, the method is not as clearly defined and may depend on the objective of the study at hand (SFA, TA, UA, TCA, CBA, CEA and ScenA). Some tools are often used in combination (e.g. the use of EPE within EMS, the use of MCA based on results from LCA, the use of SA as a complementary step to LCA or the use of CEA after LCA). This illustrates that individual tools in itself are often not sufficient for dealing with complex issues. Third, the tools differ with respect to their products. The results are in most cases quantitative. They range from the changes in the environmental performance to the costs per unit environmental performance improved. For instance, LCA results are typically in terms of the potential environmental impact for certain environmental impact categories. EPE results include a number of indicators, which in contrast to LCA, allow for identifying trends in environmental performance. Finally MCA can be used to express the environmental performance in one overall indicator. Fourth, all tools have their specific strengths. This may help in selecting the most appropriate tool for a specific study. For instance, EMS, LCA, EPE and SFA are comparable in the sense that they all aim to quantify the environmental performance. However, EMS is probably the most widely accepted by industrial companies, LCA is the most powerful tool to assess the complete lifetime of a product, EPE may be the Chapter 2: Selecting environmental systems analysis tools - 24 - In addition, the fourth characteristic (iv) refers to the analysis of technological options aiming to reduce the environmental impact of the industrial process. Such analysis may be based on Technology Assessment. TA may be a useful way to analyse pollution reduction options in terms of their potential to reduce the environmental impacts, i.e., by assessing the consequences of a new technology or a modification of an existing technology. Characteristic v points to the need to calculate the costs associated with the implementation of pollution reduction options. CEA is obviously the first choice to calculate the cost-effectiveness of pollution reduction options. This tool allows for ranking of pollution reduction options by calculating the costs per unit of environmental impact reduced. This implies that our DST will not provide total costs or benefits in monetary units. It is important that the user of the model has confidence in the results. The reliability of a model depends on the quality of the model parameters and model structure. A typical way to assess the sensitivity of model results to changes in model parameters is to perform a sensitivity analysis. Thus, SA is a tool that will assist the model development and application. Characteristics vi and vii express the overall environmental impact in terms of one single indicator and the intention to define pollution reduction strategies to reduce the overall environmental impact. MCA, as a tool, allows for assessing the overall environmental impact in one overall indicator. This tool, however, may use several methods that take into account multiple criteria and their relative weights. Characteristic vii implies analysis of pollution reduction strategies reflecting different management strategies, which can be done by scenario analysis (ScenA). The consequences of a range of alternative combinations of pollution reduction options may thus be analysed. The abovementioned seven tools (LCA, SFA, TA, CEA, SA, MCA and ScenA) will be used as a basis for our DST. This set of tools excludes EMS, EPE, TCA, CBA and UA. Although of significant importance when assessing the environmental performance and total costs of an industrial process, these tools are not the first choice options for our DST for the following reasons. EMS lacks the structured methodology aimed for in our approach. Our aim is to develop a reproducible DST. EPE is currently used on the industrial practice to assess the environmental performance but is not our first choice. EPE typically results in indicators allowing for identifying trends in the performance by considering a retrospective analysis, based on measured data made available by the industry. Our approach is different from that. We aim to outline and assess possible future developments regarding strategies on pollution reduction. TCA is not selected because our aim was not to perform a full economic analysis but to limit ourselves to an analysis of the cost-effectiveness of pollution reduction options and the costs associated with the pollution reduction strategies. Moreover, TCA is time consuming and requires a cost inventory of all internal costs of a company regardless of the relevance of the costs for the analysis. Cost Benefit Analysis, which is also an economic tool used to express all positive and negative effects of an activity in monetary units, is not a first choice for our analysis. Even though CBA has been applied at the company level, the monetarisation of the benefits of environmental management is often too uncertain to make this tool useful for studies taking a company’s perspective. Finally, Chapter 2: Selecting environmental systems analysis tools - 25 - we will not perform a full quantitative UA. Instead, we will compare our model results to company data, and we will perform a sensitivity analysis. This may be sufficient for ensuring confidence in the quality of our model. As discussed earlier, in the literature several examples can be found of combinations of ESA tools (e.g. Schmidt et al., 1996; Finkbeiner et al., 1998; Tukker et al., 1998; Marano and Rogers, 1999; Backman and Thun, 1999; Wrisberg et al., 2002; Beaver, 2002; Moberg, 2006; Hermann et al., 2006). None of these studies, however, combine the seven tools selected here to develop a DST taking a company perspective. In the following chapters, we aim to providing companies with an instrument to assess the potential environmental impact of industrial processes and to analyse options to reduce this environmental impact and the associated costs. References Assefa, G., Eriksson, O., Frostell, B., 2005. Technology assessment of thermal treatment technologies using ORWARE. Energy Conversion and Management 46 797-819. Backman, M., Thun, R., 1999. Total Cost Assessment. Recent developments and Industrial Applications. In: Backman, M., Thun, R. (Eds.), IIEE Communications, International Institute for Industrial Environmental Economics at Lunds University, Finland. Available from http://www.iiiee.lu.se/Publication.nsf/DisplayCommunications?OpenPage, last acceded on 20.12.2006. Barbirolli, G., Raggi, A., 2003. A method for evaluating the overall technical and economic performance of environmental innovations in production cycles. Journal of Cleaner Production 11 365-374. Beaver, E., 2002. LCA and Total Cost Assessment. Environmental Progress 19 (2) 130-139. Berkhout, F., Howes, R., 1997. The adoption of life-cycle approaches by industry: Patterns and impacts. Resources, Conservation and Recycling 20 71-94. Berkhout, F., Hertin, J., Azzone, G., Carlens, J., Drunen, M., Jasch, C., Noci, G., Olsthoorn, X., and others, 2001. MEPI (Measuring the environmental performance of industry), 2001. Final report to EC Environment and Climate Research programme: Research theme 4 Human Dimensions of environmental change. Björklund, A.E., 2002. Survey of Approaches to Improve Reliability in LCA. International Journal of LCA 7 (2) 64-72. Björklund, A., 2004. Environmental System Analysis – Some Ongoing Research and Ideas About Future Developments. In: Systems Approaches and Their Applications: Examples from Sweden, Kluwer Academic Publishers. Bouman, M., Heijungs, R., Van der Voet, E., van den Bergh, J.C.J.M., Huppes G., 2000. Material flows and economic models: an analytical comparison of SFA, LCA and partial equilibrium models. Ecological Economics 32 195-216. Cacuci, D.G., 2003. Sensitivity and uncertainty analysis – vol.1- Theory. In: Chapman & Hall/CRC (Eds.), A CRC Press Company. CEFIC, 1992. Technology assessment: a tool towards sustainable development by the chemical industry. European Chemical industry council (CEFIC). Position paper. CIFOR, 1999. Guidelines for Applying Multi-Criteria Analysis to the assessment of criteria and indicators, Center for International Forestry Research (CIFOR), Jakarta. Chapter 2: Selecting environmental systems analysis tools - 26 - Curran, M.A., 2004. The Status of Life-Cycle Assessment as an Environmental Management Tool. Environmental Progress. Environmental Progress 23 (4) 277-283. Cziner, K., Tuomaala, M., Hurme, M., 2005. Multicriteria decision making in process integration. Journal of Cleaner Poduction 13 (5) 475-483. EEA, 2001. Scenarios as tools for international environmental assessments, Experts' corner report. Prospects and Scenarios No 5, Environmental issue report No 24, European Environmental Agency. EMAS, 2001. Regulation (EC) No 761/2001 of the European parliament and of the council of 19 March 2001 allowing voluntary participation by organisations in a Community ecomanagement and audit scheme (EMAS). Finkbeiner, M., Wiedemann, M., Saur, K., 1998. A Comprehensive Approach Towards Product and Organisation Related Environmental Management Tools. International Journal of LCA 3 (3) 169-178. Finnveden, G., Moberg, A., 2005. Environmental systems analysis tools an overview. Journal of Cleaner Production 13 1165-1173. Frankl, P., Rubik, F., 1999. Life Cycle Assessment (LCA) in business. An overview on drivers, applications, issues and future perspectives. Global Nest: the Int. Journal 1 185-194. French, S., Geldermann, J., 2005. The varied contexts of environmental decision problems and their implications for decision support. Environmental Science & Policy 378-391. Frey, H.C., Zao, Y., 2004. Quantification of variability and uncertainty for air toxic emission inventories with censored emission factor data. Environmental Science & Technology 38 22. Fukushima, Y., Hirao, M., 2002. A structured framework and language for scenario-based life cycle assessment. International Journal of LCA 7 (6) 317 – 329. Geisler, G., Hellweg, S., Hungerbühler, K., 2005. Uncertainty analysis in life cycle assessment (LCA): case study on plant protection and implications for decision making. International Journal of LCA 10 (3) 192.1-192.3. GRI (Global Reporting Initiative), 2000. Sustainability Reporting Guidelines on Economic, Environmental and Social Performance, Boston. Guinée, J.B., 2002. Handbook on Life Cycle Assessment. Operational Guide to the ISO Standards. In: Series: Eco-efficiency in Industry and Science, Vol 7, Guinée J.B., (Eds.), Kluwer Academic Publishing, The Netherlands. Hermann, B.G., Kroeze, C., Jawjit, W., 2006. Assessing environmental performance by combining life cycle assessment, multi-criteria analysis and environmental performance indicators. Journal of Cleaner Production, in press. ISO, 1997. (ISO 14040). Environmental Management – Life Cycle assessment – Principles and Framework. International Organisation for Standardisation. ISO, 1999. (ISO 14031). Environmental Management – Environmental Performance Evaluation – Guidelines. International Organisation for Standardisation. ISO, 2004. (ISO 14001). Environmental Management Systems – Requirements with Guidance for use. International Organisation for Standardisation. ISO, 2005. The Global Use of Environmental Management System by Small and Medium Enterprises. Executive Report from ISO/TC207/SC1/ Strategic SME Group. Jasch, C., 2000. Environmental performance evaluation and indicators. Journal of Cleaner Production 8 79-88. Chapter 2: Selecting environmental systems analysis tools - 27 - Jiménez-González, C., Kim, S., Overcash, M.R., 2000. Methodology for Developing Gate-to- Gate Life Cycle Inventory Information. International Journal of LCA 5 (3) 153 – 159. Kaplan, P.O., Barlaz, M.A.., Ranjithan, S.R., 2005. A procedure for life-cycle based solid waste management with consideration of uncertainty. Journal of Industrial Ecology vol. 8 n.4. Klimont, Z., Cofala, J., Bertok, I., Amann, M., Heyes, C., Gyarfas, F., 2002. Modelling Particulate Emissions in Europe A Framework to Estimate Reduction Potential and Control Costs. IIASA Interim Report IR-02-076. Kolk, A., Mauser, A., 2002. The evolution of environmental management: from stage models to performance evaluation. Business Strategy and the Environment 11 14-31. Kytzia, S., Nathani, C., 2004. Bridging the gap to economic analysis: economic tools for industrial ecology. Progress in Industrial Ecology vol. 1 n. 1/2/3 143-163. Lee, Y.M., Ding, C.Y., 2000. Life-cycle assessment and production policy: An application to corrugated paperboard manufacture. Journal of Environmental Management 59 157–165. Marano, J.J., Rogers, S., 1999. Process optimization for life cycle improvement. Environmental Progress 18 40. Moberg, A., 2006. Environmental systems analysis tools for decision-making. LCA and Swedish waste management as an example. Licentiate thesis. Royal Institute of Technology, Department of Urban Planning and Environment, Environmental Strategies Research. Available from http://www.infra.kth.se/fms/pdf/kappaAM, last acceded on 02.12.2006. Moors, E.H.M., Mulder, K.F., Vergragt, P.J., 2005. Towards cleaner production: barriers and strategies in the base metals producing industry. Journal of Cleaner Production 13 657-668. Morgan, G.M., Henrion, M., 1990. Uncertainty: A guide to dealing with uncertainty in quantitative risk and policy analysis. Cambridge University Press, New York. Neto, B., Kroeze, C., Pulles, M., 2003. Environmental Performance of a Metals Industry facility. Industrial Production Chain Analysis. Paper presented at the second ISIE Conference ‘Industrial Ecology for a Sustainable Future’, June 29-July 2, Ann Arbor, USA. Neuman, SP., 2003. Maximum likelihood Bayesian averaging of uncertain model predictions. Stochastic Environmental Research and Risk Assessment 17 291-305. Norris, G.A., Yost, P., 2002. A transparent, interactive software environment for communicating life-cycle assessment results. Journal of Industrial Ecology vol. 5 n. 4. NRTEE (National Round Table on the Environment and Economy), 1999. Measuring Eco- Efficiency in Business: Feasibility of a core set of indicators, available in: http://www.nrteetrnee.ca, issued by the National Round Table on the Environment and the Economy, Ottawa, Ontario, Canada Olsthoorn, X., Tyteca, D., Wehrmeyer, W., Wagner, M., 2001. Environmental indicators for business: a review of the literature and standardisation methods. Journal of Cleaner Production 9 453-463. Pallottino, S., Sechi, G.M., Zuddas, P., 2005. A DSS for water resources management under uncertainty by scenario analysis. Environmental Modelling & Software 20 1189-1193. Pesonen, H-L., Ekvall, T., Fleischer, G., Huippes, G., Jahn, C., Klos, Z.S., Rebitzer, G., Sonnemann, G.W., Tintinelli, A., Weiodema B.P., Wenzeel, H., 2000. Framework for scenario development in LCA. International Journal of LCA 5 (1) 21 – 30. Pineda-Henson, R., Culaba, A.B., Mendoza, G.A., 2002. Evaluating Environmental Performance of Pulp and Paper Manufacturing Using Analytical Hierarchy process and Life-Cycle Assessment. Journal of Industrial Ecology Volume 6 Number 1. Chapter 2: Selecting environmental systems analysis tools - 28 - Pluimers, J., 2001. An environmental systems analysis of greenhouse horticulture in the Netherlands, PhD thesis Wageningen University, The Netherlands. Pun, Kit-Fai, Hui, Ip-Kee, Lewis, W.G., Lau, H.C.W., 2003. A multiple-criteria environmental impact assessment for the plastic injection molding process: a methodology. Journal of Cleaner Production 11 41–49. Quade, E., 1997. Predicting the consequences: Models and Modelling. In: H.J. Miser and E.S. Quade, (Eds.), Handbook of Systems Analysis. Volume one – overview of uses, procedures, applications, and, practice, John Wiley and Sons, Chichester, 191-218. Rahimi, M., Weidner, M., 2004. Decision Analysis Utilizing Data from Multiple Life-Cycle Impact Assessment Methods Part I: A Theoretical Basis. Journal of Industrial Ecology 8 (1–2) 93-118. Rabah, M., 1999. Cost effectiveness of abatement options for emission control in Egyptian iron foundries. Waste management 19 283-292. Rebitzer, G., Ekvall, T., Frischknecht, R., Hunkeler, D., Norris, G., Rydberg, T., Schmidt, W.P., Suh, S., Weidema, B.P., Pennington, D.W., 2004. Life cycle assessment Part 1: Framework, goal and scope definition, inventory analysis and applications. Environmental International 30 701- 720. Rebitzer, G., 2005. Enhancing the application efficiency of life cycle assessment for industrial uses, PhD thesis, Lausanne, Switzerland. Rebitzer, G., Buxmann, K., 2005. The role and implementation of LCA within life cycle management at Alcan. Journal of Cleaner Production 13 1327-1335. Robèrt, K.H., Schmidt-Bleek, B., Larderel, J.A., Basile, G. Jansen, J.L., Kuehr, R., Thomas, P.P., Suzuki, M., Hawken, P., Wackernagel, M., 2002. Strategic sustainable development – selection, design and synergies of applied tools. Journal of Cleaner Production 10 197-214. RPA, 1998. Economic Evaluation of Environmental Policies and Legislation; final Report for DGIII of the European Commission, Contract Number: ETD/97/501287. Schmidt et al., 1996 as cited in: Jensen, A. A., Hoffman L., Møller B.T., Schmidt, A., 1997. Life Cycle assessment– A Guide to approaches, experiences and information sources. European Environmental Agency. Environmental issues no. 6. Scholl, G.U., Nisius, S., 1998. The environmental benefits to German companies through application of LCA, Journal of Cleaner Production 6 247-252. Schwarz, B., 1997. Forecasting and Scenario’s. In: H.J. Miser and E.S. Quade, (Eds.), Handbook of Systems Analysis. Volume two Craft issues and procedural choices, John Wiley and Sons, Chichester, 327-367. SETAC, 1997. Life cycle assessment and conceptually related programmes. SETAC Europe Working group on life cycle assessment and conceptually related programmes. Report of the SETAC-Europe working group on LCA and conceptually related programmes. Available from www.setac.org, last acceded on 30.05.2006. Siegenthaler, C.P., Margni, M., 2005. Dissemination, Application and Assessment of LCA in Industry. International Journal of LCA 10 (5) 377 – 378. Sonesson, U., Björklund, A., Carlsson, M., Dalemo, M., 2000. Environmental and economic analysis of management systems for biodegradable waste. Resources, Conservation and Recycling 28 29-53. Sonnemann, G., Castells, F., Schumacher, M., 2004. Integrated life-cycle assessment for industrial processes. In: Lewis Publishers (Eds.), CRC Press Company. Chapter 2: Selecting environmental systems analysis tools - 29 - Sugiyama, H., Fukushima Y., Hirao, M., Hellweg, S., Hungerbühler, K., 2005. Using standard statistics to consider uncertainty in industry-based life cycle inventory databases. International Journal of LCA 10 (6) 399-405. Tan, R.B.H, Khoo, H.H., 2005. Zinc Casting and Recycling. International Journal of LCA 10 (3) 211-18. Tukker, A., Kleijn R., Van Oers L., Smeets, E., 1998. Combining SFA and LCA. The Swedish PVC analysis. Journal of Industrial Ecology vol 1 n. 4 93-116. UN Agenda 21, 1992. Available in http://www.unep.org, last acceded on 02.10.2005, United Nations Environment Programme. UNEP/SETAC, 2005. Life Cycle Approaches . The road from analysis to practice. United Nations Environment Programme. Division of technology, Industry and Economics (DTIE) Production and Consumption unit, France. Van Aardenne, J., 2002. Uncertainties in emission inventories, PhD thesis Wageningen University, The Netherlands. Van der Voet, E., van Oers, L., Guinée, J.B., de Haes, H.A.U., 1999. Using SFA indicators to support environmental policy. Environmental Science and Pollution Research 6 (1) 49-58. Walker, W.E., Harremoes, P., Rotmans, J., Van der Sluijs, J.P., Van Asselt, M.B.A., Janssen, P., Krayer von Krauss, M.P., 2003. Defining Uncertainty – a Conceptual Basis for Uncertainty Management in Model-based Decision Support. Integrated Assessment 4 (1) 5-17. WBCSD (World Business Council for Sustainable Development), 1999. Measuring Eco- Efficiency: A guide to reporting company performance. In: Accountability & Reporting Eco- Efficiency series. Wollenberg, E., Edminds, D., Buck, L., 2000. Anticipating change: Scenarios as a tool for adaptive forest management a guide. Center for International Forestry Research (CIFOR), Indonesia. WRI (World Resources Institute, 1997. BUSINESS CASE STUDY: Industrial Products Inc.: Measuring Environmental Performance. In: WRI Publications by Richard Wells, Abt Associates. Wrisberg, N., Haes, H.A.U., Triebswetter, U., Eder, P., Clift, R., 2002. Analytical tools for Environmental Design and Management in a Systems Perspective, the combined use of analytical tools. In: Wrisberg, N., Haes, H.A.U., Triebswetter, U., Eder, P., Clift, R., (Eds.), Eco-efficiency in Industry and Science, Kluwer academic publishers. Zobel, T., Almroth, C., Bresky, J., Burman, J., 2002. Identification and assessment of environmental aspects in an EMS context: an approach to a new reproducible method based on LCA methodology. Journal of Cleaner Production 10 381-396. Zopounidis, C., Doumpos, M., 2002. Multi-group discrimination using multi-criteria analysis: Illustrations from the field of finance. European Journal of Operational Research 139 371-389. - 30 - - 31 - Chapter 3: Inventory of Pollution Reduction Options for an Aluminium Pressure Die Casting Plant Belmira Neto, Carolien Kroeze, Leen Hordijk, Carlos Costa This chapter has been submitted to International Journal of Environment and Pollution Abstract This study presents a general overview of options aiming to reduce emissions to air, soil and water from an aluminium die casting plant located in Portugal. We first identify pollution reduction options and then estimate their potential to reduce the pollution and the costs associated with the implementation of these options. We identify eighteen technical reduction options that are applicable to aluminium pressure die casting companies. The options include typical end-of-pipe solutions as well as alternative techniques or still modifications in process operations from the die casting plant. We distinguish between different types of options, including, for instance, fabric filters and scrubbers; alternative desoxidation agents; modifications of the combustion process; alternative mould release application techniques; new die casting moulds and alternative equipment. Finally, we calculate the implementation costs for the company of each reduction option. The calculated net additional costs include fixed and variable costs. We conclude that there are promising opportunities to reduce the pollution from aluminium pressure die casting. Our inventory includes options with net negative costs, indicating that the company may in fact gain from implementing these options. Even though our study specifically focuses on one particular plant, the results may be interesting for the aluminium pressure die casting sector industry in general. Chapter 3: Inventory of pollution reduction options - 32 - 3.1. Introduction Aluminium pressure die casting is a manufacturing process supplying automotive industry with engineered car components (Brown, 1999). During the industrial process the aluminium alloy is molten, shaped on die casting moulds and submitted to different types of surface finishing processes in order to accomplish the client’s requirements (NADCA, 1991; USEPA, 1999; US Department of Energy, 1999; US Department of Energy, 2004). The industrial sector represented by the European Foundry Association produced about 1100 thousands tons of aluminium die castings products (APF, 2003). This production value has a share of about 33% of the European production for the non-ferrous metals alloys (CAEF, 2003). The material inputs entering the production process are aluminium ingots and/or aluminium alloy mass recycled within the plant. The production process requires several other inputs such as energy and subsidiary materials. The aluminium pressure die casting industry contributes to a number of environmental problems caused by emissions released to air, soil and water (Kim et al., 2003). For instance, the industry is a source of metal emissions to the environment that may be toxic to humans and other organisms. Moreover, the industry emits air pollutants, such as nitrogen oxides and carbon dioxide, which cause tropospheric ozone formation, acidification, human toxicity and global warming. And finally, there are waste-related problems, potentially leading to soil pollution. This is the case for the aluminium dross produced in melting or the ceramic lining waste from the furnaces. Existing studies of the aluminium pressure die casting industry focus on environmental management for the industrial sector in general and aim to provide the aluminium die casting sector with information on how to realise a more environmentally sound die casting process (e.g. Kim et al., 2003; EIPPCB, 2005; Dalquist and Gutowski, 2004) or on determining its environmental performance (e.g. Backhouse et al., 2004). Some other studies regarding the metals industry also aim to assist the industrial sector to comply with environmental regulations (Moors et al., 2005) or to analyse proposed policy instruments (Moors, 2006). Most of these studies have been performed at the level of the industrial sector, or in other words, the intended user of the study results is meant to be the die casting industry sector. Relatively few studies exist that are specific for the company level (e.g. Park et al., 2002). To our knowledge, a complete and comprehensive overview of pollution reduction options for aluminium pressure die casting plants does not exist. In this study we therefore aim at answering the following questions. • Which technical pollution reduction options are available for reducing the environmental impact of an aluminium pressure die casting plant? • What are their technical potentials and the associated costs for the plant? To answer these questions we identify emission reduction options for an aluminium die casting plant, based on an inventory of materials and energy used in the industrial process and the associated emissions of pollutants and production of waste. The identification of options available for reducing the environmental pressure is largely based on the literature Chapter 3: Inventory of pollution reduction options - 33 - and discussions with the industrial plant managers. The technical potentials to reduce the pollution and costs are either from the literature or estimated, based on the options’ characteristics. The plant serving as case study provided feedback, as well as data on materials and energy consumption, and technical details about the production process. In the following, we will first review the aluminium pressure die casting process, using information from the studied plant, including the input materials, energy and the environmental problems (section 3.2). Next, the pollution reduction options are identified and the characteristics of each option are described in terms of potential to pollution reduction and costs (section 3.3). Finally, section 3.4 presents conclusions of this study. 3.2. Aluminium Pressure Die Casting 3.2.1. The industrial process and system definition Pressure die casting is a manufacturing process that produces accurately dimensioned, sharply defined and smooth- or textured –surfaced metal car components (Kim et al., 2003). This manufacturing process includes a number of subsequent production processes and uses a variety of materials and energy resources. The most important operations are the melting of aluminium alloy, shaping it into a semi-product (casting), several operating processes for surface finishing, and finally the product cleaning and degreasing and its expedition. The technologies used for the die casting process do not differ among European countries, or between Europe and the USA (Tan and Khoo, 2005). In this study an existing aluminium pressure die casting plant is taken as an example. This plant is located in northern Portugal and provided information about its production processes and the input and output flows of materials and energy. Since our study takes a company perspective, a gate-to-gate analysis is performed, i.e., this study only considers material flows within the gates of the plant. We assume that this reflects the span of control of the plant managers, and their primary interest in assessments of the environmental aspects of the plant. The company perspective is reflected by the choices made with respect to systems boundaries and systems elements (Figure 3.1). Within the system boundaries a number of processes are considered. These include processes that are contributing to pollution or waste streams and only processes that can be managed by the plant managers. We distinguish between five sub-processes within the aluminium die casting production plant: 1) Melting, 2) Casting, 3) Finishing, 4) Internal transports and 5) Auxiliary burners (see Figure 3.1). The system boundaries are chosen such that they include all relevant processes that can be managed by the plant managers. We consider the following outputs of the system: die casting products, emissions of pollutants, liquid effluents and the production of waste. Thus the environmental pressures taken into account include, beside emissions of air pollutants, liquid effluents and waste. Liquid effluents are a mixture of water and oils (from the sub-sub-processes Pressure Die Casting and Tumbling) or of water and detergents (from Cleaning and Degreasing). These effluents are treated in the plant’s wastewater treatment sites. The solid waste includes aluminium dross (from Melting), ceramic lining (lining from holding furnaces), steel shot (from Shot Blasting) and ceramic abrasives (from Chapter 3: Inventory of pollution reduction options - 40 - 3.3.3. Description of individual reduction options In the following, eighteen reduction options are described for the following sub-processes considered: Melting, Casting, Finishing, Internal transport and Auxiliary Burners. For each option a description is made of (1) what it does and how it does it, followed by (2) an estimate of the reduction factors and the associated costs, and (3) identification side-effects of options on materials or energy used or produced. Table 3.2 and Table 3.3 overview the pollution reduction options. Table 3.2 describes the types of options considered. Table 3.3 overviews each individual option and indicates the compounds reduced by them. These options include add-on techniques (fabric filters or wet scrubbers) or more structural reduction options that may change a material or technique used. We use a so-called reduction factor (RF) to express the percentage reduction in emissions possible by some of the options (Table 3.4). The effect of options on the activity rates are presented in Table 3.5. Some of the reduction options may have unintended side-effects leading, for instance, to extra consumption of materials or energy or to extra production of materials (see also Table 3.5 for the so-called extra activities). The prices of activities and extra activities used to calculate the costs associated to each option are also presented in Table 3.5. Finally, we estimated the costs of implementing these options (Table 3.6). The costs included are regarded as additional costs for the plant (in line with Geldermann and Rentz, 2004). In Table 3.6, we present the cost parameters used to calculate the net additional costs (Cna). These are the sum of the annualised capital costs (CI), the fixed costs (CO) and the variable costs (CV) (as presented in Table 3.6). For each reduction option (τ) the annualised capital costs are calculated as a function of the investment (I), the interest rate (r) and the equipment lifetime (lt). The fixed costs are calculated as a fraction (o) of the investment. In addition, there are variable costs (CV) including the cost of the consumption of materials and energy associated with the reduction options (in line with Klimont et al., 2002). In the following, we provide details on the net additional costs for each reduction option. The reduction potential and the associated costs, for each option, may together form the basis for deciding on the opportunities for pollution reduction by the industrial sector. Chapter 3: Inventory of pollution reduction options - 41 - Table 3.3. Overview of pollution reduction options for the aluminium pressure die casting plant. subprocess (pi) sub-sub- process (pij) Types of Options Reduction Options (τ) Abbreviation Compounds reduced Fabric Filter. Reverseair type a) Melting_FF_RA Heavy metals such as Cd, Ni, Pb, Cr Fabric Filter. Pulse-Jet type b) Melting_FF_PJ Heavy metals such as Cd, Ni, Pb, Cr Fabric Filter. Mechanical Shaker type c) Melting_FF_MS Heavy metals such as Cd, Ni, Pb, Cr Wet Scrubber. Impingement-Plate type d) Melting_WS_IP Heavy metals (Cd, Ni, Pb, Cr), Cu and HF Filters and scrubbers Wet scrubber. Spraychamber type e) Melting_WS_SC Heavy metals (Cd, Ni, Pb, Cr), Cu and HF and NMVOC. Alternative desoxidation agent Granular desoxidation agent f) Melting_GA HF, Aluminium dross. Alternative degassing technique Impeller station using N2f), k) Melting_IS HF, Aluminium dross. Alternative metal loading in furnaces Compact metal loading in furnaces g) Melting_CM CO2, CO, NOx, and NMVOC (Natural gas combustion related emissions). Air enrichment with oxygen (30%O2) g) Melting_AE CO2, CO, NOx, and NMVOC (Natural gas combustion related emissions). Melting (i=1) Melting Combustion process modification Oxyfuel firing (100%O2)g) Melting_OF CO2, CO, NOx, and NMVOC (Natural gas combustion related emissions). Wet Scrubber. Packed- Bed type h) Casting_WS_PB Heavy metals (Pb, Cr), Cu, Zn, NMVOC Scrubbers Wet scrubber. Spraychamber type e) Casting_WS_SC Heavy metals (Pb, Cr), Cu, Zn, NMVOC. New mould release agent g) Casting_nMA NMVOC, liquid effluent, oils, grease and sludge. Alternative to mould release agent application Powder agent i), k) Casting_PA NMVOC, liquid effluent, oils, grease and sludge. New die casting moulds Reduce runners mass j), *) Casting_rRR Heavy metals (Cd, Ni, Pb, Cr), Cu, Zn, HF, CO2, CO, NOx, and NMVOC, aluminium dross, oils, grease and sludge. Casting (i=2) Pressure die casting Reduce scrap rate Reduce scrap rate k), *) Casting_rSR Heavy metals (Cd, Ni, Pb, Cr), Cu, Zn, HF, CO2, CO, NOx, and NMVOC, aluminium dross, oils, grease and sludge. Finishing (i=3) Trimming Reduce scrap rate Reduce scrap rate k), *) Finishing_rSR Heavy metals (Cd, Ni, Pb, Cr), Cu, Zn, HF, CO2, CO, NOx, and NMVOC, aluminium dross, oils, grease and sludge. Internal Transport (i=4) Forklift Truck on Diesel (I,II) and LPG Electrical equipment Use electric forklift trucks g) IT_eFL CO2, CO, NOx, and NMVOC, SO2 and Particulates. a) EPA-CICA Fact Sheet. EPA-452/F-03-026. USEPA (2002); b) EPA-452/F-03-025. USEPA (2002); c) EPA-452/F-03-024. USEPA (2002); d) EPA- 452/F-03-012. USEPA (2002); e) EPA-452/F-03-016. USEPA (2002). f) Brown (1999) g) Pedro (2005). Personal communication. h) EPA-CICA Fact Sheet.. EPA-452/F-03-015. USEPA (2002). i) Klüber (2005) j) INETI (2000). k) EIPPCB (2005). *) These reduction options may change the value of the metal yield. See footnote in Table 3.2. for a definition of metal yield. Chapter 3: Inventory of pollution reduction options - 42 - Melting In the sub-process Melting five types of reduction options are identified (Table 3.3). These include i) filters and scrubbers, ii) alternative desoxidation agent, iii) alternative degassing technique, iv) alternative metal loading in furnaces and v) combustion process modifications. The options included in filters and scrubbers, aim at reducing emissions of heavy metals (cadmium, nickel, lead and chromium), copper, hydrogen fluoride and non-methane volatile organic compounds (Table 3.3). Fabric filters reduce the emitted heavy metals while wet scrubbers also reduce copper, hydrogen fluoride and non-methane volatile organic compounds. The potential reduction of these compounds is presented in Table 3.4. For instance, all three fabric filters considered (abbreviated in Table 3.3 as Melting_FF_RA, Melting_FF_PJ and Melting_FF_MS) could reduce emission of cadmium, nickel, lead and chromium by 99% relative to the unabated present case (Table 3.4). The wet scrubbers (abbreviated as Melting_WS_IP and Melting_WS_SC) could reduce hydrogen fluoride and copper emissions by 99% relative to the unabated present case (Table 3.4). One of the wet scrubbers analysed (Melting_WS_SC) has also a large potential to reduce non-methane volatile organic compounds (95% relative to the unabated present case) (Table 3.4). A side-effect of these scrubbers and filters is extra waste to be disposed. This is estimated at 2.5 ton of additional waste per year for fabric filters and 4 tons for wet scrubbers (both estimates refer to the plant that served as a case study here) (Table 3.5). The amount of additional waste produced is estimated based on the efficiency of the filters and scrubbers in collecting dust plus, in case of wet scrubbers, an estimated 60% of water in the sludge. It should be noted that emission factors for the abated case and for the die casting process are not available from the literature. Nevertheless, our estimates for dust collection are in line with emission factors available from the literature for the aluminium industry (EIPPCB, 2001).The net additional costs for fabric filters vary from 26 to 84 k€/y and for wet scrubbers from 8 to 13 k€/y (Table 3.6). Another possibility to reduce the environmental impact is to change the desoxidation agent used. This could be done by using a granular agent (Melting_GA as abbreviated in Table 3.3) as opposed to the conventional agents. Desoxidation agents are used to remove impurities from the molten bath. We assume that the amount of granular agent used is the same as the amount of the conventional agent used now i.e., 6400 kg/year (included as extra activity in Table 3.5) (Foseco, 2002). The granular agent has a lower content of fluorides and therefore gives rise to lower emissions of hydrogen fluoride emissions (Brown, 1999). The potential of this option to reduce hydrogen fluoride emissions is 62% relative to the reference case. This option also reduces the amount of aluminium dross formed. This is caused by an estimated reduction of 5% in the aluminium alloy in the aluminium dross (Foseco, 2002). The net additional cost of this option is -0.2 k€/y (Table 3.6). Changing the degassing technique is a next option. It involves a new degassing technique using an impeller station using N2 (Melting_IS as of Table 3.3). This technique is also used to remove gas impurities from the molten bath. This is done by promoting an agitation of the molten bath and the subsequent release of the gas entrapped (EIPPCB, 2005). Using an impeller station does not require the use of a solid agent containing fluoride compounds, Chapter 3: Inventory of pollution reduction options - 43 - but instead uses nitrogen that, being injected into the molten bath, promotes gas impurities to escape. This option thus reduces hydrogen fluorides emissions and aluminium dross formed relative to the unabated present case. This amount of gas N2 used is estimated to be 403 m3/year (Brown, 1999; EIPPCB, 2005) (included as extra activity in Table 3.5) and the net cost is 58 k€/y (Table 3.6). Changing the metal loading in furnaces may reduce the use of natural gas, and as a result all associated emissions of pollutants. This option makes use of equipment that breaks the runners (see footnote in Table 3.2 for a description of runner), to small pieces that are again melted. This option (Melting_CM as abbreviated in Table 3.3), allows loading the melting furnaces with a more compact aluminium alloy load. This contributes to smaller the voids existing between the different parts of metal load and results in an increase in the thermal efficiency of the process. Although this process is not well documented in the available literature, it clearly explores the furnace thermal efficiency. We assume that this option will increase the furnace efficiency to the average thermal efficiency (47.5%) indicated for the shaft furnaces used to melt aluminium alloy to the pressure die casting process (EIPPCB, 2005). Therefore, the option leads to a large increase on the thermal efficiency relative to the unabated present case and subsequently reduces the activity (natural gas consumption). The potential 58% reduction in natural gas use (as indicated in Table 3.5) was calculated from the heat needed to melt the same amount of aluminium alloy. Consequently, the emissions resulting from the natural gas combustion (as CO2, CO, NOx and NMVOC) are reduced likewise. The net additional cost is, however, comparatively low (-128 k€/y, see Table 3.6) indicating that the company gains from implementing this option. Finally, the combustion process modification makes use of oxygen in the melting process. This includes two options that use either a small percentage of air enrichment with oxygen (Melting_AE as abbreviated in Table 3.3) or use only oxygen. This last case is also referred to as oxyfuel firing (Melting_OF as abbreviated in Table 3.3). These combustion modifications exploit the latent heat present in the exhaust gases. As the specific heat from the exhaust gases decreases with the increase of the amount of oxygen, a decrease in the specific consumption of natural gas is expected. As result, this option leads to an increase in the efficiency of heat production and savings in the natural gas used. The option Melting_AE is calculated to reduce natural gas consumption of 2%, relative to the present case, while the reduction on the consumption of natural gas is for the option Melting_OF of 4%, relative to the current case (Table 3.5). The emissions resulting from natural gas combustion (as CO2, CO, NOx and NMVOC) are reduced likewise. The amount of oxygen used is estimated at 4.8E+05 m3/year, (Table 3.5 for Melting_AE) and the net cost associated with this option is 59 k€/y (Table 3.6). For the option that uses oxyfuel firing 1.6E+06 m3 of oxygen is needed annually (Table 3.5 for Melting_OF) while the net cost is relatively high and estimated to be 224k€/y (Table 3.6). Casting In sub-process Casting four types of reduction options are analysed (Table 3.3). These include i) scrubbers, ii) mould release agent application, iii) new die casting moulds and iv) reducing scrap rate. Chapter 3: Inventory of pollution reduction options - 44 - • Sub-sub-process Pressure Die Casting The scrubbers aim to reduce emission of metals (such as lead, chromium, copper and zinc) and non-methane volatile organic compounds (NMVOCs) (Table 3.3). They include two wet scrubbers (Casting_WS_PB and Casting_WS_SC as abbreviated in Table 3.3), which are very effective in reducing emissions. The wet scrubber of the type packed bed (Casting_WS_PB) reduces lead, chromium, copper and zinc by 95% relative to the unabated present case, and NMVOCs by 99% (Table 3.4). The wet scrubber of the type spray chamber (Casting_WS_SC) reduces metals emissions (lead, chromium, copper and zinc) by 99% relative to the unabated case, and NMVOC by 95% (see Table 3.4). These two scrubbers can estimate 0.5 ton/year of waste to be disposed (Table 3.5). This estimate is based on the efficiency of the scrubbers to collect dust plus an estimated 60% of water in the sludge. The amount of dust estimated is in line with literature values for the aluminium industry (EIPPCB, 2001). The net additional cost of scrubbers are 195 k€/y (Casting_WS_PB) and 22 k€/y (Casting_WS_SC) (Table 3.6). The mould release agent application includes two options. One option replaces the mould release agent by an alternative one (Casting_nMA in Table 3.3). The other uses a new technique where a powder agent is applied by electrostatic deposition into the die casting mould (Casting_PA in Table 3.3). These two options aim firstly to reduce the amount of NMVOC emitted. In addition, a reduction in the specific consumption of mould agent leads to a reduction in the waste generated by the wastewater treatment plants (oils and grease and sludge). Using a new mould release agent (Casting_nMA) reduces the amount of agent needed compared to the current practice. This in turn leads to a reduction in water use of 30% (see Table 3.5), in the emissions of NMVOCs, as well as liquid effluent, oils, grease and sludge produced. On the other hand, the new mould release agent used annually is estimated to be 28m3 (as indicated an extra activity in Table 3.5). The net cost associated to this reduction option is -58 k€/y, indicating that the company gains from implementing this option (Table 3.6). The option using a powder agent (Casting_PA), replaces the old spraying technique by a new one where a lubricant (powder agent) is applied into the die casting mould, by electrostatic deposition, before each die casting operation. The powder agent used reduces the emissions of NMVOC, eliminates the production of liquid effluent and consequently of oils, grease and sludge (Klüber, 2005). When a powder agent is used, the water needed in the sub-sub-process is reduced to zero (as seen in Table 3.5). However, the use of powder agent (as an extra activity) amounts to 4020 kg/y (extra activity in Table 3.5). The net cost is 146 k€/y (as shown in Table 3.6). The reduction factor for emissions of nonmethane volatile organic compounds (NMVOC) is not available from literature, but we tentatively assume a 100% potential to reduce NMVOC emissions, when using powder agent. Likewise, the production of liquid effluent, oils, grease and sludge are zero when powder agents are used. The new die casting moulds aim to reduce the mass of runners (option abbreviated to Casting_rRR in Table 3.3). This is possible by replacing the moulds with smaller cavities to runners. Reducing the runners mass (or in other words, reduce the amount of alloy in the runners to the die casting mould by using new moulds) will reduce most of the pollutants Chapter 3: Inventory of pollution reduction options - 45 - released by the plant process (Table 3.3). This occurs due to the fact that the plant recycles internally the excess of aluminium alloy mass, i.e., the alloy mass that is not part of the final product is sent back to the furnaces to be molten again. The materials recycled internally include the excess of alloy in the die castings (runners) and the scrap (die castings products that do not fulfil the final product requirements). Thus, if the amount of alloy in die castings is reduced, the alloy recycled internally also decreases and the materials and energy used in the process are affected. An estimated value for the reduction of the runners’ mass is not easily available from the literature. Some studies (INETI, 2000) estimate, for alternative moulds, a value that may vary up to 30% reduction in the runners’ mass, when compared to the conventional die casting moulds. However, this value depends on the type of the product produced. The option (Casting_rRR) is estimated to reduce the runners’ mass by 25% relative to the plant’s current practice. The use of new die casting moulds is then estimated to reduce the amount of aluminium alloy that is recycled internally to the melting furnaces by 16%, when compared to the present situation. This is because the percentage reduction in the runners only contributes to a part of the aluminium alloy (including runners and scraps from Casting and Finishing), that feeds the melting furnaces. Several materials and energy used in the sub-processes Melting and Casting are reduced likewise relative to the unabated case (as seen in Table 3.5). The net cost of using new moulds in the die casting machines is 119 k€/y (Table 3.6). Reducing the scrap rate (Casting_rSR in Table 3.3) aims to reduce the amount of scrap (rejected die casting products not fulfilling the final product requirements). As mentioned above, this option reduces the aluminium alloy mass recycled internally and therefore reduces a large number of pollutants released by the plant process. The potential to reduce the scrap rate per sub-process is not available from the literature. Rather, the available literature values refer to the conventional overall reduction of the average scrap rate for die casting companies. About 5% of the scrap is typical of an aluminium pressure die casting company (US Department of Energy, 1999; EIPPCB, 2005). This would imply a 50% reduction in the scrap rate for the sub-sub-process pressure die casting of the plant. This option is then estimated to reduce the amount of alloy that is recycled internally to the melting by 5%. Subsequently, the different materials and energy used in the sub-processes Melting and Casting, as well as emissions of pollutants are reduced by the same amount (5%), compared to the unabated case (Table 3.5). The company may gain 30 k€/y from implementing this option (Table 3.6). Finishing • Sub-sub-process Trimming Reduction of the scrap rate (Finishing_rSR in Table 3.3) is also possible in the sub-process Finishing. For the reasons mentioned above, this option affects a large number of pollutants released by the plant. The option (Finishing_rSR) is estimated to reduce the scrap rate by 60% compared to the plant’s current scrap rate for the sub-sub-process trimming. The amount of alloy that is recycled internally to Melting is also assumed to be 5% lower, as well as the use of different materials and energy in Melting, Casting and Finishing (Table 3.5) and the pollutants emitted. Moreover, net cost indicates that the company may gain 36 k€/y from implementing this option (Table 3.6). Chapter 3: Inventory of pollution reduction options - 46 - Internal transport • Sub-sub-process Forklift truck on Diesel (I and II) and LPG Currently, three forklift trucks are used in the plant, fuelled with diesel and LPG. It is possible to replace these by electric forklift trucks (electrical equipment; IT_eFL in Table 3.3). This would reduce the use of diesel and LPG to zero, and as a result the release of combustion products (Table 3.5). Instead of diesel and LPG, electricity is needed. We did not estimate the amount of electricity needed, nor the emissions associated with electricity production. Assuming that these are taking place outside the gate of the plant, and therefore beyond our system boundaries. The net additional cost is -39 k€/y indicating that the company gains from implementing this option (see Table 3.6). Auxiliary burners • Sub-sub-process Oxyacetylene and Butane Burners The plant uses oxyacetylene and butane burners. These are of environmental concern because of emissions of combustion compounds (e.g. CO2, CO and NOx). However, they are not used regularly in the current practice at the plant. Alternatives that are more environmentally sound include, for instance, the use of electrical equipment to replace oxyacetylene burners. However, we consider their impact on the overall environmental performance small and therefore this option is not included in the current analysis. Chapter 3: Inventory of pollution reduction options - 47 - Table 3.4. Reduction factors (RF) for options applicable to the aluminium pressure die casting plant. The reduction factors express the theoretical potential to reduce emissions by the add-on techniques considered. The units express the percentage of reduction for each pollutant (x) for each reduction option (τ). released at each sub-sub-process and relative to the unabated situation. See section 3.3.3 for a description of the pollution reduction options. Reduction Factors (RF) for each pollutant (x) subprocesses sub-sub- processes Reduction Option (τ) x=Al x=Cd x=Ni x=Pb x=Cr x=Cu x=HF x=NMVOC x=Zn x=Fe Melting_FF_RA a) 99.9% 99% 99% 99% 99% n.e. n.e. n.e. n.e. n.e. Melting_FF_PJ b) 99.9% 99% 99% 99% 99% n.e. n.e. n.e. n.e. n.e. Melting_FF_MS c) 99.9% 99% 99% 99% 99% n.e. n.e. n.e. n.e. n.e. Melting_WS_IP d) 99% 99% 99% 99% 99% 99% 99% n.e. n.e. n.e. Melting Melting Melting_WS_SC e) 99% 99% 99% 99% 99% 99% 99% 95% n.e. n.e. Casting_WS_PB f) 95% n.e. n.e. 95% 95% 95% n.e. 99% 95% 95% Casting Pressure Die Casting Casting_WS_SC e) 99% n.e. n.e. 99% 99% 99% n.e. 95% 99% 99% n.e. = no effect a) USEPA Air pollution control technology fact sheet. EPA-CICA Fact Sheet. EPA-452/F-03-026. USEPA (2002). b) USEPA Air pollution control technology fact sheet. EPA-CICA Fact Sheet. EPA-452/F-03-025. USEPA (2002). c) USEPA Air pollution control technology fact sheet. EPA-CICA Fact Sheet. EPA-452/F-03-024. USEPA (2002). d) USEPA Air pollution control technology fact sheet. EPA-CICA Fact Sheet. EPA-452/F-03-012. USEPA (2002). e) USEPA Air pollution control technology fact sheet. EPA-CICA Fact Sheet. EPA-452/F-03-016. USEPA (2002). f) USEPA Air pollution control technology fact sheet. EPA-CICA Fact Sheet. EPA-452/F-03-015. USEPA (2002) Chapter 3: Inventory of pollution reduction options - 48 - Table 3.5. Effect of each reduction option on the activity rates (Act) and Extra Activities (Xα ) for the aluminium pressure die casting plant. The values refer to the use of a certain material or energy (α). The values in percentage express the reduction on each activity rate (Actα) for each reduction option (τ) and for each sub-sub-process, relative to the unabated present situation. The units presented for the extra activities reflect the extra amount of materials (Xα ), required by the use of a reduction option and related to the unabated current situation. See section 3.3.3. for a description of the pollution reduction options. Reduction in activity rates (Actα) Reduction Option (τ) desoxidation agent degassing flux Natural gas Mould Release Agent Water Hydraulic Oil Tip Lubricant Other oils Steel Shot Ceramic Abrasives Detergent Diesel LPG Melting_FF_RA n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_FF_PJ n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_FF_MS n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_WS_IP n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_WS_SC n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_GA 100% b) n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_IS n.e. 100% b) n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_CM n.e. n.e. 58% e) n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_AE n.e. n.e. 2% f) n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_OF n.e. n.e. 4% f) n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Casting_WS_PB n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Casting_WS_SC n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Casting_nMA n.e. n.e. n.e. 100% b) 30% h) n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Casting_PA n.e. n.e. n.e. 100% b) 100% b) n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Casting_rRR 16% j) 16% j) 16% j) 16% j) 16% j) 16% j) 16% j) 16% j) n.e. n.e. n.e. n.e. n.e. Casting_rSR 5% k) 5% k) 5% k) 5% k) 5% k) 5% k) 5% k) 5% k) n.e. n.e. n.e. n.e. n.e. Finishing_rSR 5% l) 5% l) 5% l) 5% l) 5% l) 5% l) 5% l) 5% l) 5% l) 5% l) 5% l) n.e. n.e. IT_eFL n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 100% b) 100% b) unit price 1.32 €/kg m) 2.8 €/kg m) 0.33 €/m3 m) 1.53 €/liter m) 1.5€/m3 m) 1.05 €/liter m) 1.37€/liter m) 0.74 €/liter m) 0.59 €/kg m) 1.02€/kg m) 1.86€/liter m) 0.91 €/liter n) 1.23 €/kg n) * values referring to electricity or extra water use were not estimated. n.e. = no effect. n.s. = not specified. a) Estimated based on the amount of avoided air emissions. b) Assuming a full replacement of the agents currently used. Melting_GA: use of a granular agent replacing the currently used. Melting_IS: use of gas N2, replacing the solid agent. Casting_nMA: use of alternative mould release agent. Casting_PA: replace the liquid agent and water by a solid powder mould release agent. IT_eFL: use of electric forklift trucks replacing the fuelled (Diesel and LPG) currently used. c) Estimated from Foseco (2002). Personal communication. d) Estimated from Brown (1999) and EIPPCB (2005). e) Estimated based on the average furnace thermal efficiency for the type of furnace used in the plant. EIPPCB (2005). f) “Estimated” based on the latent heat present in the exhaust gases. g) “Estimated” based on the reduction of natural gas use reported to the plant unabated situation. h) “Estimated” based on the reduction of mould release agent use reported to the plant unabated situation. i) “Estimated” based on use reported to the number of die casting shots produced annually by the company and the indication from product use from Klüber (2005). j) “Estimated” based on the reduction of the aluminium alloy that is recycled internally. This option affects the materials and energy used on the sub-processes Melting and Casting. k) “Estimated” based on the reduction of the aluminium alloy that is recycled internally. This option affects the materials and energy used on the sub-processes Melting and Casting. l) “Estimated” based on the reduction of the aluminium alloy that is recycled internally. This option affects the materials and energy used on the sub-processes Melting, Casting and Finishing. m) Pedro (2005), Personal communication. n) GALP energy (2004), Personal communication. o) Foseco (2005), Personal communication. p) Praxair (2005), Personnal communication. q) Klüber (2005). Chapter 3: Inventory of pollution reduction options - 49 - Table 3.5. (cont.). Extra Activity (Xα) Reduction Option (τ) Electricity Water Waste to disposal Granular agent Gas N2 Oxygen New mould release agent Powder Agent Melting_FF_RA * n.e. 2.5 ton/yr a) n.e. n.e. n.e. n.e. n.e. Melting_FF_PJ * n.e. 2.5 ton/yr a) n.e. n.e. n.e. n.e. n.e. Melting_FF_MS * n.e. 2.5 ton/yr a) n.e. n.e. n.e. n.e. n.e. Melting_WS_IP * * 4 ton/yr a) n.e. n.e. n.e. n.e. n.e. Melting_WS_SC * * 4 ton/yr a) n.e. n.e. n.e. n.e. n.e. Melting_GA n.e. n.e. n.e. 6400 kg/yr c) n.e. n.e. n.e. n.e. Melting_IS * n.e. n.e. n.e. 403 m3/yr d) n.e. n.e. n.e. Melting_CM * n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_AE n.e. n.e. n.e. n.e. n.e. 4.8E+05 m3/yr g) n.e. n.e. Melting_OF * n.e. n.e. n.e. n.e. 1.6E+06 m3/yr g) n.e. n.e. Casting_WS_PB * * 0.5 ton/yr a) n.e. n.e. n.e. n.e. n.e. Casting_WS_SC * * 0.5 ton/yr a) n.e. n.e. n.e. n.e. n.e. Casting_nMA n.e. n.e. n.e. n.e. n.e. n.e. 28 m3/yr h) n.e. Casting_PA * n.e. n.e. n.e. n.e. n.e. n.e. 4020 kg/yr i) Casting_rRR n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Casting_rSR n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Finishing_rSR n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. IT_eFL * n.e. n.e. n.e. n.e. n.e. n.e. n.e. unit price n.s. 1.5€/m3 m) 200€/ton (dust) m) 220 €/ton sludge m) 37 €/ton (dross) m) 51 €/ton (oils and grease) m) 1.3€/kg o) 127 €/m3 p) 0.13 €/m3 p) 1.7 €/liter m) 55 €/kg q) Chapter 4: Modelling the environmental impact - 56 - 4.1. Introduction Aluminium is a widely used metal, in particular in the automotive industry. The need to reduce vehicle fuel consumption by reducing the weight of the car, has increased the interest in aluminium. For instance, the total mass of aluminium in a European car roughly doubled between 1990 and 2000 (EIPPCB, 2005). The expected growth of the use of aluminium to achieve lighter cars has an effect on the aluminium die casting industry. Aluminium pressure die casting is a manufacturing process in the non-ferrous industries, producing engineered aluminium alloy products, such as car components. Aluminium castings dominate the non-ferrous sector, comprising roughly 80 percent of the light alloy castings on the European aluminium market (CAEF, 2003). Pressure die casting is a widely used casting process for aluminium alloys and about two-thirds of all aluminium castings are used in automotive industry (Brown, 1999). The aluminium pressure die casting industry contributes to a number of environmental problems (Kim et al., 2003). For instance, it is a source of metal emissions to the environment that may be toxic to humans and other organisms. Moreover, this industry contributes to air pollution problems through emissions of gases that contribute to tropospheric ozone formation, acidification, human toxicity and global warming. And finally, there are waste-related problems, potentially leading to soil pollution. Today, the industrial sector has to meet environmental goals, aiming to reduce the environmental impact of the industrial activities (Finkbeiner et al., 1998; Silvo et al., 2002). In many countries environmental laws exist that, for instance, regulate the emissions of a number of pollutants, or include restrictions of the use of toxic compounds or waste handling. The aluminium die casting industry shows the worldwide trends of implementing environmental management systems to quantify their environmental performance (Neto et al., 2003; Hillary, 2004; Zobel and Burman, 2004). Despite existing regulations, it is not easy to answer the question of how the environmental impact of an individual company can be reduced most effectively. There are too many pollutants involved, and too many reduction options available, to easily get a good overview of the situation. A complicating factor is that many industrial processes result in more than one pollutant. In addition, reduction options typically, not only reduce the pollutant that they are aiming to reduce, but may have positive and negative side effects on other pollutants. And finally, industrial companies are not only interested in the most effective way to reduce emissions, but also in the most efficient way, in order to limit the costs of environmental control (Geldermann and Rentz, 2004). Existing environmental systems analysis tools may assist in getting insight into this complexity. For instance, Life Cycle Assessment (LCA) is a tool aiming to specifying the environmental consequences of products or services over its entire lifetime (Guinée, 2002; Rebitzer et al., 2004). Substance Flow Analysis (SFA) focuses on specific substances, either within a region or through its entire life cycle; typical examples include studies of nitrogen flows or flows of a specific metal (Kytzia and Nathani, 2004; Finnveden and Moberg, 2005). Multi-Criteria Analysis (MCA) is a tool to support the selection of the best combination of outcomes that have different Chapter 4: Modelling the environmental impact - 57 - dimensions, and it can assist in identifying trade-offs between different criteria and finding the best solutions (Wrisberg et al., 2002). Scenario analysis typically results in a set of answers to “What… if” questions illustrating the consequences of a range of alternative decisions (Schwarz, 1997; Pluimers, 2001). Technology assessments are used to analyse technological options to reduce the environmental impact. Some authors defend that many possibilities to reduce the environmental burden of industrial production are present, such as, optimisation of the environmental performance through good housekeeping, end-of pipe techniques, substitution of unfriendly products or by technology innovation (Moors et al., 2005). Cost-effectiveness analysis (CEA) reveals the costs per unit of avoided emission (Rabah, 1999; Pluimers 2001; Klimont et al., 2002). Because of the complexity of most environmental issues, the above briefly described analytical tools are seldom appropriate as a stand-alone tool for analysing environmental issues. In environmental analyses, therefore, often a combination of tools is used to analyse a particular problem. Integrated Assessment (IA) Models typically combine a number of tools. However, these IA Models seldom take a company perspective, but are rather developed to assist policy makers (Alcamo et al., 1990; Carmichael et al., 2004; Ball et al., 2005). Industrial companies may use the systems analysis tools to analyse how to keep the environmental impact of their processes within the limits, set by environmental legislation or how to minimise the impacts in an economically feasible way. This is not a simple task, the tools mentioned above are not, by themselves, appropriate to analyse an environmental problem at the company scale. For instance, LCA is typically developed for the analysis of a product, but not of a company. Industrial companies use EPIs, but the result is a long list of emission estimates that do not give an answer to the question of what the overall environmental performance is, or what the best way is to reduce these emissions. MCA is a useful tool to assess the overall environmental performance, but in itself not easily applied at the company level when the set of emission estimates is not consistent with the structure of the MCA. Scenario analysis is usually applied to investigate trends at the sector or national level, but not often at the company level, due to the site-specific information that would be needed for that. From the above it may be clear that there is a need for decision support systems, to help industrial management to decide on environmental control options for their particular plant. The purpose of this study is therefore to develop a model to analyse options to reduce the environmental impact of aluminium die casting. This model will take a company perspective, so that it can be used as a decision support tool for the environmental management. It will allow the plant management to decide on the environmental strategy to follow. We refer to our model as MIKADO: Model of the environmental Impact of an Aluminium Die casting plant and Options to reduce this impact. In the next sections, we will first describe the MIKADO approach, the model parameters and activities of an aluminium die casting plant. The later section will present the result of a sensitivity analysis, showing the sensitivity of the model results to uncertainties in selected parts of the model. Chapter 4: Modelling the environmental impact - 58 - 4.2. Model Description 4.2.1. Model Design and Structure The MIKADO structure and the modelling approach are based on the work of Van Langen who has developed object-oriented software for designing processes (DESIRE) (Van Langen, 2002). This software, and the language used in it, provides a structure and a grammar to define objects, objects’ properties and methods at multiple layers. The language has been developed to allow to model processes. Van Langen shows that his approach can be used in designing models for estimating the emissions from industrial processes. One of Van Langen’s case studies deals with an emissions inventory model developed as a prototype system for an environmental inventory of brick and tile fabrication in the Netherlands (Van Langen, 2002). The model described in this chapter (Chapter 4) is an application of this approach and uses as an interface to model user a software tool called EstimatER developed by the European Topic Centre on Air and Climate Change to analyse and assess alternative pollution reduction options (ETC/ACC, 2001). The basic “object” in MIKADO is a process. The object covers the full process and has information and material exchange with the environment. The material exchanges of the process with its environment consist of: a) the raw material, energy and any other subsidiary materials needed for the process and, b) the output in terms of the products and any environmental pressures that might be caused by the process. The information exchange of the process with its environment concerns the activity rate. Whether this information about the activity rate is an input or an output is a matter of perspective: if we are interested in managing the process, the activity rate could be seen as an input to the object. If, on the other hand, the object is to describe a process with an endogenous mechanism to run it, it can be regarded as an output. The latter will mainly occur in dynamic models of processes that contain positive or negative feedback loops. In our approach we aim for a model that is to be used by the plant’s management. Therefore, we subdivide the object process into several sub-processes or even, sub-sub-processes. Typically, a process in a plant can be decomposed into a series of consecutive and possibly parallel sub-processes or even sub-sub-processes that form the production line. This nested approach is useful in further specifying the process and process characteristics. In this study we are aiming for a steady-state model that describes the environmental pressures caused by a process. We therefore will regard the information about the process activity rate as an input to the object. This rate can be expressed in different ways: it could be related to one of the inputs in the system or the required outputs. The choice will depend on the type of process modelled. In this study, the process is the production of die cast aluminium car parts. To this end an existing small/medium size enterprise located in the northern part of Portugal served as a case study. Since we are building this model to be used from a company’s management’s perspective, the model considers a production rate of approximately 3000 tons of aluminium die casting products as the model driver. In the model, the production rate of the process is used to calculate all necessary inputs, all outputs and all environmental pressures. Obviously, the exact functions describing the dependence of the inputs, outputs and pressures from the production are determined by the characteristics of the process. The model Chapter 4: Modelling the environmental impact - 59 - allows for manipulation of such process characteristics to implement possible reduction options influencing the environmental problems. By manipulating the process characteristics all functions might change. The DESIRE approach, as implemented in the user interface tool, provides the functionality for these manipulations. The objects within our model structure are nested. This nested structure allows for describing the process characteristics for different sub-process (or sub-sub-processes). This is schematically presented in Figure 4.1 for the case plant on aluminium die casting. The production line of the die casting process consists of the following subprocesses: 1) Melting, 2) Casting, 3) Finishing. The system also includes as subprocesses: 4) Internal Transport and 5) Auxiliary Burners. These are considered subprocesses that are independent of the annual production rate. In addition, the company owns two wastewater treatment plants that are part of the die casting production line. These plants treat liquid effluents from Casting and Finishing. The sub-processes of the production line (Melting, Casting and Finishing) are connected in series since the output from one sub-process is used as an input for the next one. The alloy entering the process includes ingots and alloy recycled internally. The molten alloy, output from the Melting, feeds the Casting sub-process, yielding the raw products. The raw products, in turn, are finished and leave the system as final products. A small part of the aluminium leaves the system as emissions, either to air, to water or as solid waste. Obviously aluminium is not the only resource flowing through the system. Energy is needed to melt the alloy, a range of subsidiary materials is needed for many different purposes and investments, and operation costs need to be paid. In the model, all of these are derived from the production rate. Chapter 4: Modelling the environmental impact - 60 - Figure 4.1. The aluminium pressure die casting plant’s production line. The figure includes the sub-processes and sub-sub-processes that we included in our model (see also Table 4.1). The scheme includes the alloy mass flow throughout the production line. Part of the alloy mass flows exiting Casting and Finishing are recycled internally. For simplification, these flows of recycled alloy are excluded from the figure. MIKADO has been structured so that each sub-process receives all inputs from the earlier sub-processes it needs to deliver the (semi)products. Wherever needed, a further detail in the model is defined by decomposition of a sub-process into sub-sub- processes. The overall structure in terms of sub-sub-process of the die casting plant is described in detail in the next sections. In summary, the model structure is based on the mass flows through the successive steps in the production line giving rise to environmental problems. The inputs to MIKADO at the level of the aluminium die casting production line include, beside the subsidiary materials and energy, the alloy mass flows as raw products use (ingots) and/or alloy mass recycled within the production line. In p uts Casting • Holding Furnaces • Pressure Die Casting Finishing • Trimming • Surface Treatment: • Grinding, • Shot Blasting, • Tumbling. • Cleaning and Degreasing Molten allo y Raw products Products Out p uts PROCESS Env i ronmental Problems Meltin g Chapter 4: Modelling the environmental impact - 61 - In addition, some of the costs of process operation and investments are considered. The outputs of MIKADO include products, semi-products, emissions of pollutants, waste and liquid effluents, alloy mass to be recycled within the process and the costs of emission control. Moreover, the environmental performance of the plant is assessed in terms of one overall indicator. 4.2.2. Model Formulation The MIKADO’s core is formed by conservation of aluminium alloy mass throughout the production line. The alloy mass flow presented in the model is determined by the production rate. The production rate and the mass of alloy leaving the system at each sub-process level are readily available from the company managers. But, within the MIKADO structure, the activity data used refer essentially to the alloy mass flow entering each sub-process, so the raw data supplied was converted in order to refer to the tonnage of aluminium alloy mass inputs per year into each sub-process. So, the model calculates the alloy input needs at each sub-sub-process level, for the production rate, by using the values of alloy mass emissions leaving the sub-sub-process and also the amount of alloy mass recycled internally. In the manufacturing process the losses of aluminium alloy during the process are emitted to the air or leave the system as solid wastes or liquid effluents. These average values are company specific and were made available by the company managers. The alloy losses occur in all the sub-processes from the production line. The losses in the sub-process Melting are due to air emissions (0.04% of the mass of alloy inputs) and aluminium dross (0.72 % of the mass of alloy inputs). In the sub-process Casting a small amount is lost as air emissions (0.0005% of the alloy entering the Casting sub-process). In addition, in the Casting process the runners and biscuits (40% of the alloy entering Casting) and scrap that is internally recycled (6% of the alloy entering Casting) are produced. Finally, the sub-process Finishing produces aluminium burrs (4% of the alloy entering Finishing) and scrap to be internally recycled (7.5% of the alloy entering Finishing). All these losses are obviously compensated by the approximately 6% higher input of ingots as compared with finished products. The alloy mass losses in the liquid effluents are neglected in this study because aluminium losses value less than 0.001% of the mass of input alloy. The aluminium mass flow diagram is presented in Figure 4.2 for the annual production rate (approximately 3000 tons of aluminium die casting products). The figure moreover, includes the alloy mass leaving the system for the sub-processes: 1) Melting, 2) Casting and 3) Finishing. From the figure, it is also clear that the molten alloy entering subprocess Melting is at least twice the production rate, this is due to the fact that the alloy mass includes the alloy recycled internally and the ingots of aluminium alloy. Thus, the shot weight (the shot is the semi-product from Casting that includes the final products plus the excess of materials (runners and biscuits) needed to allow the molten metal to fulfil the die casting moulds) consists, for the specific company, of a mixture of approximately 50% ingot and 50% internal recycled aluminium alloy. Chapter 4: Modelling the environmental impact - 62 - Figure 4.2. Alloy mass flow on the existing aluminium pressure die casting plant. Table 4.1 shows the alloy mass input in each sub-sub-process and the related activities (materials or energy usage) at the sub-sub-process level for the industrial plant. The materials and energy inputs are identified for each sub-sub-process level and refer to the alloy mass flow entering each sub-process. For instance, the amount of natural gas used in the sub-process Melting is directly dependent of the amount of alloy input to the same sub-process. These activities are summed up at the firm level. alloy losses = 1% alloy losses < 0.1% alloy losses = 2% Products (47%)Ingots (50%) Meltin g (p 1 ) Castin g (p 2 ) Finishin g (p 3 ) Recycled (50%) Chapter 4: Modelling the environmental impact - 63 - Table 4.1. Alloy mass in-flow (AL) and type of activities (α) by sub-sub-process (pij). Based on an existing aluminium pressure die casting plant and assuming no reduction options implemented on the plant. Activities Process (p) sub-process (pi) sub-sub-process (pij) AL α Melting (i=1) Melting Annual mass of alloy input to Melting Desoxidation agent Degassing Flux Natural gas Holding Furnaces Annual mass of alloy input to Holding Furnaces Ceramic lining Casting (i=2) Pressure Die Casting a) Annual mass of alloy input to Pressure Die Casting Mould release agent Water Hydraulic oil Tip Lubricant Other Oils Antifoam Sodium hydroxide Polyelectrolyte Flocculation agent Trimming Annual mass of alloy input to Trimming --- Grinding Annual mass of alloy input to Grinding --- Shot Blasting Annual mass of alloy input to Shot Blasting Steel shot Surface Treatment Tumbling b) Annual mass of alloy input to Tumbling Water Ceramic abrasives Splitting agent Finishing (i=3) Cleaning and Degreasing a) Annual mass of alloy input to Cleaning and Degreasing Water Detergent Antifoam Sodium hydroxide Polyelectrolyte Flocculation agent Forklift Trucks on Diesel (I and II) Diesel Internal Transports c) (i=4) Forklift Truck on LPG --- LPG Oxyacetylene burners Acetylene and Oxygen Die Casting company Auxiliary Burners c) (i=5) Butane burners --- Butane a) Both liquid effluents leaving the sub-sub-processes Pressure Die Casting and Cleaning and Degreasing are treated in the same wastewater treatment plant. Thus, some of the agents (activities, such as: antifoam, polyelectrolyte, etc. ) used in the treatment plant are allocated to these sub-sub-processes. b) The liquid effluent leaving this process is treated in a specific wastewater treatment plant. The agent needed (splitting agent) is allocated to this sub-sub-process. c) The activities from sub-processes Internal transports and Auxiliary burners are considered to be independent of the annual production rate. Emissions and waste production are calculated by the model as a function of the activity rate (Act), within sub-sub-process (pij). The activity rate measures the use of materials or energy in each sub-sub-process. The emissions are calculated assuming a linear relation between the activity rate and a specific emission of a pollutant (x). The Chapter 4: Modelling the environmental impact - 64 - proportionality constant is called the emission factor (EF). The total emission (Ex) is calculated by summing all emissions of pollutant (x) resulting from the use of all activities (α) in all sub-sub-processes (pij). Equations 1 to 14 are used to calculate the emissions, the activity rates, the environmental impact and the costs. Box 4.1 presents all the equations and describes all the parameters and variables. Box 4.1. Mathematical Formulation of the model (Equations 1 to 14). ( ) ∑∑ = jα x,αij,αpx, ijiji EF*ActE (Equation 1) ijij αijαAF*ALAct = (Equation 2a) ijαALAct ij = (Equation 2b) ijij αα ActAct = (Equation 2c) () ∑ = i i p px,x EE (Equation 3) () ∑∑ ⎟ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎜ ⎝ ⎛ = z z z x xz,px, pWF* NF CF*E M i i (Equation 4) () ∑ = i i p p MM (Equation 5) ∑∑∑ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛− = jατ τx, τx,,αij,τ,αpx, 100 RF100 *EF*ActE ijiji (Equation 6) ( ) ∑∑∑ = jXατ τx,,Xαij,τ,Xαpx, ijiji EF*ActE (Equation 7) ijij Xαijτ,XαAF*ALAct = (Equation 8a) ijτ,XαALAct ij = (Equation 8b) 1)r)((1 r)(1 *r*ICI τ τ lt lt ττ −+ + = (Equation 9) τττ o*ICO = (Equation 10) ( ) ∑∑ += ij αzerocase,αingzerocasep,zerocase ijij P*ActP*ALCV (Equation 11) ()( ) ∑∑∑∑∑ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎣ ⎡ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ + ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ += ijXα XαXα jα ααingp ijijijij P*ActP*ActP*ALCV (Equation 12) () CVCOCIC τ ττ ++= ∑ (Equation 13) zerocasena CCC −= (Equation 14) Chapter 4: Modelling the environmental impact - 65 - Box 4.1 (cont.) Description of parameters and variables present on equations 1 to 14. i px E, = emission of pollutant x within the sub-process pi in (kg of pollutant x /year) or (m3 of pollutant x /year). x = index for type of pollutant emitted such as: metals (Al, Cd, Ni, Pb, Cr, Cu, Zn, ….), CO2, NOx, CO, NMVOC, etc. p = index for process p (Table 4.1). i = index for sub-process from process p (Table 4.1). j = index for sub-sub-process within sub-process i from process p (Table 4.1). α = index for type of activity, referring to energy or materials such as: use of natural gas, mould release agent, water, desoxidation agent, etc. (Table 4.1). α Act = activity rate (Act) expressing the use of a certain material or energy (α) in (unit activity/year) (Table 4.3). A L = aluminium alloy mass inflow in (ton alloy /year). α AF = activity factor (AF) expressing the unit of activity (α) used by the amount of aluminium alloy mass flow in (unit activity/ton alloy) (Table 4.3). x EF , α = emission factor (EF) for pollutant x, related to a certain type of activity (α) in (kg/ unit activity) or (m3/ unit activity) (Table 4.4). x E = total emission of pollutant x in (kg of pollutant x /year) or (m3 of pollutant x /year). i p M = environmental impact from sub-process pi (unitless). z = index for type of environmental impact category: Human toxicity, terrestrial ecotoxicity, global warming, acidification, photochemical ozone formation, abiotic depletion, aquatic toxicity and solid waste production. xz CF , = characterisation factor (CF) for environmental impact category z due to emission of pollutant x (Table 4.8). z NF = normalisation factor (NF) for environmental impact category z (Table 4.9). z WF = weighting factor (WF) for environmental impact category z (unitless) (Table 4.10). M = overall environmental impact (unitless). i pz M, = environmental impact for a specific environmental impact category (z) resulting from sub-process pi (unitless). τ ,x RF = reduction factor for pollutant x due to the reduction option τ in (%) (Table 4.5). τ = index for reduction option (Table 4.2). τα , Act = activity rate (Act) related with a certain type of activity (α) that may change by the reduction option τ in (unit activity/year) (Table 4.5). τα ,,x E F = emission factor (EF) for pollutant x, related to a certain type of activity (α), that may change by the reduction option τ in (kg/ unit activity) or (m3/ unit activity). α X = index for type of extra activity (Xα), induced by the reduction option τ and referring to energy or materials such as: use of gas N2, oxygen, powder agent or waste to be disposed (Table 4.5). τα ,X Act = activity rate (Act) expressing the use of an extra material or energy (Xα), induced by the reduction option τ in (unit activity/year) (Table 4.5). τα ,,xX E F = emission factor (EF) for pollutant x, related to a certain type of an extra activity (Xα) induced by the reduction option τ in (kg/unit activity) or (m3/ unit activity). Xα AF = activity factor (AF) expressing the unit of an extra activity (Xα) used by the amount of aluminium alloy mass flow in (unit activity/ton alloy). τ CI = annualised capital cost due to the option τ in (k€/year) (Table 4.6). τ I = investment due to option τ in (k€) (Table 4.6). r = interest rate in (fraction/year) (r=0.07) (USEPA, 2002). τ lt = lifetime of reduction option τ in (years) (Table 4.6). τ CO = fixed costs for reduction option τ in (k€/year) (Table 4.6). τ o = fraction of investment indicating the fixed costs for reduction option τ in (fraction/year) (Table 4.6). Chapter 4: Modelling the environmental impact - 72 - 4.3.2. Emission factors The emission factor describes the relation between the activity rate and the emission for a specific pollutant. Emission factors are calculated from annual activity rates for each sub-process. There are different ways to quantify emissions. These include direct measurements, mass balance calculations, process based modelling and the emission factor approach (Frey and Small, 2003). The emission factor approach is the simplest one and typically used in environmental studies of economic sectors (e.g. Pluimers, 2001; Winiwarter and Schimak, 2005) or by country (e.g. IPCC Guidelines, 1997; Zárate et al., 2000). The emission factor calculation performed here differs in scope in terms of process and location from the literature investigated. The emission factors presented in this chapter are process specific and are mainly derived from average emission measurements carried out at the facility in combination with the known alloy flows, energy consumption and the use of subsidiary materials. When no measures were available, emission factors were then estimated based on mass balance calculations, specific literature data or provided by suppliers. The company’s suppliers made materials characteristics available. The emission factors for each sub-sub-process are presented in Table 4.4. Chapter 4: Modelling the environmental impact - 73 - Table 4.4. Emission Factors (EF) by sub-process (pij) referred to the tonnage of molten alloy used in each sub-sub-process. For sub-processes Internal Transport and Auxiliary Burners, EF is referred to the annual fuel use. Assuming no reduction options implemented. sub-process (pi) sub-sub-process (pij) Pollutant (x) xij ij EF , , α Emission Factor Units Melting Melting 9 Aluminium 9 Cadmium 9 Nickel 9 Lead 9 Chromium 9 Copper 9 Hydrogen Fluoride 9 Hydrogen Fluoride 9 Aluminium dross 9 Aluminium dross 9 CO 9 CO2 9 NOx 9 NMVOC 0.3945 a) 0.000196 a) 0.000151 a) 0.000947 a) 0.000124 a) 0.00168 a) 0.0534 b) 0.224 b) 0.949 c) 0.788 c) 30.03 d) 65.11 d) 186 d) 2.1 d) kg / ton molten alloy kg / ton molten alloy kg / ton molten alloy kg / ton molten alloy kg / ton molten alloy kg / ton molten alloy kg / kg degassing flux kg/kg desoxidation agent kg / kg degassing flux kg/kg desoxidation agent g/GJ kg/GJ g/GJ g/GJ Holding Furnaces 9 Aluminium 9 Zinc 9 Lead 9 Chromium 9 Copper 9 Iron 9 Ceramic lining wasted 0.00489 a) 0.00763 a) 0.000611 a) 0.000458 a) 0.00366 a) 0.0370 a) 0.237 e) kg/ton casted alloy kg/ton casted alloy kg/ton casted alloy kg/ton casted alloy kg/ton casted alloy kg/ton casted alloy kg/ton casted alloy Casting Pressure Die Casting 9 NMVOC 9 Liquid effluent 9 Sludge 9 Oils and grease 0.00892 a) 0.76 f) 7 5 kg/l mould release agent m3 / ton alloy kg/ton casted alloy kg/ton casted alloy Shot Blasting 9 Steel Shot 0.27 c) kg/kg steel shot Surface Treatment Tumbling 9 Ceramic abrasives 9 Liquid effluent 9 Sludge 0.43 c) 0.000064 f) 0.07 kg/kg ceramic abrasives m3 / ton alloy kg/ton alloy Finishing Cleaning and Degreasing 9 Liquid effluent 9 Sludge 9 Oils and grease 0.0074 f) 0.07 0.05 m3 / ton alloy kg/ton alloy kg/ton alloy Fork lift trucks on Diesel (I) 9 CO2 9 NOx 9 CO 9 Particulates 9 SO2 9 NMVOC 76.92 g) 0.54 g) 0.08 g) 0.01 g) 0.11 g) 0.03 g) kg/GJ kg/GJ kg/GJ kg/GJ kg/GJ kg/GJ Fork lift trucks on Diesel (II) 9 CO2 9 NOx 9 CO 9 Particulates 9 SO2 9 NMVOC 95.69 g) 0.64 g) 0.15 g) 0.02 g) 0.13 g) 0.07 g) kg/GJ kg/GJ kg/GJ kg/GJ kg/GJ kg/GJ Internal Transport Fork lift trucks on LPG 9 CO2 9 NOx 9 CO 9 NMVOC 67.48 g) 1.67 g) 0.02 g) 0.02 g) kg/GJ kg/GJ kg/GJ kg/GJ Oxyacetylene burners 9 CO2 67.80 h) kg/GJ Auxiliary Burners Butane burners 9 CO2 9 NOx 9 CO 9 NMVOC 9 Particulates 65.41 i) 0.0688 i) 0.0096 i) 0.0027 i) 0.0027 i) kg/GJ kg/GJ kg/GJ kg/GJ kg/GJ a) Emissions to air. Emission factor derived from annual average pollutant concentrations (mg/m3) measured at the industrial plant. Pedro (2005). Personal communication. b) Emission to air. Emission factor derived from fluorine contents of desoxidation agent and degassing flux. Pedro (2005). Personal communication c) Solid waste. Emission factor derived from the composition of desoxidation agent and degassing flux used. Pedro (2005). Personal communication d) Emission to air. Average value for emission factor related with natural gas use, derived from range present. EMEP/CORINAIR (2004). e) Solid waste. Emission factor from EIPPCB (2005). f) Liquid effluent. Implied emission factor; the model includes a mass balance for water. The emission factor is derived from annual water consumption and losses at the industrial plant. Pedro (2005). Personal communication. g) Emissions to air. Emission factor made available from Salvador Caetano, S.A and the annual forklift workinh hours. Monteiro (2004). Personal communication. h) Emission to air. Emission factor derived from annual acetylene consumption. Pedro (2005). Personal communication. i) Emission to air. Emission factor related with butane use. Pedro (2005). Personal communication. Chapter 4: Modelling the environmental impact - 74 - 4.3.3. Reduction Factors and Extra Activities The implementation of reduction options leads to a decrease in the pollution. Several reduction options were defined for the die casting plant; they are process specific and were proposed by industrial facility managers or found in specialised literature. Table 4.2 (section 4.2.2.) gives an overview of the 18 pollution prevention options included in our model. These reduction options are either add-on technologies, the replacement of an existing technique or a change in process operation. These reduction options reduce the original emission factors, change the activity rates (such as: energy consumption) or in the case of add-on technologies, might add a reduction factor responsible for the pollution abatement. When add-on technologies are used, the reduction factors of one or more pollutants are well known and available in literature. Table 4.5 includes the values of reduction factors (RF) per reduction option using an add-on technology. In addition, the reduction options may also influence the activity rate itself either by altering the amount of materials or energy used or by introducing an extra activity in the industrial process or through a combination of both situations. Table 4.5 also includes the changes in the activity rates (Actα) and extra activity rates (ActXα) relative to the zero case when an individual reduction option (τ) is implemented. The table only presents the activities and extra activities affected by each individual reduction option. Chapter 4: Modelling the environmental impact - 75 - Table 4.5. Reduction factors (RFx,τ), reduction in activity rates (Actα) and extra activity rates (ActXα) caused by reduction options. See Chapter 3 (section 3.3.3.) for a detailed description of the reduction options. Reduction factor (RFx,τ) Reduction in activity rates (Actα) Reduction Options (τ) x = Al x=Cd x=Ni x=Pb x=Cr x =Cu x=HF x =NMVOC x=Zn x=Fe α=desoxidation agent α=degassing flux α=natural gas α=mould release agent α=water Melting_FF_RA a) 99.9% 99% 99% 99% 99% n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_FF_PJ b) 99.9% 99% 99% 99% 99% n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_FF_MS c) 99.9% 99% 99% 99% 99% n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_WS_IP d) 99% 99% 99% 99% 99% 99% 99% n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_WS_SC e) 99% 99% 99% 99% 99% 99% 99% 95% n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_GA n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 100% h) n.e. n.e. n.e. n.e. Melting_IS n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 100% h) n.e. n.e. n.e. Melting_CM n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 58% k) n.e. n.e. Melting_AE n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 2% l) n.e. n.e. Melting_OF n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 4% l) n.e. n.e. Casting_WS_PB f) 95% n.e. n.e. 95% 95% 95% n.e. 99% 95% 95% n.e. n.e. n.e. n.e. n.e. Casting_WS_SC e) 99% n.e. n.e. 99% 99% 99% n.e. 95% 99% 99% n.e. n.e. n.e. n.e. n.e. Casting_nMA n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 100% h) n.e. Casting_PA n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 100% h) 100% h) Casting_rRR n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 16% p) 16% p) 16% p) 16% p) 16% p) Casting_rSR n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 5% p) 5% p) 5% p) 5% p) 5% p) Finishing_rSR n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 5% q) 5% q) 5% q) 5% q) 5% q) IT_eFL n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. = no effect. a) to f) from USEPA (2002): a) EPA-452/F-03-026. b) EPA-452/F-03-025. c) EPA-452/F-03-024. d) EPA-452/F-03- 012. e) EPA-452/F-03-016. f) EPA-452/F-03-015. g) Estimated based on the amount of avoided air emissions. h) Assuming a full replacement of the agents currently used. Melting_GA: use of a granular agent replacing the currently used. Melting_IS: use of gas N2, replacing the solid agent. Casting_nMA: use of alternative mould release agent. Casting_PA: replace the liquid agent and water by a solid powder mould release agent. IT_eFL: use of electric forklift trucks replacing the fuelled (Diesel and LPG) currently used. i) Estimated from Foseco (2002). Personal communication. j) Estimated from Brown (1999) and EIPPCB (2005). k) “Estimated” based on the average furnace thermal efficiency for the type of furnace used in the plant. From EIPPCB (2005). l) “Estimated” based on the latent heat present in the exhaust gases. n) “Estimated” based on the reduction of natural gas use reported to the plant unabated situation. o) “Estimated” based on the reduction of mould release agent use reported to the plant unabated situation. p) “Estimated” based on use reported to the number of die casting shots produced annually by the company and the indication from product use from Klüber (2005). q) “Estimated” based on the reduction of the aluminium alloy that is recycled internally. This option affects the materials and energy used on the sub-processes Melting and Casting. r) “Estimated” based on the reduction of the aluminium alloy that is recycled internally. This option affects the materials and energy used on the sub-processes Melting, Casting and Finishing. Chapter 4: Modelling the environmental impact - 76 - Table 4.5. (cont.) Reduction in activity rates (Actα)Extra activity rates (ActXα) caused by reduction options Reduction Options (τ) α= hydraulic oil, tip lubricant and other oils α=steel shot α=ceramic abrasives α=detegent α= diesel α= LPG Xα= waste to disposal Xα=granular agnet Xα= gas N2 Xα=oxygen Xα= new mould release agent Xα= powder agent Melting_FF_RA a) n.e. n.e. n.e. n.e. n.e. n.e. 2.5 ton/yr g n.e. n.e. n.e. n.e. n.e. Melting_FF_PJ b) n.e. n.e. n.e. n.e. n.e. n.e. 2.5 ton/yr g n.e. n.e. n.e. n.e. n.e. Melting_FF_MS c) n.e. n.e. n.e. n.e. n.e. n.e. 2.5 ton/yr g n.e. n.e. n.e. n.e. n.e. Melting_WS_IP d) n.e. n.e. n.e. n.e. n.e. n.e. 4 ton/yr g) n.e. n.e. n.e. n.e. n.e. Melting_WS_SC e) n.e. n.e. n.e. n.e. n.e. n.e. 4 ton/yr g) n.e. n.e. n.e. n.e. n.e. Melting_GA n.e. n.e. n.e. n.e. n.e. n.e. n.e. 6400 kg/yr i) n.e. n.e. n.e. n.e. Melting_IS n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 403 m3/yr j) n.e. n.e. n.e. Melting_CM n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Melting_AE n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 4.8E+05 m3/yr m) n.e. n.e. Melting_OF n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 1.6E+06 m3/yr m) n.e. n.e. Casting_WS_PB f) n.e. n.e. n.e. n.e. n.e. n.e. 0.5 ton/yr g n.e. n.e. n.e. n.e. n.e. Casting_WS_SC e) n.e. n.e. n.e. n.e. n.e. n.e. 0.5 ton/yr g n.e. n.e. n.e. n.e. n.e. Casting_nMA n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 28 m3/yr n) n.e. Casting_PA n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. 4020 kg/yr o) Casting_rRR 16% p) 16% p) 16% p) n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Casting_rSR 5% p) 5% p) 5% p) n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. n.e. Finishing_rSR 5% q) 5% q) 5% q) 5% q) 5% q) 5% q) n.e. n.e. n.e. n.e. n.e. n.e. IT_eFL n.e. n.e. n.e. n.e. 100% h) 100% h) n.e. n.e. n.e. n.e. n.e. n.e. n.e. = no effect. a) to f) from USEPA (2002): a) EPA-452/F-03-026. b) EPA-452/F-03-025. c) EPA-452/F-03-024. d) EPA-452/F-03- 012. e) EPA-452/F-03-016. f) EPA-452/F-03-015. g) Estimated based on the amount of avoided air emissions. h) Assuming a full replacement of the agents currently used. Melting_GA: use of a granular agent replacing the currently used. Melting_IS: use of gas N2, replacing the solid agent. Casting_nMA: use of alternative mould release agent. Casting_PA: replace the liquid agent and water by a solid powder mould release agent. IT_eFL: use of electric forklift trucks replacing the fuelled (Diesel and LPG) currently used. i) Estimated from Foseco (2002). Personal communication. j) Estimated from Brown (1999) and EIPPCB (2005). k) “Estimated” based on the average furnace thermal efficiency for the type of furnace used in the plant. From EIPPCB (2005). l) “Estimated” based on the latent heat present in the exhaust gases. n) “Estimated” based on the reduction of natural gas use reported to the plant unabated situation. o) “Estimated” based on the reduction of mould release agent use reported to the plant unabated situation. p) “Estimated” based on use reported to the number of die casting shots produced annually by the company and the indication from product use from Klüber (2005). q) “Estimated” based on the reduction of the aluminium alloy that is recycled internally. This option affects the materials and energy used on the sub-processes Melting and Casting. r) “Estimated” based on the reduction of the aluminium alloy that is recycled internally. This option affects the materials and energy used on the sub-processes Melting, Casting and Finishing. Chapter 4: Modelling the environmental impact - 77 - 4.3.4. Investments and Variable Costs The total annual costs (C) (Equation 13 of Box 4.1) are calculated based on fixed and variable costs. When a reduction option is used, investments are needed. The total annual fixed costs are calculated by two components: the investment costs and the fixed operational costs. Variable costs are related to materials and energy uses and to the production rate. Table 4.6 shows an overview of cost-related parameters per reduction option. Table 4.7 summarises the unit prices for the activity data. Table 4.6. Cost parameters (I, lt, oj) used for the calculation of the overall annual fixed costs (CI+CO). Model results for the variable costs (CV). These fixed and variable costs of reduction options are applicable to the die casting facility. The variable cost for zero case (CVzerocase) is 5315 k€/year. See Chapter 3 (section 3.3.3.) for a detailed description of the reduction options. Reduction Options (τ) Investment ( I ) in (k€) Lifetime (lt) in (years) Annualised Capital Cost (CI) m) in (k€/year) Fraction of investments (oτ) n) in (fraction/year) Fixed cost (CO) o) in (k€/year) Variable cost (CV) p) in (k€/year) Fabric Filter. Reverse-air type 675 a) 20 a) 64 0.03 20 5315 Fabric Filter. Pulse-Jet type 206 b) 20 b) 19 0.03 6 5315 Fabric Filter. Mechanical Shaker type 574 c) 20 c) 54 0.03 17 5315 Wet Scrubber. Impingement- Plate type 86 d) 15 d) 9 0.03 3 5316 Wet scrubber. Spray-chamber type 49 e) 15 e) 5 0.03 1 5316 Granular desoxidation agent 0 0 0 0 0 5315 Impeller station using N2 55 f) 10 l) 8 0.03 2 5364 Compact metal loading in furnaces 140 g) 10 l) 20 0.04 6 5162 Air enrichment with oxygen (30%O2) 0 0 0 0 0 5374 Oxyfuel firing (100%O2) 170 h) 10 l) 24 0.03 5 5510 Wet Scrubber. Packed-Bed type 1396 i) 15 i) 153 0.03 42 5315 Wet scrubber. Spray-chamber type 155 e) 15 e) 17 0.03 5 5315 New mould release agent 0 0 0 0 0 5257 Powder agent 220 j) 10 l) 31 0.04 9 5421 Reduce runners’ mass 1140 k) 10 l) 162 0.04 46 5226 Reduce scrap rate 0 0 0 0 0 5285 Reduce scrap rate 0 0 0 0 0 5279 Use electric forklift trucks 57 g) 10 l) 8 0.02 1 5267 a) to e) and i) USEPA Air pollution control technology fact sheet - EPA-CICA Fact Sheet. a) EPA-452/F-03-026; b) EPA-452/F-03-025; c) EPA-452/F-03-024; d) EPA-452/F-03-012; e) EPA-452/F-03-016; i) EPA-452/F-03-015. f) EIPPCB (2005). g) Pedro (2005). Personal communication. h) Praxair (2005). Personal communication. j) Klüber (2005); k) INETI (2000); l) Assumed to be 10 years; m) The annualised capital costs (CI) is calculated by Equation 9 (see Box 4.1). n) USEPA (2002) and Klimont et al. (2002). o) The fixed cost (CO) is calculated by equation 10 (see Box 4.1). p) The variable costs (CV) is calculated by equation 12 (see Box 4.1), and include costs of all relevant inputs (5315 k€/year in the zero case). Chapter 4: Modelling the environmental impact - 78 - Table 4.7. Price of aluminium ingot, activities (Pα) and extra-activites (PXα). The aluminium die casting plant provided the prices presented for the zero case. The prices for material use on the reduction options were in some cases provided by the case plant but mostly provided by the die casting industry suppliers. Parameter Price unit References Aluminium ingot 1.54 €/kg Pedro (2005). Personal communication. Antifoam 2.1 €/liter Pedro (2005). Personal communication. Ceramic abrasives 1.02 €/kg Pedro (2005). Personal communication. Degassing Flux 2.8 €/kg Pedro (2005). Personal communication. Desoxidation Agent 1.32 €/kg Pedro (2005). Personal communication. Detergent 1.86 €/liter Pedro (2005). Personal communication. Diesel 0.91 €/liter GALP (2004). Personal communication. Flocculation agent 0.25 €/liter Pedro (2005). Personal communication. Gas N2 127 €/m3 Praxair (2005). Personal communication. Granular agent 1.3 €/kg Foseco (2005). Personal communication Hydraulic oil 1.05 €/liter Pedro (2005). Personal communication. LPG 1.23 €/kg GALP, 2004. Personal communication. Mould release agent 1.53 €/liter Pedro (2005). Personal communication. Natural gas 0.33 €/m3 Pedro (2005). Personal communication. New mould release agent 1.7 €/liter Pedro (2005). Personal communication. Other Oils 0.74 €/liter Pedro (2005). Personal communication. Average value of three different oils used in the die casting machines. Oxygen 0.13 €/m3 Praxair (2005). Personal communication. Polyelectrolyte 2.84 €/liter Pedro (2005). Personal communication. Powder agent 55 €/kg Klüber (2005). Personal communication. Sodium hydroxide 0.17 €/liter Pedro (2005). Personal communication. Splitting agent 3.37 €/kg Pedro (2005). Personal communication. Steel Shot 0.59 €/kg Pedro (2005). Personal communication. Tip lubricant 1.37 €/liter Pedro (2005). Personal communication. Waste to disposal (dust) 200 €/ton Pedro (2005). Personal communication. Waste to disposal (aluminium dross) 37 €/ton Pedro (2005). Personal communication. Waste to disposal (oils ands grease) 51 €/ton Pedro (2005). Personal communication. Waste to disposal (sludge) 220 €/ton Pedro (2005). Personal communication. Water 1.5 €/m3 Pedro (2005). Personal communication. Chapter 4: Modelling the environmental impact - 79 - 4.3.5. Environmental impact assessment Potential environmental impacts are assessed for depletion of natural resources, air emissions, solid wastes and liquid effluents resulting from the industrial plant. The potential environmental impact categories or environmental problems (z) resulting from the operation of the industrial process, include: • Human toxicity • Terrestrial ecotoxicity • Global warming • Acidification • Photochemical ozone formation • Abiotic depletion • Aquatic toxicity • Solid waste production. For the environmental problem Abiotic depletion the consumption of natural gas is used as an indicator for the use of non-renewable resources. The impact category aquatic toxicity indicates the amount of liquid effluent produced by the company and solid waste is an indicator of the amount produced. Following the current practice in Life Cycle Assessment and Multi-Criteria Analysis, the environmental impact assessment in MIKADO includes three steps (Pennington et al., 2004): 1) Characterisation, 2) Normalisation and 3) Weighting. All emissions contributing to a specific environmental problem were aggregated in one single value by multiplication by a characterisation factor (CF). For the Characterisation step the methodology of Guinée (Guinée et al., 2002) was used. CF expresses the relative contribution of each pollutant to a specific environmental problem. Table 4.8 shows the CF used in the characterisation step. The emissions are quantified in kilograms 1.4 dichlorobenzene (DCB) for human toxicity and ecotoxicity, in kilograms of antimony for natural resources depletion, in kilograms CO2 for global warming, in kilograms of SO2 for acidification, in kilograms of ethylene for ozone precursors. In the Normalisation step we divide the potential impact for each environmental problem (value from Characterisation) by the impact score for a reference situation. This way, the relative contribution of the process is related to a reference situation (region, country or the whole world). The normalisation factors (NF) applied here (see Equation 4 on Box 4.1) use the Western Europe 1995 as a reference situation (Huijbregts et al., 2003) (Table 4.9). Exceptions are the NF for ATP and SW; these values are developed from emissions from Western European territory in the period 1990-1994 (Blonk, 1997). For the NF for solid waste we choose the maximum value of the range by Blonk (1997). Chapter 4: Modelling the environmental impact - 80 - Four different methods for the weighting were used: a) considering all environmental problems equally important, b) Panel method I (Kamp, 2005) c) Panel method II (Kortman et al., 1994) and d) Distance to target method (Goedkoop, 1995). In addition, the model user may define the set of valuation factors for each environmental problem. Table 4.10 lists the weighting factors used. Table 4.8. Characterisation Factors (CF) per pollutant (x) for each environmental impact category (z). Environmental impact category (z) Pollutant (x) CF a) CF Units NOx 1.20E+00 Particulates 8.20E-01 Cd 1.50E+05 Ni 3.50E+04 Pb 4.70E+02 Cr 3.40E+06 Cu 4.30E+03 Zn 1.00E+02 HF 2.90E+03 NMVOC 1.40E+04 Human Toxicity Potential (HTP inf) b) SO2 9.60E-02 kg 1.4-DCB eq. / kg Abiotic Depletion Potential (ADP) Ultimate reserves and extraction rates Natural gas 1.87E-02 kg antimony eq. / m3 Global Warming Potential (GWP 100) c) CO2 1.00E+00 kg CO2 eq. / kg NOx 5.00E-01 SO2 1.20E+00 Acidification Potential Average Europe (AP Huijbregts, 1999; average Europe total, A&B) HF 1.60E+00 kg SO2 eq. / kg Cd 8.10E+01 Ni 1.20E+02 Pb 1.60E+01 Cr 3.00E+03 Cu 7.00E+00 Zn 1.20E+01 HF 2.90E-03 Ecotoxicity Potential terrestrial (ECP inf) d) NMVOC 2.50E-03 kg 1.4-DCB eq. / kg CO 2.70E-02 NOx 2.80E-02 NMVOC 3.73E-01 Photochemical ozone formation potential (POCP Jenkin & Hayman, 1999 and Derwent et al. 1998; high NOx) SO2 4.80E-02 kg ethylene eq. / kg Aquatic toxicity (ATP) Not available Solid waste (SW) Not available a) See CML (2002). b) HTP inf. (Time horizon infinite) c) GWP100 (Time horizon = 100 years) d) ECP inf. (Time horizon infinite) Chapter 4: Modelling the environmental impact - 81 - Table 4.9. Normalisation Factors (NF) for Western Europe per environmental impact category (z). The reference situation is assumed to be Western Europe in 1995 (Huijbregts et al., 2003). Environmental impact category (z) NFz unit Human Toxicity Potential (HTP) 7.6E+12 1.4-DCB eq./yr Abiotic Depletion Potential (ADP) 1.5E+10 kg antimony eq./yr Global Warming Potential (GWP) 4.8E+12 kg CO2 eq./yr Solid waste (SW) a) 54E+10 kg/yr Acidification Potential (AP) 2.0E+10 kg SO2 eq./yr Ecotoxicity Potential (ECP) 4.7E+10 1.4-DCB eq./yr Photochemical Ozone Formation Potential (POCP) 8.2E+09 kg ethylene eq. /yr Aquatic toxicity (ATP) b) 4.4E+14 m3 aquatic ecotoxicity /yr a), b) These values are developed for the reference situation : Western European territory in the period 1990-1994 (Blonk, 1997). a) The NF for solid waste was assumed to be the maximum value in the range (9.7– 54*1010) (Blonk, 1997). Chapter 4: Modelling the environmental impact - 88 - For Sensitivity Analysis I the results, presented in Table 4.12, indicate that the MIKADO is most sensitive to variations in the emission factors used to calculate Melting-related emissions of HF, (case SA6, ±5% change in M relative to the zero case), and Casting-related emissions of chromium (SA11, ±5%). The model is also relatively sensitive to the characterisation factor (CF) for chromium (SA16, ±5% change in M relative to zero case). The other emission factors and CF have a relatively small effect (less then 5%) on M. In Sensitivity Analysis II, it was observed that the MIKADO is sensitive to some, but not all parameters. For instance, the calculated value of M appears to be relatively sensitive to the variations of the parameter thermal efficiency, (case SA28: -3 to +5% change in M relative to the situation when the reduction option is implemented). Model results show that the calculated costs of emission control appear to be most sensitive to variations in the thermal efficiency parameter (case SA28) presenting a +1% cost variation. No effects on costs calculations are seen for the other parameters. Sensitivity Analysis III indicates that the calculated overall environmental impact (M) is relatively highly sensitive to changes in the runners’ mass (SA38, changing the value of M by -12% to +17% relative to the zero case). A comparatively minor effect is observed in the scrap rates from Casting (SA39) and Finishing (SA40) and burrs fractions (SA48) (as seen in Table 4.12). In the Combined cases, the changes in the calculated values of M related to the zero case were investigated (Figure 4.5) for Combined Cases 1 and 3. Combined Case 1 shows a - 14 to +16% change in M relative to the zero case. For Combined Case 3 this is -16% to +23%. In Combined Case 2 (Figure 4.6), the results indicate that the model is sensitive showing a change in M of -9 to +20%, when compared with M for the Combined case 2 (default case). Changes in total costs were analysed for the Combined Cases. Model runs show for Combined case 2 cost variation of ±1% when compared with the total costs for the situation when the associated reduction option were used, and for Combined case 3 a cost change of ±3% related with the zero case. Chapter 4: Modelling the environmental impact - 89 - 0.5 0.6 0.7 0.8 0.9 1.0 zero case Combined case 1 (lower) Combined case 1 (higher) Combined case 3 (lower) Combined case 3 (higher) Combined cases 1 and 3 Environmental Impact (M)*10-6 Figure 4.5. The calculated overall environmental impact (M) for two combined cases, in each of which was changed the value of a parameter to a lower or higher value, in order to test the sensitivity of the model results to changes in parameter values. The zero case uses parameter values as presented in earlier sections of this chapter. The valuation method used considers all environmental problems to be equally important. 0.2 0.3 0.4 Combined case 2 (default case) Combined case 2 (lower) Combined case 2 (higher) Combined case 2 Environmental Impact (M)*10-6 Figure 4.6. The calculated overall environmental impact (M) for Combined case 2, in each of which the value of a parameter was changed to a lower or higher value, in order to test the sensitivity of the model results to changes in parameter values. The calculation of M for Combined case 2 (default case) uses the default parameters for the situation when the reduction options were implemented. The valuation method used considers all environmental problems to be equally important. Chapter 4: Modelling the environmental impact - 90 - The results of the sensitivity analysis (Table 4.12) indicate that MIKADO appears to be relatively sensitive to changes associated with alloy mass flow. The parameter SA38 (runners’ mass, this parameter is strictly related with the increase of metal yield) has an effect on the majority of model results (such as alloy mass entering the process and the activities rates), as seen in Table 4.5. The same is true for the case where multiple combinations of parameters are used. Our analysis of the cases used leads to the conclusion that again a variation of the alloy mass flow related parameters (Combined case 3) gives rise to a relatively large change in M and in total costs. Although outside the scope of this study, future studies may add to this partial sensitivity analysis by combining different parameters from SAI to SAIII. In addition, parameters related to other environmental problems other than human toxicity and from sub-processes other than Melting and Casting may be included in a sensitivity analysis. 4.5. Discussion and Conclusions This study describes a model that assesses the potential environmental impact of emissions of environmental pollutants from a small to medium sized plant supplying car manufacturers with aluminium die casting products. The model also includes a number of options for emission reduction, and can be used to calculate their technical potentials to reduce the environmental impact as well as the associated costs. These calculations can be done for individual reduction options, or for combinations of options. Our study takes a company perspective. As a result, our model is an environmental decision support tool, meant to assist the management of the company in deciding on environmental policies. MIKADO can be used to perform scenario analysis to analyse the impact on the environment of different strategies, while taking into account both economical and ecological consequences of decision-making. Our model results indicate that more than 90% of the environmental impact of the company is from the sub-processes Melting and Casting. Moreover, the results indicate that the environmental impact is mostly associated with human toxicity problems (caused by metal emissions, and emissions of ozone precursors), and the abiotic depletion of natural gas. This conclusion is relatively insensitive to the environmental impact assessment methods used. This may not be too surprising since the main compounds released by the plant are metals, including heavy metals, and some volatile organic compounds both contributing to human toxicity problems. The results of the sensitivity analysis indicate that variations in individual model parameters may change the calculated overall environmental impact only to a limited extent. Parameters to which the model is most sensitive include those related with alloy mass flow, and in particular, the ones highly related with the increase of metal yield (as the runners’ mass). As mentioned above, MIKADO takes a company perspective. This is apparent from the choice of (1) model components and (2) system boundaries, as well as (3) the selection of reduction options used in the model. In short, we modelled those parts of the production process that can be influenced by the company management. Below, we will elaborate on this. Chapter 4: Modelling the environmental impact - 91 - First, the processes that are explicitly taken into account in the model are those that can be influenced by the company management to improve the environmental performance. The modelling approach taken acknowledges that the industrial system is primarily influenced by the process owner and therefore takes the process operator point of view. We modelled the production process, by zooming into its sub-process level. An insight into the sub-sub-process level allows for the identification of the intervention needed in terms of managing the environmental performance of the industrial process. This quantitative information may make it possible to prioritise environmental management decisions made by board managers (Wright et al., 1998). The model developed is based on specific data that was made available by the plant that served as a case study in our analysis. Second, the system boundaries are chosen from a company perspective. We follow a limited chain analysis. This implies that we do not perform a full Life Cycle Assessment, but rather limit ourselves to those system elements that can be influenced by company management (Pineda-Henson et al., 2002). This implies, in accordance with other studies, that, in most cases, the system boundaries were set at the gates of the business concerned (Finkbeiner et al., 1998; Zobel et al., 2002; Backhouse et al., 2004). Although the activities outside the site may generate significant environmental impacts the possibilities that the facility has to influence or control them are limited (Zobel et al., 2002). Third, we only included reduction options that can be taken by company management. We limited ourselves to existing and available abatement techniques. The options listed include both add-on techniques and more structural changes in the production process. The latter may best suit the company’s pro-activeness. The method presented here is not a mere pollution oriented approach, since the model user may in the analysis select reduction options not only because of their potential in reducing pollution, but also due to the fact that these options may lead to a double goal on reduction pollution and simultaneously bring a cost benefit to the firm. One of the strengths of MIKADO is the integrated approach that it takes in analysing, simultaneously, all the relevant environmental problems caused by the aluminium die casting plant. The compounds analysed contribute to several environmental problems including human toxicity, terrestrial ecotoxicity, aquatic toxicity, depletion of resources such as natural gas, acidification, global warming, emissions of ozone precursors and solid waste production. The model takes into account that some pollutants contribute to more than one problem, that some sub-processes emit more than one pollutant, and that some reduction options affect more than one pollutant. As a result, an analysis of the impact of the reduction options included in the MIKADO will reveal their integrated effect on the environmental performance of the company. MIKADO is developed in such a way that a user can easily select options to be analysed (tailor-made structure), given the production line as defined in the model. Moreover, the model is transparent and understandable, making it reproducible. The data included in the model refer to consumption of energy and materials and the output data, besides the annual production, refer to the emissions and wastes. The model allows the user to analyse the causal chain of activities at the sub-process level and the associated environmental aspects. Chapter 4: Modelling the environmental impact - 92 - An important feature of MIKADO is that it expresses the environmental performance of the company in one single indicator. This is done on the basis of Multi-Criteria Analysis (Pennington et al., 2004). We argue, in agreement with some authors, that this makes the model interesting for company management (Haes, 2000; Daniel et al., 2004; Krajnc and Glavič, 2005). Our approach is also in line with Olsthoorn et al. (2001) who propose a method for the aggregation of different environmental aspects in one single indicator. However, we realise that this approach includes a valuation step, in which the different environmental problems are weighed, that introduces subjectivity in the model: the question whether one environmental problem is more problematic than another is a political question, not a scientific one. Therefore, we included several valuation approaches in our model. More importantly, MIKADO is designed so that any user can change the valuation factors according to their own judgement (in line with Bengtsson and Steen, 2000). In MIKADO, most emissions are quantified using a simple emission factor approach (IPCC Guidelines, 1997; Sakamoto and Tonooka, 2000; Winiwarter and Schimak, 2005). This implies that emissions are calculated as a function of a certain emission factor and activity rate (Pluimers, 2001). The reduction options may have an effect on the emission factor or on the activity level. The costs of reduction options are calculated as total annual costs. MIKADO allows the user to calculate the overall environmental impacts and the total costs of an individual reduction option or a combined strategy. This approach is appropriate, because it makes it possible to model a complex industrial process in a relatively simple way, while making it possible to take into account all the relevant interrelations between sub-processes, pollutants and reduction options. MIKADO can only be used for scenario analysis, answering “what…if” type questions (e.g. ‘What would the effect on the environment and the costs be if we would implement the following options?…). This model cannot be used for cost optimisation. Nevertheless, we consider the model flexible enough to allow the user to get a good overview of the cost-effectiveness of a large set of different scenarios. This way, the model creates a plausible possibility space that users can explore in order to identify the set of choices and trade-offs that they are willing to accept (in line with Carmichael et al., 2004). Any model approach has its limitations. A weak point in our analysis is that by taking the company perspective, we do not account for the environmental impact of the production of raw materials, nor for the environmental impact of the use of the products after they leave the company. Another limitation of the model is the absence of electricity uses and its related consumption costs. Although it may contribute significantly for the cost assessment of the implementation of some reduction options, these data was not available from the company in a disaggregated level for the situation zero case. However, we consider this an inherent consequence of our modelling approach taken. Another weakness is the impact assessment methodology. More specifically, we used the amount of liquid effluent and the solid waste produced as indicators due to the lack of characterisation factors for aquatic toxicity and solid waste production. Also, the normalisation factors used are, because of lack of data, not based on the local situation, but on Western Europe. This, however, is justified by the fact that it is common practice in multi-criteria and Life Cycle Assessments (e.g. Hertwich Chapter 4: Modelling the environmental impact - 93 - and Hammitt, 2000; Huijbregts et al., 2003; Geldermann and Rentz, 2005). Another limitation of our study is that the model was not validated with independent data, because such data do not exist. However, we based our model on data from a specific plant and were able to simulate the processes in this plant in a satisfactory way. It should also be noted that although the model includes a large number of environmental problems, we do not take into account noise, vibration and odour. The MIKADO approach is simple, but complete. It is based, among other environmental systems analysis tools, on relevant parts of life cycle impact assessment, environmental systems management and Multi-Criteria Analysis, which are welldeveloped tools in covering ecological aspects of decision-making (Finnveden and Moberg, 2005). Although several tools are available for assessment of the environmental performance of industrial processes, these methods may not be able to effectively evaluate the impacts associated with decision-making at the company level, nor provide opportunities to be more pro-active in environmental management (Kolk and Mauser, 2002). We consider our model an important step towards fulfilling the need for a tool that assesses both environmental and economic impacts of organisations. References Alcamo, J., Shaw, R., Hordijk, L., 1990. The RAINS model of acidification. Sciences and Strategies in Europe, In: Alcamo, J., Shaw, R., Hordijk, L., (Eds.), Kluwer Academic Publisher: Dordrecht. Backhouse, C.J., Clegg, A.J., Staikos, T., 2004. Reducing the environmental impacts of metal castings through life-cycle management. Progress in Industrial Ecology Vol. 1 Nos. 1/2/3. Ball, M., Calaminus, B., Kerdoncuff, P., Rentz, O., 2005. Techno-economic databases in the context of integrated assessment modelling. Environmental Modelling & Software 20 1189-1193. Bengtsson, M., Steen, B., 2000. Weighting in LCA – Approaches and Applications. Environmental progress Vol. 19 number 2. Blonk, H., 1997. Three reference levels for normalisation in LCA. In: Blonk, H., (Eds.), Ministry of Housing, Spatial Planning and the Environment (VROM), The Netherlands. Brown, J.R., 1999. Foseco Non-Ferrous Foundryman´s Handbook. In: Brown, J.R., (Eds.), Butterworth-Heinemann, Oxford. CAEF (The European Foundry Association), 2003. Statistics from the CAEF countries members on the production of ferrous and non-ferrous alloys. Data from 1999 to 2003. In: Revista de Fundição 234 Asssociação Portuguesa de Fundição, (Eds.), Portugal (in Portuguese). The statistics from the non-ferrous metals castings industry are also available from http://www.caefeurofoundry.org/library/documents/T28-non-ferrous_37317.pdf, last acceded on 07.11.2006. Carmichael, J., Tansey, J., Robinson, J., 2004. An integrated assessment modeling tool. Global Environmental change 14 171-183. CML, 2002. Characterisation factors for Impact Assessment. Center of Environmental Science, Leiden University. The Netherlands. Available from http://www.leidenuniv.nl/cml/lca2/index.html, last acceded on 06.11.2002. Daniel, S.E., Tsoulfas, G.T., Pappis, C.P., Rachaniotis, N.P., 2004. Aggregating and evaluating the results of different Environmental Impact Assessment methods. Ecological Indicators 4 125-138. Chapter 4: Modelling the environmental impact - 94 - EMEP/CORINAIR Emission Inventory Guidebook, 2004. Combustion in manufacturing industry, (Eds.) European Environment Agency. ETC/ACC, 2001. European Topic Centre on Air and Climate Change - Topic Centre of European Environment Agency. Available from http://airclimate.eionet.eu.int/country_tools/ae/EstimatER.htm, last acceded on 20.12.2006. EIPPCB (European Integrated Pollution Prevention and Control Bureau), 2005. Draft on Reference document on Best Available Techniques in the Smitheries and Foundries Industry, European Commission. Available from http://eippcb.jrc.es/, last acceded on 10.02.2005. Finkbeiner, M., Wiedemann, M., Saur, K., 1998. A Comprehensive Approach Towards Product and Organisation Related Environmental Management Tools. International Journal of LCA 3 (3) 169-178. Finnveden, G., Moberg, A., 2005. Environmental systems analysis tools an overview. Journal of Cleaner Production 13 1165-1173. Frey, H.C., Small, M.J., 2003. Integrated Environmental Assessment, Part I. Estimating emissions. Journal of Industrial Ecology vol.7, number 1. Foseco, 2002. Development, Evaluation and Application of Granular and Powder Fluxes in Transfer Ladles, Crucible and Reverbatory Furnaces. In: Foundry Practice issue 237, Foseco International Limited (Eds.), England. Geldermann, J., Rentz, O., 2004. The reference installation approach for the techno-economic assessment of emission abatement options and the determination of BAT according to the IPPC- directive. Journal of Cleaner Production 12 389-402. Geldermann, J., Rentz, O., 2005. Multi-criteria Analysis for Technique Assessment, Case Study from Industrial Coating. Journal of Industrial Ecology Volume 9 number 3. Goedkoop, M., 1995. The Eco-indicator 95, final report, The Netherlands. Guinée, J.B., 2002. Handbook on Life Cycle Assessment. Operational Guide to the ISO Standards. In: Series: Eco-efficiency in Industry and Science, Vol 7, Guinée J.B., (Eds.), Kluwer Academic Publishing, The Netherlands. Haes, H.A.U., 2000. Weighting in Life-Cycle Assessment. Is There a Coherent Perspective?, Journal of Industrial Ecology volume 3 number 4. Hertwich, E.G., Hammitt, J.K., 2000. A Decision-Analytic Framework for Impact Assessment Part I: LCA and Decision Analysis. International Journal of LCA 5 . Hillary, R., 2004. Environmental management systems and the smaller enterprise. Journal of Cleaner Production 12 561-569. Huijbregts, M.A.J., Breedveld, L., Huppes, G., Koning, A., Oers L., Suh, S., 2003. Normalisation figures for environmental life-cycle assessment The Netherlands (1997/1998), Western Europe (1995) and the world (1990 and 1995). Journal of Cleaner Production 11 737-748. INETI (Instituto Nacional de Engenharia, Tecnologia e Inovação), 2000. Guia Técnico sectorial - Sector da Metalurgia e Metalomecânica (in Portuguese) (Technical Guide for the Metallurgical and Metal-related Industrial Sectors, Lisbon, Portugal. Available from http://www.ineti.pt/, last acceded on 14.06.2004. IPCC Guidellines, 1997. Revised IPCC Guidelines for National Greenhouse Gás Inventories, Houghton, J.T., Meira Filho, L.G., Lim, B., Treanton, K., Mamaty, I., Bonduki, Y., Griggs, D.J., Callender, B.A., (Eds.), available in http://www.ipcc-nggip.iges.or.jp/public/gl/invs1.htm, last acceded on 20.12.2006. Chapter 4: Modelling the environmental impact - 95 - Jasch, C., 2000. Environmental performance evaluation and indicators. Journal of Cleaner Production 8 79-88. Kamp, W. van de, 2005, Environmental impact of organically produced tomatoes in wintertime; imported from Spain versus home-grown in glasshouses. Master Thesis, Wageningen University, the Netherlands. Kim, S.K., Cho, H., Jo, H., Han, M., Lim, S., Hur, T., 2003. Material Life Cycle Assessment for Diecasting Process. Materials Science Forum vols. 462-432 3353-3358. Klimont, Z., Cofala, J., Bertok, I., Amann, M., Heyes, C., Gyarfas, F., 2002. Modelling Particulate Emissions in Europe A Framework to Estimate Reduction Potential and Control Costs. IIASA Interim Report IR-02-076. Klüber, 2005. Klubertec HP 1-810 - Powdered release agent for aluminium and magnesium pressure diecasting. Product information and safety data sheet. Available from http://www.klueber.com, last acceded on 27.05.2005. Kolk, A., Mauser, A., 2002. The evolution of environmental management: from stage models to performance evaluation. Business Strategy and the Environment 11 14-31. Kortman, J.G.M., Lindeijer, E.W., Sas, H., Sprengers, M., 1994. Towards a single indicator for emissions - An exercise in aggregating environmental effects, IVAM, Interfaculty Department of Environmental Sciences, University of Amsterdam, The Netherlands. Krajnc, D., Glavič, P., 2005. A model for integrated assessment of sustainable development. Resources, Conservation and Recycling 43 189-208. Kytzia, S., Nathani, C., 2004. Bridging the gap to economic analysis: economic tools for industrial ecology. Progress in Industrial Ecology Vol.1 numbers 1/2/3. Moors, E.H.M., Mulder, K.F., Vergragt, P.J., 2005. Towards cleaner production: barriers and strategies in the base metals producing industry. Journal of Cleaner Production 13 657-668. Neto B., Kroeze C., Pulles M., 2003. Environmental Performance of a Metals Industry facility. Industrial Production Chain Analysis. Paper presented at the second ISIE Conference ‘Industrial Ecology for a Sustainable Future’, June 29-July 2, Ann Arbor, USA. Olsthoorn, X., Tyteca, D., Wehrmeyer, W., Wagner, M., 2001. Environmental indicators for business: a review of the literature and standardisation methods. Journal of Cleaner Production 9 453-463. Pennington, D.W., Potting, J., Finnveden, G., Lindeijer, E., Jolliet, O., Rydberg, T., Rebitzer, G., 2004. Life cycle assessment Part 2: Current impact assessment practice. Environmental International 30 721-739. Pineda-Henson, R., Culaba, A.B., Mendoza, G.A., 2002. Evaluating Environmental Performance of Pulp and Paper Manufacturing Using Analytical Hierarchy process and Life-Cycle Assessment. Journal of Industrial Ecology Volume 6 Number 1. Pluimers, J., 2001. An environmental systems analysis of greenhouse horticulture in the Netherlands. PhD thesis. Wageningen University, The Netherlands. Rabah, M., 1999. Cost effectiveness of abatement options for emission control in Egyptian iron foundries. Waste management 19 283-292. Rebitzer, G., Ekvall, T., Frischknecht, R., Hunkeler, D., Norris, G., Rydberg, T., Schmidt, W.P., Suh, S., Weidema, B.P., Pennington, D.W., 2004. Life cycle assessment Part 1: Framework, goal and scope definition, inventory analysis and applications. Environmental International 30 701- 720. Chapter 4: Modelling the environmental impact - 96 - Sakamoto, Y., Tonooka, Y., 2000. Estimation of CO2 emission for each process in the Japanese steel industry: a process analysis. International Journal of Energy Research 24 625-632. Schwarz, B., 1997. Forecasting and Scenario’s. In: H.J. Miser and E.S. Quade, (Eds.), Handbook of Systems Analysis. Volume two Craft issues and procedural choices, John Wiley and Sons, Chichester, 327-367. Seppälä, J., Koskela, S., Melanen, M., Palperi, M., 2002. The Finish metals industry and the environment. Resources, Conservation and Recycling 35 61-76. Silvo, K., Melanen, M., Honkasalo, A., Ruonala, S., Lindström, M., 2002. Integrated pollution prevention and control – the Finnish approach. Resources, Conservation and Recycling 35 45-60. USEPA, 2002. EPA Air Pollution Control Cost Manual, EPA/452/B-02-001, sixth edition, available from http://www.epa.gov/ttn/catc/products.html, last acceded 20.12.2006. Van Langen, P.H.G., 2002. The Anatomy of Design: Foundations, Models and Applications, PhD thesis, Vrije University Amsterdam, The Netherlands. Winiwarter, W., Schimak, G., 2005. Environmental software systems for emission inventories. Environmental Modelling & Software 20 1469-1477. Wright, M., Allen, D., Clift, R., Sas, H., 1998. Measuring Corporate Environmental Performance. The ICI Environmental Burden System. Journal of Industrial Ecology Vol.1 Number 4. Wrisberg, N., Haes, H.A.U., Triebswetter, U., Eder, P., Clift, R., 2002. Analytical tools for Environmental Design and Management in a Systems Perspective, the combined use of analytical tools. In: Wrisberg, N., Haes, H.A.U., Triebswetter, U., Eder, P., Clift, R., (Eds.), Eco-efficiency in Industry and Science, Kluwer academic publishers. Zárate, IO., Ezcurra, A., Lacaux, JP., Dinh, P., 2000. Emission factor estimates of cereal waste burning in Spain. Atmospheric Environment 34 3183-3193. Zobel, T., Almroth, C., Bresky, J., Burman, J., 2002. Identification and assessment of environmental aspects in an EMS context: an approach to a new reproducible method based on LCA methodology. Journal of Cleaner Production 10 381-396. Zobel, T., Burman, J., 2004. Factors of importance in identification and assessment of environmental aspects in an EMS context: experiences in Swedish organizations. Journal of Cleaner Production 12 13-27. - 97 - Chapter 5: Strategies to Reduce the Environmental Impact of an Aluminium Pressure Die Casting Plant: a Scenario Analysis Belmira Neto, Carolien Kroeze, Leen Hordijk, Carlos Costa, Tinus Pulles This chapter has been submitted to Journal of Environmental Management Abstract This study explores a model (MIKADO) to analyse scenarios for the reduction of the environmental impact of an aluminium die casting plant. Our model calculates the potential to reduce emissions, and the costs associated with implementation of reduction options. We first present model results for a situation in which no reduction options are assumed to be implemented (so-called zero case, reflecting the current practice in the plant). Second, we perform a systematic analysis of reduction options. Finally, seven types of reduction strategies are analysed, assuming the simultaneous implementation of different reduction options. These strategies are analysed with respect to their potential to reduce emissions, environmental impact and costs associated with the implementation of options. These strategies were found to differ largely in their potential to reduce the environmental impact of the plant (10 – 87%), as well as in the costs associated with the implementation of options (-268 to +277 k€/year). We were able to define eleven strategies, reducing the overall environmental impact by more than 50%. Of these, two have net negative costs, indicating that the company may in fact earn money through their implementation. Chapter 5: Strategies to reduce the environmental impact - 104 - Table 5.1. The relative contribution of the emissions of an aluminium die casting plant to the overall environmental impact (M) by sub-sub-process and for the zero case. (Valuation method: All problems equally important) (units: % relative to the M for zero case). Process (p) subprocess (pi) sub-sub-process (pij) Environmental impact categories(z) Pollutants / Liquid effluent / Solid wastes (x) associated with the environmental impact categories Contribution to M (%) Human toxicity HF Cr NMVOC Cd + Ni + Pb + Cu + NMVOC 9% 6% 2% <1% Abiotic depletion Natural gas a) 17% Global warming CO2 7% Solid waste production Aluminium dross 4% Acidification NOx HF 2% 2% Terrestrial ecotoxicity Cr Cd + Ni + Pb + Cu + HF+NMVOC 2% <1% Melting (i=1) Melting Photochemical ozone formation CO + NMVOC + NOx <1% Human toxicity Cr Zn + Pb + Cu 21% <1 % Terrestrial ecotoxicity Cr Zn + Pb + Cu 4% <1 % Holding Furnaces Solid waste production Ceramic lining <1% Human toxicity NMVOC 18% Terrestrial ecotoxicity NMVOC <1% Solid waste production Oils & Grease Sludge 1% 1% Photochemical ozone formation NMVOC <1% Casting (i=2) Pressure Die Casting Aquatic toxicity Liquid effluent <1% Shot Blasting Solid waste production Steel Shot <1% Solid waste production Ceramic Abrasives Sludge <1% <1% Tumbling Aquatic toxicity Liquid effluent <1% Solid waste production Sludge Oils and Grease <1% <1% Finishing (i=3) Cleaning and Degreasing Aquatic toxicity Liquid effluent <1% Human toxicity NMVOC NOx + SO2 + Particulates 2% <1% Global warming CO2 <1% Acidification NOx SO2 1% <1% Terrestrial ecotoxicity NMVOC <1% Internal Transport (i=4) Fork-lift Truck on Diesel and LPG Photochemical ozone formation CO + NMVOC + NOx + SO2 <1% Human toxicity NMVOC + NOx + Particulates <1% Global warming CO2 <1% Acidification NOx <1% Terrestrial ecotoxicity NMVOC <1% Die Casting company Auxiliary Burners (i=5) Oxyacetylene and Butane burners Photochemical ozone formation CO + NMVOC + NOx <1% a) The natural gas consumption contributes to the impact category - depletion of natural resources. Chapter 5: Strategies to reduce the environmental impact - 105 - 5.4. Systematic analysis of individual pollution reduction options The environmental impact of industrial activities can be reduced by end-of-pipe technologies or by changes in the operation process (Beaumont and Tinch, 2004; Moors, 2006). We identified 18 options to reduce the potential environmental impact for the die casting plant (see Chapter 3, section 3.3.3.). These options are implemented at the sub-sub-processes level. We organise the options in different types (Table 5.2). Within the types we consider the options to be mutually exclusive (e.g. a fabric filter cannot be applied with another option that, simultaneously, is part of the same subsub-process and the same type “filters and scrubbers”). For each option the potential to reduce emissions is estimated, as well as the costs involved (see Chapter 3, section 3.3.3.). The options considered were described in the literature or were proposed by the managers of the plant that serves as a case in this study. Thus, the options may be considered up-to-date techniques to be used by any company in the aluminium pressure die casting sector. Most options affect more than one pollutant and most compounds are affected by more than one option (Table 5.2). MIKADO includes options for all the sub-processes within the plant and includes end- of-pipe techniques (such as fabric filters), as well as more structural reduction options (Table 5.2). An example of a more structural option is replacing the desoxidation agent by a less polluting agent (granular desoxidation agent). Other examples include using a different technique for a) degassing of the molten alloy (Impeller station), b) the combustion process (air enrichment with oxygen or oxyfuel firing), c) mould release (by use of a lower concentrated agent maintaining the technique or changing the technique by using a powder agent). Yet another example is d) an alternative metal loading (compact metal load in melting furnaces), which may improve the thermal efficiency, by introducing in the furnaces a more compacted material in substitution to the remains of casting. The model also includes options aiming at increasing the metal yield (where metal yield is the ratio of production to molten alloy). These options either reduce the mass of runners in the die casting moulds or reduce the scrap rate produced in the subprocesses Casting and Finishing. The metal yield increases due to the reduction of the mass of aluminium alloy that feeds the melting furnaces and hence the energy needed. Thus, these options not only reduce the amount of aluminium that is recycled internally, but also decrease the use of subsidiary materials that are directly dependent on the amount of molten alloy. As a result, many emissions will be reduced. Chapter 5: Strategies to reduce the environmental impact - 106 - Table 5.2. Overview of pollution reduction options for a die casting plant (from Neto et al., submitted (Chapter 4)). See Chapter 3 (section 3.3.3.) for a description of the reduction options. subprocess (pi) subsub- process (pij) Types of Options Reduction Options (τ) Abbreviation Compounds reduced Fabric Filter. Reverse-air type a) Melting_FF_RA Heavy metals such as Cd, Ni, Pb, Cr Fabric Filter. Pulse-Jet type b) Melting_FF_PJ Heavy metals such as Cd, Ni, Pb, Cr Fabric Filter. Mechanical Shaker type c) Melting_FF_MS Heavy metals such as Cd, Ni, Pb, Cr Wet Scrubber. Impingement-Plate type d) Melting_WS_IP Heavy metals (Cd, Ni, Pb, Cr), Cu and HF Filters and scrubbers Wet scrubber. Spray-chamber type e) Melting_WS_SC Heavy metals (Cd, Ni, Pb, Cr), Cu and HF and NMVOC. Alternative desoxidation agent Granular desoxidation agent f) Melting_GA HF, Aluminium dross. Alternative degassing technique Impeller station using N2f), k) Melting_IS HF, Aluminium dross. Alternative metal loading in furnaces Compact metal loading in furnaces g) Melting_CM CO2, CO, NOx, and NMVOC (Natural gas combustion related emissions). Air enrichment with oxygen (30%O2) g) Melting_AE CO2, CO, NOx, and NMVOC (Natural gas combustion related emissions). Melting (i=1) Melting Combustion process modification Oxyfuel firing (100%O2)g) Melting_OF CO2, CO, NOx, and NMVOC (Natural gas combustion related emissions). Wet Scrubber. Packed-Bed type h) Casting_WS_PB Heavy metals (Pb, Cr), Cu, Zn, NMVOC Scrubbers Wet scrubber. Spray-chamber type e) Casting_WS_SC Heavy metals (Pb, Cr), Cu, Zn, NMVOC. New mould release agent g) Casting_nMA NMVOC, liquid effluent, oils, grease and sludge. Alternative to mould release agent application Powder agent i), k) Casting_PA NMVOC, liquid effluent, oils, grease and sludge. New die casting moulds Reduce runners mass j), *) Casting_rRR Heavy metals (Cd, Ni, Pb, Cr), Cu, Zn, HF, CO2, CO, NOx, and NMVOC, aluminium dross, oils, grease and sludge. Casting (i=2) Pressure die casting Reduce scrap rate Reduce scrap rate k), *) Casting_rSR Heavy metals (Cd, Ni, Pb, Cr), Cu, Zn, HF, CO2, CO, NOx, and NMVOC, aluminium dross, oils, grease and sludge. Finishing (i=3) Trimming Reduce scrap rate Reduce scrap rate k), *) Finishing_rSR Heavy metals (Cd, Ni, Pb, Cr), Cu, Zn, HF, CO2, CO, NOx, and NMVOC, aluminium dross, oils, grease and sludge. Internal Transport (i=4) Forklift Truck on Diesel (I and II) and LPG Electrical equipment Use electric forklift trucks g) IT_eFL CO2, CO, NOx, and NMVOC, SO2 and Particulates. a) to e) and h) USEPA Air pollution control technology fact sheet. EPA-CICA Fact Sheet. USEPA (2002). a) EPA-452/F- 03-026; b) EPA-452/F-03-025; c) EPA-452/F-03-02;. d) EPA-452/F-03-012; e) EPA-452/F-03-016; f) Brown (1999); g) Pedro (2005). Personal communication; h) EPA-452/F-03-015; i) Klüber (2005); j) INETI (2000); k) EIPPCB (2005). *) These reduction options may change the value of the metal yield. For some reduction strategies, where these three reduction options were used, the runners’ mass and scrap rates were modified leading to values of metal yield equal to 57% (EIPPCB, 2005). Chapter 5: Strategies to reduce the environmental impact - 107 - 5.4.1. Effectiveness of reduction options in reducing the environmental impact The potential to reduce the overall environmental impact differs largely for the total 18 reduction options analysed (Figure 5.3). The two most effective options are found to be two wet scrubbers associated with Casting (Casting_WS_PB and Casting_WS_SC), each of which may reduce the overall impact by about 40% (Figure 5.3, Table 5.3). This reduction is mainly achieved by a reduction in emissions of toxic compounds. The two least effective reduction options are calculated to reduce the overall impact by less than 1%. These include more structural reduction options, such as the use of an impeller station for degasification (Melting_IS), and using a small percentage of oxygen for the fuel combustion (Melting_AE). The other 14 reduction options have more intermediate results, reducing M from 4 to 20%. A reduction in M of about 20% is calculated for wet scrubbers in Melting (Melting_WS_SC) as a result of reducing emissions of hydrogen fluoride, chromium and NMVOC. The option to use powder release agents (Casting_PA) in Casting, may reduce M by 20% as a result of a decrease in NMVOC emissions. A smaller reduction is observed for the options that compact the metal load (Melting_CM, 17% reduction in M), the new die casting moulds (Casting_rRR, 15% reduction in M) and the new mould release agent (Casting_nMA, 11% reduction in M). The other options reduce M by less than 10%. It can be concluded that the choice of the valuation method has a small effect on the calculated effect of options on the overall environmental impact M (Figure 5.3). Therefore, in the following we only analyse the results for the valuation factors that consider all environmental problems equally important. Chapter 5: Strategies to reduce the environmental impact - 108 - 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100% zero case Melting_IS Melting_AE Melting_OF IT_eFL Finishing_rSR Casting_rSR Melting_GA Melting_FF_RA Melting_FF_PJ Melting_FF_MS Casting_nMA Casting_rRR Melting_CM Melting_WS_IP Casting_PA Melting_WS_SC Casting_WS_PB Casting_WS_SC Individual reduction options M (%) all problems equally important Panel method I Panel method II Distance to target Figure 5.3. Overall environmental impact (M) of the plant for the zero case and for alternative cases in which one reduction option is assumed to be implemented. The overall environmental impact (M) is calculated for four multi-criteria analyses (MCA) using different valuation methods: All problems equally important; Panel Method I (Kamp, 2005); Panel Method II (Kortman et al., 1994); and Distance to target (Goedkoop, 1995). (units: % relative to M for the zero case). See Chapter 3 (section 3.3.3.) for a description of the reduction options. Chapter 5: Strategies to reduce the environmental impact - 109 - Table 5.3. Relative contribution of sub-processes (pi) and sub-sub-processes (pij) to the overall environmental impact (M) for the zero case and for alternative cases in which a single reduction option is assumed to be implemented. (the valuation method used is the one that considers all problems equally important) (units: % relative to M for the zero case). See Table 5.2 for an explanation of the abbreviations. Reduction Options (τ) subprocess (pi) sub-sub- process (pij) Zero case Melting_IS Melting_AE Melting_OF IT_eFL Finishing_rSR Casting_rSR Melting_GA Melting_FF_RA Melting_FF_PJ Melting_FF_MS Casting_nMA Casting_rRR Melting_CM Melting_WS_IP Casting_PA Melting_WS_SC Casting_WS_PB Casting_WS_SC (1) Melting Melting 51% 51% 50% 47% 51% 48% 48% 44% 44% 44% 44% 51% 43% 34% 33% 51% 31% 51% 51% Holding Furnaces 24% 24% 24% 24% 24% 23% 23% 24% 24% 24% 24% 24% 20% 24% 24% 24% 24% 1% 0% (2) Casting Pressure Die Casting 20% 20% 20% 20% 20% 19% 19% 20% 20% 20% 20% 9% 17% 20% 20% 0% 20% 2% 3% Shot Blasting <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% Surface Treatment Tumbling <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% (3) Finishing Cleaning and Degreasing <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% (4) Internal Transports Fork-lift Truck on Diesel (I and II) and LPG 4% 4% 4% 4% 0% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% (5) Auxiliary Burners Oxyacetylene and Butane burners <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% <1% Overall Contribution (M) 100% 100% 100% 96% 96% 95% 95% 93% 93% 93% 93% 89% 85% 83% 82% 80% 80% 59% 59% Chapter 5: Strategies to reduce the environmental impact - 110 - Next, the effects of single reduction options on the calculated potential impact per environmental impact category (Mz) are investigated (Figure 5.4). The model includes 15 options to reduce emissions of compounds with a high human toxicity. Their calculated potential to reduce the environmental impact (Mz for HTP) ranges from 2 to 65%. The most effective options are the two different types of wet scrubbers located in Casting (Casting_WS_PB and Casting_WS_SC). These two options are calculated to reduce Mz by about 65% each. Six reduction options were analysed for decreasing natural gas use. The most effective reduction option (Melting_CM) may reduce the calculated environmental impact (Mz for ADP) by about 60%. The other options have a smaller effect on natural gas use, and lead to a 2-15% reduction in the environmental impact (Mz). The least effective (2% reduction) is the option to reduce the oxygen concentration in combustion (Melting_AE). A reduction in the greenhouse gas emissions (contributing to global warming) was calculated for seven options. A relatively large reduction (more than 50%) in Mz (for GWP), is calculated for the option to compact the metal load (Melting_CM). A smaller reduction (15%) is calculated for the option to use new moulds in the die casting process (Casting_rRR). Reductions smaller than 10% are calculated for options associated with Internal Transport (IT_eFL), reducing scrap rate (Casting_rSR and Finishing_rSR) and the use of oxygen in combustion (Melting_AE and Melting_OF). The environmental impact from solid waste production (Mz for SW) may be reduced by up to 30% by the options analysed. On the other hand, some add-on techniques, meant to reduce other emissions may, as a side effect, increase the production of waste. This is true for options collecting dust that lead to an increase of the amount of waste (in particulate form or sludge) that needs to be disposed of. The largest reduction (about 30%) in SW is calculated for the option to use a powder mould release agent (Casting_PA). The use of die casting moulds (Casting_rRR) and of new mould release agents (Casting_nMA) may reduce Mz by about 15%. The other options reduce Mz by less than 5%. We include eleven options to reduce acidification in our analyses. These are meant to reduce the impact for acidifying compounds (Mz for AP). The most effective reduction option (Melting_OF), may reduce Mz by over 45%. The other options reduce the calculated Mz by 1% such as (Melting_IS and Melting_AE) to 35% (Melting_WS_IP and Melting_WS_SC). In total, ten options were analysed to reduce emissions of compounds contributing to terrestrial ecotoxicity problems (Mz for ECP). The two most effective are add-on techniques for the sub-process Casting and include two different types of wet scrubbers (Casting_WS_PB and Casting_WS_SC). These are calculated to reduce Mz by about 75% each. The other options have intermediate results ranging from 5% (for all the options reducing scrap rate) to 20% for all the add-on techniques in Melting. Twelve options to reduce emissions of tropospheric ozone precursors were analysed. They may reduce the environmental impact (Mz for POCP) by 1% (Melting_AE) to 40% (Casting_PA, Casting_WS_PB and Casting_WS_SC). Eight intermediate options are found to reduce emissions of ozone precursors from 5% to 30%. Chapter 5: Strategies to reduce the environmental impact - 111 - Five options to reduce emissions of compounds contributing to aquatic toxicity (Mz for ATP) were analysed. The largest reduction was found for the option to use powder agent (90% reduction in ATP). The option to use a new mould release agent (Casting_nMA) reduces this Mz by 40% while the other options from 5 to 15%. Figure 5.4. As figure 5.3, but for each environmental impact category (z). The results are only presented for options affecting the Mz in question, and for the valuation factor that considers all environmental problems equally important (units: % relative to Mz for the zero case). Human toxicity 0% 20% 40% 60% 80% 100% zero case Melting_CM IT_eFL Finishing_rSR Casting_rSR Melting_GA Melting_FF_RA Melting_FF_PJ Melting_FF_MS Casting_rRR Casting_nMA Melting_WS_IP Melting_WS_SC Casting_PA Casting_WS_PB Casting_WS_SC Individual reduction options Mz (%) Abiotic depletion 0% 20% 40% 60% 80% 100% zero case Melting_AE Melting_OF Finishing_rSR Casting_rSR Casting_rRR Melting_CM Individual reduction options Mz (%) Global warming 0% 20% 40% 60% 80% 100% zero case Melting_AE Melting_OF Finishing_rSR Casting_rSR IT_eFL Casting_rRR Melting_CM Individual reduction options Mz (%) Acidification 0% 20% 40% 60% 80% 100% zero case Melting_AE Melting_IS Finishing_rSR Casting_rSR Casting_rRR IT_eFL Melting_GA Melting_CM Melting_WS_IP Melting_WS_SC Melting_OF Individual reduction options Mz (%) Terrestrial ecotoxicity 0% 20% 40% 60% 80% 100% zero case Finishing_rSR Casting_rSR Casting_rRR Melting_FF_RA Melting_FF_PJ Melting_FF_MS Melting_WS_IP Melting_WS_SC Casting_WS_PB Casting_WS_SC Individual reduction options Mz (%) Photochemical ozone formation 0% 20% 40% 60% 80% 100% zero case Melting_AE Finishing_rSR Casting_rSR Melting_WS_SC Casting_rRR IT_eFL Casting_nMA Melting_CM Melting_OF Casting_WS_SC Casting_WS_PB Casting_PA Individual reduction options Mz (%) Aquatic toxicity 0% 20% 40% 60% 80% 100% zero case Finishing_rSR Casting_rSR Casting_rRR Casting_nMA Casting_PA Individual reduction options Mz (%) Solid waste produced 0% 20% 40% 60% 80% 100% 120% zero case Melting_WS_IP Melting_WS_SC Melting_FF_RA Melting_FF_PJ Melting_FF_MS Melting_GA Finishing_rSR Casting_rSR Casting_nMA Casting_rRR Casting_PA Individual reduction options Mz (%) Chapter 5: Strategies to reduce the environmental impact - 112 - 5.4.2. Costs of reduction options The implementation of pollution reduction options results in costs for the company. The total costs for each individual reduction option include fixed cost (equipment investment and the fixed operational cost) and variable costs (dependent on the equipment or materials use). In another chapter (Chapter 4) (Neto et al., submitted), we provide details on the cost parameters used in the analysis. In line with Geldermann and Rentz (2004) the costs are considered in terms of net additional cost for each reduction option. The calculated net additional costs for an option reflect the extra costs or savings (in some cases the company saves money by implementing an option) associated with the implementation of a reduction option, relative to the zero case. Thus, these costs may be negative if the implementation of a reduction option brings revenue for the company. The calculated net additional costs (Cna) of reduction options range from -128 k€/year, for the option to compact metal load (Melting_CM), to +224 k€/year for the option when to implement oxyfuel firing in Melting (Melting_OF) (Figure 5.5). Six reduction options are calculated to have negative costs, indicating that by implementing these options, the company, in fact, may earn money. These include the options to use a granular desoxidation agent (Melting_GA, -0.24 k€/year), some of the options to increase the metal yield (Casting_rSR, -30 k€/year, and Finishing_rSR, -36 k€/year), the use of electric forklift trucks in internal transports (IT_eFL, -39 k€/year) and options using a new mould release agent (Casting_nMA, -58 k€/year). The largest savings are calculated for the option to compact the metal load (Melting_CM, -128 k€/year). The four most costly reduction options include the use of oxyfuel firing (Melting_OF, 224 k€/year), a wet scrubber in Casting (Casting_WS_PB, 195 k€/year), the use of powder mould release agent (Casting_PA, 146 k€/year) and the use of new moulds on Casting (Casting_rRR, 119 k€/year). Among the non-paying options are eight options with costs below 100 k€/year. Chapter 5: Strategies to reduce the environmental impact - 113 - -150 -100 -50 0 50 100 150 200 250 Melting_OF Casting_WS_PB Casting_PA Casting_rRR Melting_FF_RA Melting_FF_MS Melting_AE Melting_IS Melting_FF_PJ Casting_WS_SC Melting_WS_IP Melting_WS_SC zero case Melting_GA Casting_rSR Finishing_rSR IT_eFL Casting_nMA Melting_CM Individual Reduction Options Net additional Costs (Cna) (k€/year) Figure 5.5. Net additional Costs (Cna) for the aluminium die casting plant for cases in which it is assumed that one of the reduction options is implemented. See Table 5.2 for an explanation of reduction options. We also calculated the cost-effectiveness (CE) of reduction options. We define costeffectiveness as the net additional costs (Cna) per avoided overall environmental impact (M) (Equation 1, following Pluimers (2001)). The cost-effectiveness for the reduction options is also calculated for each impact category (Mz) (Equation 2). () τ τ + − = zerocasezerocase na MMM C CE , (Equation 1) () τ τ + − = zerocasezzerocasez na MMM C CE z ,, , (Equation 2) Chapter 5: Strategies to reduce the environmental impact - 120 - 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% zero case RS I-A RS I-B RS I-C RS I-D RS I-E RS I-F RS I-G RS II-A RS II-B RS II-C RS III-A RS III-B RS III-C RS IV-A RS IV-B RS IV-C RS V-A RS VI-A RS VI-B RS VI-C RS VI-D RS VII-A RS VII-B RS VII-C RS VII-D Reduction Strategies M (%) Aquatic toxicity Photochemical ozone formation Terrestrial ecotoxicity Acidification Solid waste production Global warming Abiotic depletion Human toxicity Figure 5.6. Overall environmental impact (M) for the zero case and for the alternative reduction strategies. The valuation method used on the overall environmental impact assessment considers that all environmental problems to be equally important. See Table 5.5 for a description of the reduction strategies. (unit: % relative to M for the zero case). Chapter 5: Strategies to reduce the environmental impact - 121 - Reduction Strategies IV: combining the most cost effective options Combining the most cost effective options is probably the most interesting from a company perspective. In reduction strategy IV-A, we combine the six options reducing M and having simultaneously, a net negative costs (Melting_CM, Casting_nMA, Casting_rSR, Finishing_rSR, IT_eFL and Melting_GA). In strategies IV-B and IV-C we also consider options having relatively low costs, but that are relatively effective in reducing the environmental impact (Melting_WS_SC, Casting_WS_SC and Casting_PA) (Table 5.4). The overall impact is calculated to be reduced by 45% (IV-A) to 86% (IV-C) (Figure 5.6, Table 5.5). The cost associated with these strategies are calculated to be negative ranging from -268 k€/y (IV-A) to -51 k€/y (IV-C). These results indicate that relatively large reductions (up to 85%) in environmental impact are possible while gaining money. Reduction Strategies V: combining add-on techniques Some add-on techniques are known to have a relatively large potential to reduce emissions of specific compounds. We selected relatively effective options that also have relatively low cost (Figure 5.3, Figure 5.5). These options include two wet scrubbers (Melting_WS_SC, Casting_WS_SC). Reduction strategy V-A is calculated to reduce the overall environmental impact (M) by 61% at an associated cost of 29 k€/y (Figure 5.6, Table 5.5). Reduction Strategies VI: combining more structural reduction options Four different combinations of more structural reduction options are analysed (VI-A to VI-D). The options increasing the metal yield are analysed separately (see Reduction Strategies type VII). The reduction strategies are calculated to reduce the overall environmental impact by 39% (VI-A) to 49% (VI-D) relative to the zero case (Figure 5.6, Table 5.5). The costs associated with these strategies range considerably from -224 k€/y (VI-A) to +91 k€/y (VI-D). Clearly, strategy VI-A is the most interesting, given the negative costs. Reduction Strategies VII: increasing the metal yield Increasing the metal yield is generally considered an important strategy to reduce pollution from the metals sector industry and in particular to the case plant studied. Here we selected options that aim to decrease the alloy mass inputs returning to melting furnaces and as such reduce all materials and energy needed in the die casting process. Options that increase the metal yield (MY) are those that reduce the scrap rate (Casting_rSR and Finishing_rSR) and those that reduce the runner’s mass by using new die casting moulds (Casting_rRR) (Table 5.2). Four different reduction strategies (VII- A to VII-D) are defined to increase the metal yield (for comparison: the MY in the zero case is 47%). These strategies were calculated to reduce the overall environmental impact from 10% (VII-A) to 16% (VII-D) (Figure 5.6, Table 5.5). The costs associated with these strategies range from 149 k€/y (VII-A) to 101 k€/y (VII-D). It is interesting to note that increasing the metal yield, although generally considered an important Chapter 5: Strategies to reduce the environmental impact - 122 - strategy, is not very effective in reducing the environmental impact, and is relatively costly. Compared to the other strategies that we analysed, increasing the metal yield is perhaps not the first choice. Comparing the different strategies From the above, it may be clear that the different strategies that we analysed differ largely in their potential to reduce the environmental impact of the company (10 – 87%) as well as in the costs associated with the implementation of options (-268 to +277 k€/year). We were able to define 11 strategies reducing the overall environmental impact by more than 50%. Of these, two have net negative costs, indicating that the company may in fact earn money by implementing them. The largest effect on the environment (87% reduction in M) is calculated for strategy I-G, which in fact focuses on reducing the human toxicity. This is mainly because compounds with a human toxicity effect have the largest share on the overall impact (M), making strategy I-G very effective to reduce the overall impact. However, this strategy, is rather costly (118 k€/year). A similar reduction (86%) could be obtained while gaining 51 k€/year for reduction strategy IV-C, which is a combination of relatively cost effective options. If we combine only the highly cost effective options, the savings are even larger (-268 k€/year) while reducing the environmental impact by almost 45% (IV-A). 5.6. Discussion and Conclusions This study explores a model (MIKADO) that assesses options to reduce the environmental impact of a plant supplying car manufacturers with aluminium die casting products. MIKADO includes a number of options for emission reduction, and can be used to calculate their technical potentials to reduce the environmental impact as well as the associated costs. We analysed individual reduction options, as well as reduction strategies, in which options are combined. MIKADO may support environmental decision making, by assisting the management of the company in answering “what …if” type questions (e.g. ‘What would the effect on the environment and on the costs be if we implement the following options?). First, we analysed the so-called zero case, assuming that none of the reduction options is implemented. The overall environmental impact of the plant is mostly associated with human toxicity (caused by metal emissions and emissions of ozone precursors), and abiotic depletion of natural gas. These two environmental problems account for about 75% of the overall environmental impact. This may be not too surprising since the main compounds released by the industry are metals, including heavy metals and some volatile organic compounds, both contributing to human toxicity problems. More than 90% of the overall environmental impact of the company comes from the subprocesses Melting and Casting. More specifically, we conclude that there are four relatively large sources of environmental pollution in the aluminium die casting plant: emissions of chromium and NMVOCs from Casting, the use of natural gas in Melting and emissions of hydrogen fluoride from Melting. These four are responsible for about two-thirds of the overall environmental impact. Chapter 5: Strategies to reduce the environmental impact - 123 - Second, the 18 individual reduction options were analysed systematically with respect to their potential to reduce the environmental impact of the company, and the associated costs. The individual options may reduce the environmental impact by up to 40%. The largest reductions in environmental impact were calculated for two different types of wet scrubbers in Casting. These scrubbers are particularly effective in reducing emissions having a large effect on human toxicity. The cost associated with the implementation differs largely for the 18 options. Six options have net negative costs, implying that the company may in fact earn money by implementing them. These include the option to compact the metal load; the use of a new mould release agent, the use of electric fork-lift trucks, the reduction of the scrap rates and the use of a granular agent. These options are also the most cost effective options. Of these, compaction of the metal load may be the most interesting, given its relatively large effect on the environment (17% reduction of M). We defined seven different types of reduction strategies in which reduction options are combined. The strategies defined include combinations of reduction options that aim I) to reduce the largest environmental problem (human toxicity); II) to reduce the use of natural gas, III) to reduce a specific pollutant emission (chromium); IV) to combine the most cost effective reduction options; V) to combine only add-on techniques; VI) to combine more structural reduction options or VII) to increase the metal yield. These strategies differ largely in their environmental impact (10 – 87% reduction) and net additional costs (-268 to +277 k€/year) (Table 5.5). The most effective strategy is a combination of options to reduce human toxicity problems (I-G). This strategy reduces the overall environmental impact by 87%, however at relatively high costs (118 k€/y). A similar reduction in M (86%) can be obtained by combining relatively cost effective options (IV-C), at net negative costs (-51 k€/y). The best paying strategy (IV-A) is reducing the environmental impact by 45% while gaining 268 k€/year. References Alcamo, J., Shaw, R., Hordijk, L., 1990. The RAINS model of acidification. Sciences and Strategies in Europe. Kluwer Academic Publisher, Dordrecht. Beaumont, N.J., Tinch, R., 2004. Abatement cost curves: a viable management tool for enabling the achievement of win-win waste reduction strategies?. Journal of Environmental Management 71 207-15. Blonk, H., 1997. Three reference levels for normalisation in LCA. Ministry of Housing, Spatial Planning and the Environment (VROM), The Netherlands. Brown, J.R., 1999. Foseco Non-Ferrous Foundryman´s Handbook, Butterworth-Heinemann, Oxford. Castillo, E.F., Mora, M., 2000. Mathematical modelling as a tool for environmental evaluation of industrial sectors in Columbia. Waste Management 20 617-23. Choi, A.C.K., Kaebernick, H., Lai, W.H., 1997. Manufacturing processes modelling for environmental impact assessment. Journal of Materials Processing Technology 70 231-38. EIPPCB (European Integrated Pollution Prevention and Control Bureau), 2005. Draft on Reference document on Best Available Techniques in the Smitheries and Foundries Industry, European Commission. Available from http://eippcb.jrc.es/, last acceded on 10.02.2005. Chapter 5: Strategies to reduce the environmental impact - 124 - Finnveden, G., Moberg, A., 2005. Environmental systems analysis tools: an overview. Journal of Cleaner Production 13 1165-73. Frey, H.C., Small, M.J., 2003. Integrated Environmental Assessment, Part I. Estimating emissions. Journal of Industrial Ecology vol.7 number 1. Gäbel, K., Forsberg, P., Tillman, A., 2004. The design and building of a life cycle-based process model for simulating environmental performance, product performance and cost in cement manufacturing. Journal of Cleaner Production 12 77-93. Gäbel, K., Tilmann, A., 2005. Simulating operational alternatives for future cement production. Journal of Cleaner Production 13 1246-57. Geldermann, J., Rentz, O., 2004. The reference installation approach for the techno-economic assessment of emission abatement options and the determination of BAT according to the IPPC- directive. Journal of Cleaner Production 12 389-02. Goedkoop, M., 1995. The Eco-indicator 95, final report. The Netherlands. Available from http://www.pre.nl/eco-indicator95/ei95-reports.htm, last acceded 10.09.2005. Guinée, J.B., 2002. Handbook on Life Cycle Assessment. Operational Guide to the ISO Standards. In: Series: Eco-efficiency in Industry and Science Vol 7, Kluwer Academic Publishing, The Netherlands. Huijbregts, M.A.J., Breedveld, L., Huppes, G., Koning, A., Oers, L., Suh, S., 2003. Normalisation figures for environmental life-cycle assessment The Netherlands (1997/1998), Western Europe (1995) and the world (1990 and 1995). Journal of Cleaner Production 11 737-48. INETI (Instituto Nacional de Engenharia, Tecnologia e Inovação), 2000. Guia Técnico sectorial - Sector da Metalurgia e Metalomecânica (in Portuguese) (Technical Guide for the Metallurgical and Metal-related Industrial Sectors, Lisbon, Portugal. Available from http://www.ineti.pt/, last acceded on 14.06.2004. IPCC (Intergovernmental Panel on Climate Change) Guidellines, 1997. Revised IPCC Guidelines for National Greenhouse Gás Inventories. In: Houghton JT, Meira Filho LG. Lim B, Treanton K, Mamaty I, Bonduki Y, Griggs DJ, Callender BA, editors. Available from http://www.ipccnggip.iges.or.jp/public/gl/invs1.htm, last acceded on 09.01.2006. Kamp, W., 2005. Environmental impact of organically produced tomatoes in wintertime; imported from Spain versus home-grown in glasshouses. Master Thesis, Wageningen University, The Netherlands. Kim, S.K., Cho, H., Jo, H., Han, M., Lim, S., Hur, T., 2003. Material Life Cycle Assessment for Diecasting Process. Materials Science Forum vols. 462-432 3353-58. Klüber, 2005. Klubertec HP 1-810 - Powdered release agent for aluminium and magnesium pressure diecasting. Product information and safety data sheet. Available from http://www.klueber.com, last acceded on 27.05.2005. Kortman, J.G.M., Lindeijer, E.W., Sas, H., Sprengers, M., 1994. Towards a single indicator for emissions - An exercise in aggregating environmental effects. IVAM: Interfaculty Department of Environmental Sciences, University of Amsterdam, The Netherlands. Moors, E.H.M., 2006. Technology strategies for sustainable metals production systems: a case study of primary aluminium production in The Netherlands and Norway. Journal of Cleaner Production 14 1121-38. Neto, B., Kroeze, C., Hordijk, L., Costa, C., submitted. Modelling the Environmental Impact of an Aluminium Pressure Die Casting Company and Options for Control. Submitted to Environmental Modelling & Software. Chapter 5: Strategies to reduce the environmental impact - 125 - Pennington, D.W., Potting, J., Finnveden, G., Lindeijer, E., Jolliet, O., Rydberg, T., Rebitzer, G., 2004. Life cycle assessment Part 2: Current impact assessment practice. Environmental International 30 721-39. Pluimers, J., 2001. An environmental systems analysis of greenhouse horticulture in the Netherlands. PhD thesis. Wageningen University, The Netherlands. Romero-Hernandez, O., 2005. Applying Life Cycle Tools and Process Engineering to Determine the Most Adequate Treatment Process Conditions. A tool in Environmental Policy. International Journal for Life Cycle Assessment 10 (5) 355-63. Tan, R.B.H., Khoo, H.H., 2005. Zinc Casting and Recycling. International Journal of LCA 10 (3) 211-18. USEPA, 2002. EPA Air Pollution Control Cost Manual, EPA/452/B-02-001, sixth edition, homepage, Available from http://www.epa.gov/ttn/catc/products.html, last acceded on 07.11.2006. Winiwarter, W., Schimak, G., 2005. Environmental software systems for emission inventories. Environmental Modelling & Software 20 1469-77. Zárate, I.O., Ezcurra, A., Lacaux, J.P., Dinh, P., 2000. Emission factor estimates of cereal waste burning in Spain. Atmospheric Environment 34 3183-93. - 126 - - 127 - Chapter 6: Conclusions and Discussion 6.1. Introduction The overall objective of this thesis was to develop a decision support tool (DST) to analyse options to reduce the environmental impact of an industrial company. A model was developed that allows for the assessment of the potential environmental impact resulting from emissions of environmental pollutants, as well as the effectiveness of reduction options and the associated costs. In this chapter the conclusions are drawn. In section 6.2. conclusions are drawn from the answers to the four research questions as well as with respect to the overall objective of this thesis. We also address the stepwise systems analysis procedure taken, and the applicability of the methodological approach to other industries. Section 6.3 includes a discussion of the results of the study. This discussion focuses on the environmental systems analysis approach taken (6.3.1), the uncertainties involved (6.3.2) and the implications of our DST for industry in general (6.3.3.). The chapter ends with recommendations for future studies (6.4.). 6.2. Conclusions In Chapter 1 we have formulated four research questions in order to meet the objectives of the thesis. In the following, we present the research questions and our conclusions. Research question 1: “What existing environmental systems analysis methods and tools can in principle be combined in a decision support tool (DST) and used to analyse the environmental performance of a plant from a company perspective?”. o It can be concluded that careful selection of environmental systems analysis tools is important. In this thesis, selection of tools was based on the desired characteristics of the DST to be developed. o We aim for an analysis that: 1) takes a company perspective; 2) includes environmental and economic aspects of decision making, 3) includes a complete coverage of the potential environmental impacts and 4) allows for an assessment of the consequences of a set of alternative strategies on pollution reduction. o We conclude that for such analysis a DST is needed that i) considers a gate- to-gate approach; ii) considers the processes within the company that are relevant for the assessment of the environmental impact; iii) uses company’s specific data easily available from the process owner; iv) considers up-to-date and company specific pollution reduction options; v) provides information on the costeffectiveness of the reduction options; vi) can be used to express the company’s environmental performance in one overall indicator; and lastly vii) may be used to explore possible user-defined pollution reduction strategies. Chapter 6: Conclusions and Discussion - 128 - o Based on these characteristics, we conclude that the following seven ESA tools form a good basis for our DST: Life Cycle Assessment (LCA), Substance Flow Analysis (SFA), Multi-Criteria Analysis (MCA), Technology Assessment (TA), Sensitivity Analysis (SA), Scenario Analysis (ScenA) and Cost-Effectiveness Analysis (CEA). Research question 2: “Which technical pollution reduction options are available for reducing the environmental impact of an aluminium pressure die casting plant? What are their technical potentials to reduce this impact, and the associated costs for the plant?” o We conclude that in an analysis of the environmental performance of an aluminium die casting company, it is important to consider five main subprocesses: 1) Melting, 2) Casting, 3) Finishing, 4) Internal Transports and 5) Auxiliary Burners. Each sub-process includes a series of process operations here referred to as sub-sub-processes. o Eighteen technical reduction options have been identified that could be applied by the aluminium pressure die casting plant. The options aim at reducing the different pollutants emitted by a specific sub-sub-process level within the company. The emissions of pollutants include air emissions, liquid effluents or solid wastes. The options considered may be typical end-of-pipe solutions (including fabric filters and scrubbers) or be more structural and use of alternative agents or techniques; modify the combustion process; use new die casting moulds; reduce the scrap rate and use electrical equipment. o We conclude that the technical potentials to reduce the environmental impact vary for the different types of options. Relatively large reduction potentials exist, for example, for fabric filters or wet scrubbers which reduce emissions of heavy metals by 99% (e.g. cadmium and nickel). Some options affect a more than one pollutant. This may happen in different ways. Options may reduce the use of materials that cause environmental problems (e.g. new mould release agent); or replace materials by less pollutant ones (e.g. the use of granular desoxidation agent and the use of nitrogen in the impeller station) or modify the production process (e.g. compact metal load or the use of oxygen) and consequently decrease the use of natural gas. Some options have side effects such as the use of additional materials, energy or the production of additional pollutants, or both. Examples include filters and scrubbers (additional electricity use and production of waste). o The associated implementation costs vary with the type of option. Some options are more expensive than others: calculated net additional costs vary from - 128 to 224 k€/year. The net additional costs are calculated as the sum of annualised capital costs with fixed and variable costs of a certain option. Net additional costs may be negative if the implementation of a reduction option brings revenues for the facility. Chapter 6: Conclusions and Discussion - 129 - Research question 3: “How can a model be developed that can be used from a company perspective to analyse options to reduce the environmental impact of aluminium pressure die casting?” o We conclude that developing a model from a company perspective requires careful definition of system boundaries, processes considered and reduction options included. o The model developed (MIKADO) calculates the main mass flows within the production chain as well as sub-processes and sub-sub-processes that lead to environmental problems. Only those processes are considered that can be managed by the plant managers. Likewise, the pollution reduction options included are applicable by the plant management. o The model calculates the required input of materials and energy based on the production rate. The model inputs include raw materials (aluminium alloy ingots), energy (natural gas and other fuels used on internal transports and in auxiliary burners) and subsidiary materials such as the desoxidation and degassing agents, mould release agents, lubricants, water, steel shot, ceramic abrasives, detergent and other agents used in the wastewater treatment plants. These inputs, the so-called activity rates, are quantified for specific activities and include, for instance, the use of energy or a specific material. The activities rates are used to calculate emissions of pollutants, using an emission factor. o In MIKADO the emissions are directly linked to the materials and energy use in the facility sub-process or sub-sub-process and include air emissions (as aluminium (Al), zinc (Zn), cadmium (Cd), nickel (Ni), lead (Pb), chromium (Cr), copper (Cu) and iron (Fe)); natural gas combustion related emissions (CO2, CO, NOx, NMVOC); hydrogen fluoride (HF) emissions resulting from the use of desoxidation and degassing flux; non-methane volatile organic compounds (NMVOCs) resulting from the mould release agent spraying technique; solid wastes (aluminium dross; furnace linings, steel shot, ceramic abrasives and burrs), liquid effluents and fuel-related combustion emissions resulting from Internal Transports and Auxiliary Burners. o From a company perspective is important that the model output is transparent and relevant. Important MIKADO outputs include emissions of pollutants, the potential environmental impact for a number of environmental impact categories and the overall potential environmental impact resulting from the emission of these pollutants. The environmental impact assessment methodology used follows current approaches in Life Cycle Assessment and Multi-Criteria Analysis. o MIKADO can be used to calculate emission reduction and costs associated with the implementation of pollution reduction options. These costs are regarded as additional costs for the company. For industrial users of the model it is essential that the implementation costs of pollution reduction for the company are quantified. Chapter 6: Conclusions and Discussion - 136 - Figure 6.1. The procedure followed: a stepwise approach. • Stepwise approach Although the sequence of steps followed is based on other systems analysis approaches, there are some differences in the individual steps, between our approach and other studies. The number of steps considered is similar to the studies performed by Jawjit, (2006), Pluimers (2001) and Findeisen and Quade (1997), which are based on the studies by Wilson (1984) and Checkland (1979). However, we observe some differences between our stepwise approach and that of others. In the following, we discuss three of these differences. The first difference is related to Step 2: Evaluation and selection of existing ESA tools. We do not know of other systems analyses in which the selection of analytical tools is described in this level of detail. Selection of tools is not explicitly mentioned as a step in the analysis by other authors describing the methodology of systems analysis (e.g. Findeisen and Quade (1997), Wilson (1984) and Checkland (1979)). Also authors applying environmental systems analysis only discuss the usefulness of the tools but not their selection. (e.g. Pluimers (2001) and Jawjit (2006)). Moreover, Finnveden and Moberg (2005) overview existing ESA tools and conclude that there is a lack of ESA tools that can assess both environmental and economic impacts of organisations and companies. Our approach for the selection of tools may be useful for other studies. It can serve as an example of how to successfully combine a selection of the currently available ESA tools, in order to assist companies in performing studies on environmental impacts of organisations. The characteristics of our DST also can serve as an example of a DST, fulfilling the expectations of companies, to assess environmental performance. We argue that a selection procedure of tools based on characteristics of the desired model is essential in any study aiming at environmental decision support. Step 2: Evaluation and selection of existing ESA tools Step 1: Problem definition Step 3: Identification of pollution reduction options Step 4: Model building Step 5: Model application Step 6: Evaluation of the methodological approach Chapter 6: Conclusions and Discussion - 137 - The second difference is related to Step 5: Model application. In the work by Pluimers (2001), Wilson (1984) and Checkland (1979), the systems analysis includes an optimisation analysis. This results in a selection of the optimal system. For instance, Pluimers (2001) analysed cost-optimal strategies to reduce the environmental impact of greenhouse horticulture in the Netherlands. Our approach is different. We perform another type of systems analysis answering “what if” type of questions. This implies analysis of future trends by using the model. Some systems analysts advise to analyse future trends before building a model (Findeisen and Quade, 1997). Our approach is different, but in line with, for instance, Jawjit (2006). We aim at providing assistance in deciding on a limited number of alternatives (decision analysis). Therefore, we develop a model to analyse user-defined scenarios for pollution reduction by the industry. The model is used to analyse scenarios and thus the scenario analysis is carried out after model building. Advantages of MIKADO over other approaches include its flexibility, transparency and user-friendliness as scenario generator. The user can analyse various environmental management strategies, expressing alternative environmental objectives, and use it to assist decision making. By repetitive analyses, a user can decide on a preferred strategy. However, the model can not be used for optimisation analysis, aiming, for instance, at finding cost-optimal solutions. One may argue that this is a shortcoming of MIKADO. However, a disadvantage of optimisation analysis is the risk for theoretical optima while the results may be more difficult to interpret by the plant managers. Thirdly we look at Step 6: Evaluation of the methodological approach. This step is not mentioned in the systems analysis literature. However, including it as a separate step, it ensures reflecting on the applicability of the approach to other industries or sectors. In this chapter (Chapter 6), we therefore explicitly address the environmental systems analysis approach in terms of iterations performed, the sequence of steps taken and the ESA tools used to the development of our decision support tool. We evaluate the model uncertainties and reveal the implications of the thesis results for industry in terms of the usefulness to combine the seven ESA tools, the identification of the eighteen pollution reduction options and the consequences of the results from the scenario analysis for the aluminium pressure die casting sector. • Comparison with other model studies Our DST MIKADO was built to be used by industrial company managers. This is different from many other models which are mostly meant to be used by environmental policy makers or environmental analysts. MIKADO focuses on the environmental management in an industrial plant. It can be classified as a deterministic model. An alternative would be a stochastic approach, which would have made a quantitative assessment of uncertainties possible. MIKADO is a steady-state model. No dynamics are described in emissions, environmental impact, nor in the demand for production. Further, it describes future trends by calculating (steady-state) results for different years. MIKADO also calculates the cost-effectiveness of pollution reduction and the environmental impact in physical units (e.g. €/unit CO2 equivalents). Other models allow for cost benefit analysis (e.g. the MERGE model by Manne (1995)). MIKADO is also designed to perform scenario analysis and not for optimisation analysis, such as some others (e.g. Pluimers (2001), Brink (2003) and the RAINS model (Alcamo et al., 1990)). Instead, we aim for a flexible tool to analyse possible scenarios based on “what Chapter 6: Conclusions and Discussion - 138 - if” type of questions. A strong point of MIKADO is its capability to evaluate combinations of options, defined by the user, to reduce the environmental impact of an industrial plant. Our results may be expressed in terms of the overall environmental and the associated costs for the company of the pollution reduction strategy. We have chosen to develop our DST MIKADO by combining a selected set of ESA tools, because individual tools are in itself not sufficient. We agree with Wrisberg et al. (2002) that combining tools is needed to overcome weaknesses of individual tools, and because single tools typically are not addressing all relevant questions. Moreover, our integrated environmental study requires combined knowledge from different scientific disciplines (as suggested by Huggett (1993)). Our DST combines parts of seven tools. However, other tools exist (see Chapter 2). As discussed earlier (Chapter 2, section 2.4) they were not included here as a first choice based on the criteria set to select the tools. Nevertheless, they may contain useful elements for studies at the company level. Potentially interesting are, for instance, environmental performance evaluation, cost benefit analysis and total cost assessment. These tools can provide additional information to the plant managers. However, we argue that our selection is sufficient regarding the nature of the plant, and the environmental problems at stake. 6.3.2. Uncertainties Uncertainties in model-based decision support tools may be associated, among others, with model structure and model parameters (Van der Sluijs, 1997; Walker et al., 2003). In the following we address these categories of uncertainties for our DST MIKADO. Uncertainties associated with the structure of MIKADO relate, for instance, to the system boundaries, the processes and pollutants included and the multi-criteria analysis used to assess the plant’s overall environmental impact. Structural uncertainties may be due to incomplete knowledge of the system with respect to the potential environmental effect of the plant. Another source of uncertainty is the perspective of the plant manager that formed the basis for some model characteristics. Plant managers may have a narrow view on environmental management, because of current legislation, or simply by lack of knowledge. This could lead to an incomplete assessment of environmental issues in the model. In our case, we avoided such problems by considering not only production processes and pollution reduction options considered important by plant managers, but also those that appeared relevant from the literature. This resulted in a final structure for MIKADO different than initially defined. For instance, the plant managers initially did not consider filters and scrubbers and alternatives to mould release agent application relevant, because these options reduce pollutants for which environmental standards were already met. As a result, the plant managers did not consider these pollutants of primary interest. However, we decided to nevertheless include these options in MIKADO, because they have a large potential to reduction of the overall environmental impact of the plant. Moreover, implementing these may result in cost benefit for the company. The combination of scrubbers, alternative mould release agent and other more structural options is among the most cost effective strategies for pollution reduction. Another source of structural uncertainty relates to the multi-criteria methods used to assess the overall environmental impact. We used different sets of weighting factors that are not plant specific. This is on the one hand a strong point of the DST. Each set of weighting factors reflects a view on valuating environmental problems, and using several sets illustrates the relative importance of these views on the Chapter 6: Conclusions and Discussion - 139 - environmental assessment. However, the factors included in MIKADO do not necessarily reflect the preferences of the managers of the specific plant for which MIKADO was developed. Therefore, in future analyses we recommend that model users define their own sets of weighting factors. We also recommend using multiple multi-criteria methods, revealing the consequences of subjective choices. A second category of uncertainties is associated with the values of parameters used in MIKADO. These include, for instance, the values used in the environmental impact assessment (i.e. characterisation and normalisation factors). For two environmental problems caused by the plant (aquatic toxicity potential and solid waste produced) no characterisation factors were available, and we therefore used the amount of liquid effluent and solid wastes produced as indicators. In addition, the normalisation factors are uncertain because they are not specific for the region where the plant is located. Rather, they were developed for Western Europe (adopted from Huijbregts et al. (2003)). Other parameter uncertainties are related to the description of the reference case (i.e. presenting the current practice in the aluminium pressure die casting plant) and the pollution reduction options. The uncertain values associated to MIKADO inputs include, for instance, emission factors, costs and pollution reduction factors. Parameters uncertainty may be caused by the system’s inherent variability. This may cause extrapolation errors (e.g. emission factors for air emissions from metals and combustion emissions are extrapolated to annual values), measurement errors (e.g. aluminium alloy mass flows estimated by the company), reporting errors (e.g. reports of activity levels or annual air emissions by the company), or errors in technical developments (e.g. incomplete knowledge associated to reduction potentials, costs and side-effects of the new technologies to pollution reduction). We have addressed the uncertainties of our DST MIKADO only partially by performing a sensitivity analysis. To this end, we analysed the influence of changes in the input parameter values to MIKADO results. The results in terms of environmental impact and costs were compared with the situation presenting the current industrial operation practice (i.e. for the case in which the input parameter values were not changed). This revealed which parts of the model are relatively robust, and which parts are more sensitive to uncertainties. The sensitivity analysis was performed in Chapter 4 and consisted of three sets of analyses in which we changed 48 of the more than 200 parameters. The three sets of analyses performed allowed for analysing (a) the model sensitivity to changes in model parameters for the current industrial operation practice; (b) the model sensitivity to changes in values of a number of parameters that are associated with reduction options and their costs; (c) and the model sensitivity to changes in parameters associated with the alloy mass flow. The partial sensitivity analysis shows that the modelled changes in environmental impact are relatively sensitive to changes in one parameter related to the mass of aluminium alloy recycled internally in the plant. However, the analysis performed could be more complete, as mentioned in Chapter 4 (section 4.4). Alternative sensitivity analyses could include, for instance, changes in model parameters related to other problems than human toxicity or be extended to other processes in the plant. Therefore, one of our recommendations for future studies is a more systematic analysis of model uncertainties that can make use of, for instance, Monte Carlo simulation (see section 6.4). Various alternative methods to assess uncertainties exist, that were not applied here. These range from qualitative assessments of uncertainties to quantitative statistical Chapter 6: Conclusions and Discussion - 140 - approaches. Qualitative uncertainty analysis methods include, for example, data quality rating (such method, used in LCA studies, assigns alphabetical or numerical scores to inputs and parameters to express the uncertainty in a qualitative scale (high-low) (Björklund, 2002). Other methods for qualitative uncertainty analysis include expert’s judgment and qualitative discussion. Quantitative uncertainty analysis include, among others, a comparison of model results with direct measurements (Van Aardenne, 2002), error propagation (this method provides a systematic way of obtaining the uncertainties in results of measurements and computations) (Morgan and Henrion, 1990), uncertainty importance analysis (used in LCA studies this method calculates how the uncertainty of different parameters contributes to the total uncertainty of the result) (Björklund, 2002), Finally, Monte Carlo simulation (also used in LCA studies) allows to generate random values for all uncertain parameters, so-called input scenarios, and for these input scenarios the model outputs are estimated (Kaplan et al., 2005). The methods abovementioned were not explored in this thesis, but we agree that further analyses of the uncertainties are of utmost importance. It may improve the quality of our model or provide insights that can prioritise research needs for the plant’s industrial sector. The exploration of uncertainties can focus on the relatively uncertain process input values, on a large set of parameters values or on the model structure. In summary, we consider that further studies on the decision support tools uncertainties could be done by including a more complete sensitivity analysis addressing the model inputs and model parameters. It may be, followed by a more systematic uncertainty analysis for the significant parameters (e.g. by performing Monte Carlo simulation). The results obtained could be useful for qualitative uncertainty analysis methods including expert’s judgment and qualitative discussion on both the model parameters and model structure used in this thesis. Finally, it can be argued that the stepwise approach taken in this ESA procedure, including the iterations performed, contributed to a reduction of uncertainties. We continuously aimed at using the most reliable sources of information and whenever available we confronted the plant data with industry specific data from the literature. Moreover, model results and parameters were discussed with experts from the plant or compared with actual measurements made in this plant. Nevertheless, uncertainties in the model can not be avoided, but we reduced the uncertainties by refining our DST MIKADO by carrying out several systems analyses iterations in the thesis. All in all, we consider our model adequate for its purpose. The model structure and the model parameters are in line with company specific information, or based on the most appropriate literature. MIKADO can therefore be considered up-to-date and makes use of the best quality data available. 6.3.3. Implications of the results for industry We will now discuss the implications of the results of this study for the aluminium pressure die casting industry, as well as for the metals industry and other industry in general. This study illustrates how the combined use of seven tools (Life Cycle Assessment, Substance Flow Analysis, Multi-Criteria Analysis, Technology Assessment, Sensitivity Analysis, Scenario Analysis and Cost-Effectiveness Analysis) is useful in assessing options to reduce the environmental impact of an industrial plant. The combination of Chapter 6: Conclusions and Discussion - 141 - these seven analytical tools proves to be a solid basis for a DST (MIKADO) that helps the company to consider environmental and economic aspects of decision-making. MIKADO refers to a specific plant in Portugal. However, other industrial companies may also benefit from the results of this thesis. In particular, they may use the method applied here as a tool to improve the company’s environmental management, and use the same tools to assess the company’s environmental performance. We also argue that this method when applied for the same industrial sector may be a valuable instrument for comparison of environmental performances, among different plants. This may be possible by comparing, for instance, the cost-effectiveness of scenarios on pollution reduction, for different plants from same industrial sector. MIKADO is not only useful for industry purposes, but also for other potential participants in environmental management assessments. First, the use of a DST like ours may be useful for environmental policy makers in providing information on the pollution reduction by available techniques that may be implemented in a plant. Second, environmental systems analysts may consider this combination of ESA tools as an interesting example. This study can serve as an example of how to select and combine tools. • Pollution reduction One of most interesting findings of this study for industry is that the aluminium die casting company studied here can earn money by implementing pollution reduction options. This may hold for similar companies as well. The analysis of single reduction options indicates that for some options the annual savings exceed the annual costs. In fact, a significant number of the options are paying options and thus are very promising. They include to compact metal load, the use of a new mould release agent and the scrap rate reduction in Casting and Finishing. Among these options, the ones that appear to be the most cost attractive are the compact metal load and the use of a new mould release agent. This can contribute to reduce the overall environmental impact while the company gains. The other alternatives to pollution reduction are not paying options. In fact, some of the options are expensive, such as the add-on techniques fabric filters and scrubbers. However, these options also have a large reduction potential. Consequently, these costly options can not be ignored and may show to be useful as the environmental policies become more restrict in terms of limiting the amount of emissions released. Therefore, this study indicates that the companies in general and the plant studied in particular may benefit (environmentally and economically) from a proactive behaviour concerning environmental performance. This study also showed that aluminium pressure die casting in general contributes to eight environmental problems, but the largest share of the overall environmental impact is associated with two problems. These two problems are human toxicity (caused by metals emissions and emission of ozone precursors) and abiotic depletion of natural gas. Furthermore, the majority of the overall environmental impact for this industry is caused by sub-processes Melting and Casting. These results suggest that the efforts of the industrial sector to reduce pollution should be focused on these two environmental Chapter 6: Conclusions and Discussion - 142 - problems. In addition, we have listed eighteen options to reduce the pollution that may be used by similar plants. It is interesting to discuss the implication of the results for the Portuguese environmental policy makers. On the one hand, we have seen that there is ample opportunity to reduce pollution by the plant studied, but on the other hand, it should be noted that the plant studied meets the environmental regulations in terms of pollutants emissions. In addition, it should be noted that our model system focuses not only on the pollutants currently regulated, but also on other potential environmental problems including: several emissions, the depletion of natural resources, an in-depth analysis of the industrial production process, and a wide range of pollution reduction options. In summary, the model system goes far beyond the current national environmental policies and thus one may then consider that the current environmental policies are unlikely to effectively reduce the overall environmental impact of industry. Therefore, MIKADO can assist policy makers in deciding on future environmental policy, because MIKADO shows how the environmental perfornmance of an industrial plant that already meets current environmental standards can be further improved. MIKADO can also be used as a communication tool. Companies may have different views on environmental management, and our DST can assist decision makers to illustrate the consequences of having different objectives. The seven different scenarios analysed in this thesis are examples of the types of studies that can be performed using MIKADO. They are useful for reflecting on consequences of different management strategies. They may also assist the dialogue between industry and the environmental authorities when the concern is the reduction of the environmental impact by a company or an industrial sector. MIKADO users can formulate other scenarios reflecting, for instance, user-defined combinations of reduction options. MIKADO could also be used as a communication tool by using it in participatory scenario analysis. In participatory scenario analysis the story lines of the scenarios can be formulated by stakeholders, which may include the plant managers, the national association for metals industry and representatives of national environmental authorities. MIKADO can then be used to quantitatively analyse these qualitative scenarios. 6.4. Recommendations for future studies The overall objective of this thesis was to develop a decision support tool to analyse options to reduce the environmental impact of an industrial company. An integrated environmental assessment model (our DST MIKADO) was developed for calculating the effectiveness of reduction options and the associated costs for the industrial plant. In the following, some recommendations for further studies are presented. Uncertainties in MIKADO may be reduced. For instance, experimental studies, on specific model parameters (e.g. emission factors) are needed. Such analyses may in particular focus on emissions that contribute significantly the overall environmental impact, such as emissions of hydrogen fluoride (so far estimated based on literature) and the emissions of chromium and non-methane volatile organic compounds (so far based on few samples monitored). In addition, the environmental impact assessment factors used in the model can be improved. To reduce model uncertainty, studies on more appropriate characterisation factors for some environmental problems and on Chapter 6: Conclusions and Discussion - 143 - site-specific normalisation factors for Portugal would be useful. These new factors would replace the currently used factors, which mostly came from Western Europe. Following that line, the number of the valuation methods used in MIKADO may be increased. This implies the need for a valuation method that can be easily implemented by the company. Or, as an alternative, the company may develop an internal valuation method itself. The resulting valuation factors preferably express the company’s specific environmental management strategy. A more systematic analysis of model uncertainties is also recommended and can be performed by using, for instance, Monte Carlo simulation. Future research may aim to make MIKADO more complete. This may hold, for instance, for the electricity used by the plant. We took into account the energy conservation in the plant by considering the reduction in the use of natural gas needed for melting. This can be achieved by efficiency improvement of the melting process. As discussed in Chapter 3, electricity used in the plant is not explicitly accounted for in our analysis. This is because of the choices made in terms of the system boundaries; we consider electricity production not manageable by the plant managers. The purpose of the tool is to assist environmental management in the company. Although the environmental problems outside the gates may in fact be significant, the ability of the management board to reduce these external effects is limited. For the same reason machining is now not included in MIKADO. In reality, raw products leave the plant to be machined, and then return to be cleaned by degreasing. As this process takes place outside the gates of the plant, the pollution caused by it was not included in our analysis. Future analyses may, however, include these. In addition, nuisance, odour, vibration, heat wasted and desiccation may be included, even though we assume that their contribution to the overall environmental impact is small. Likewise, the emissions that may occur during plant maintenance operations or in emergency situations, and that are not taken into account in the current version of our DST, may be include in future studies. One may even reconsider the system boundaries and include flows of materials or energy taking outside the plant. The current version of MIKADO only includes pollution reduction options that are currently available. With time, new options to reduce pollution could be included in the model. Alternatively, it would be interesting to analyse more user-defined scenarios with the current version of MIKADO. Finally, MIKADO may serve as an example for other plants, including a wider range of potential industrial users. It should be noted that MIKADO requires first of all a userfriendly interface. Our model can be modified to make it applicable to other plants from the aluminium pressure die casting sector. This would require validation of plantspecific parameters for other plants, or perhaps the whole aluminium die casting sector. An equally interesting option would be to extrapolate MIKADO to the metals industry in general, or even other industries. This can be done by using the methodology underlying MIKADO and apply it to develop decision support tools for other industrial plants. This would reveal the general applicability of our decision support tool and provide a good basis to further explore the general usefulness of the analytical tools combined here, in the systems analyses of companies in general. This could result in different MIKADO versions, filling the current lack of analytical tools that assess both environmental and economic aspects of industrial managenent at the plant level. Chapter 6: Conclusions and Discussion - 144 - References Alcamo, J., Shaw, R., Hordijk, L., 1990. The RAINS model of acidification (Regional Acidification Information and Simulation). Sciences and Strategies in Europe, In: Alcamo, J., Shaw, R., Hordijk, L., (Eds.), Kluwer Academic Publisher: Dordrecht, The Netherlands. Backhouse, C.J., Clegg, A.J., Staikos, T., 2004. Reducing the environmental impacts of metal castings through life-cycle management. Progress in Industrial Ecology Vol. 1 Nos. 1/2/3. Björklund, A.E., 2002. Survey of Approaches to Improve Reliability in LCA. International Journal of LCA 7 (2) 64-72. Brink, C., 2003. Modelling cost-effectiveness of interrelated emission reduction strategies – the case of agriculture in Europe. Wageningen University, PhD thesis. Checkland, P.B., 1979. Techniques in soft systems practice. Building conceptual models, Journal of Applied Systems Analysis 6 41-49. Dalquist, S., Gutowski, T., 2004. Life cycle analysis of conventional manufacturing techniques: die casting. Web.mit.edu/ebm/publications, acceded on 8.5.2006. EMAS Organisations, 2006. Available from http://ec.europa.eu/environment/emas/about/participate/sites_en.htm, last acceded on 15.12.2006. Findeisen, W., Quade, E. 1997. The Methodology of Systems Analysis: An Introduction and Overview, in: Miser, H., Quade, E. (eds.) Handbook of Systems Analysis, Wiley, New York. Finnveden, G., Moberg, A., 2005. Environmental systems analysis tools an overview. Journal of Cleaner Production 13 1165-1173. Huggett, R.J., 1993. Modelling the human impact on nature: systems analysis of environmental problems. Oxford University Press, Oxford. Huijbregts, M.A.J., Breedveld, L., Huppes, G., Koning, A., Oers L., Suh, S., 2003. Normalisation figures for environmental life-cycle assessment The Netherlands (1997/1998), Western Europe (1995) and the world (1990 and 1995), Journal of Cleaner Production 11 737-748. Jawjit, W., 2006. An environmental systems analysis of the Kraft Pulp Industry in Thailand. PhD thesis Wageningen University, The Netherlands. Kaplan, P.O., Barlaz, M.A., Ranjithan, S.R., 2005. A procedure for life-cycle based solid waste management with consideration of uncertainty. Journal of Industrial Ecology vol. 8 n.4 Kim, S.K., Cho, H., Jo, H., Han, M., Lim, S., Hur, T., 2003. Material Life Cycle Assessment for Diecasting Process. Materials Science Forum vols 462-432 3353-58. Manne, A., Mendelsohn, R., Richels, R.G.,1995. MERGE: A model for evaluating regional and global effects of GHG reduction policies. Energy Policy 23: 17-34. Morgan, G.M., Henrion, M., 1990. Uncertainty: A guide to dealing with uncertainty in quantitative risk and policy analysis. Cambridge University Press, New York. Pluimers, J., 2001. An environmental systems analysis of greenhouse horticulture in the Netherlands. PhD thesis. Wageningen University, The Netherlands. Rabah, M., 1999. Cost effectiveness of abatement options for emission control in Egyptian iron foundries. Waste Management 19 283-292. Tan, R.B.H, Khoo, H.H., 2005. Zinc Casting and Recycling. International Journal of LCA 10 (3) 211-18. Chapter 6: Conclusions and Discussion - 145 - Van Aardenne, J., 2002. Uncertainties in emission inventories. PhD thesis Wageningen University, The Netherlands. Van der Sluijs, J.P., 1997. Anchoring amid uncertainty. On the management of uncertainties in risk assessment of anthropogenic climate change. PhD thesis Utrecht University, The Netherlands. Walker, W.E., Harremoes, P., Rotmans, J., Van der Sluijs, J.P., Van Asselt, M.B.A., Janssen, P., Krayer von Krauss, M.P., 2003. Defining Uncertainty – a Conceptual Basis for Uncertainty Management in Model-based Decision Support. Integrated Assessment 4 (1) 5-17. Wilson, B., 1984. Systems: concepts, methodologies and application. John Wiley & Sons, Chichester, United Kingdom. Wrisberg, N., Haes, H.A.U., Triebswetter, U., Eder, P., Clift, R., 2002. Analytical tools for Environmental Design and Management in a Systems Perspective, the combined use of analytical tools. In: Wrisberg, N., Haes, H.A.U., Triebswetter, U., Eder, P., Clift, R., (Eds.), Eco-efficiency in Industry and Science, Kluwer academic publishers.