European LCI Database for Coiled Flat Stainless Steel Products

Transcription

1 European LCI Database for Coiled Flat Stainless Steel Products Methodology Report Prepared by ECOBILAN for: EUROFER Stainless Producers Group April 2000 ECOBILAN L016 MR5 April /156

2 Table of contents Table of contents 2 1. Introduction Objectives Phasing Project organisation Goal and scope General methodology and standardisation for Life Cycle Assessment (LCA) Goals of the study Scope of the study System function and functional unit, products included in the study System boundaries Countries and sites included in the study, geographic coverage of LCI results Data collection Data categories Site Inventory Site breakdown into modules Process stage modules Utility modules Site data requirements Inputs Waste and by-products Air emissions Water emissions Data collection questionnaire presentation Data quality Data audit Questionnaire check procedures and treatment Calculation procedures used for site inventories Allocation procedure applied for by-products (inputs and outputs) General principles 47 ECOBILAN L016 MR5 April /156

3 4.2 Allocation of the process stage inventories between several stainless steel products Allocation of utility inventories Energy allocation Effluent allocation On-site transportation Site upstream and downstream inventory Electricity Electricity Model Sensitivity analysis on electricity grid Stainless steel and non stainless steel scrap Ferro-alloys and pure metals (chromium, nickel and molybdenum) Industrial gases (O 2, N 2, Argon) Other raw materials production Cut-off criteria for raw material production Production models Site external transportation Transportation models Semi-finished product transportation Raw materials and consumables transportation Site downstream processes Oily waste incineration Dust treatment Energy indicators Definition and use Calculation Primary energy Fuel and feedstock Remarks concerning the use of net caloric value (NCV) Conclusion 75 APPENDIX 1: Project deliverables 78 APPENDIX 2: Examples of spreadsheet results Flow Variability 80 APPENDIX 3: Stainless steel product manufacturing route flow diagrams 90 APPENDIX 4: Process stage flow diagrams 97 ECOBILAN L016 MR5 April /156

4 APPENDIX 5: Example of a process stage questionnaire 108 APPENDIX 6: Missing emission calculation and emission ratio 111 APPENDIX 7: Statistics on stainless steel workshops data quality 116 APPENDIX 8: System expansion applied to stainless steel by-products 117 APPENDIX 9: Cut-off rules criteria detailed by process stage 123 APPENDIX 10: Power plant partitioning 128 APPENDIX 11: Allocation of waste water and flows related to their treatment Background and issues Water emissions Ideal approach The emission factor method Emission factor definition Emission factor allocation method Recording of waste water and related flows Water pollution imported by water sources Water outputs Waste water treatment plant consumption Waste water treatment plant consumption recording Common waste water treatment plant consumption allocation Solid waste and by-products of waste water treatment plant 137 APPENDIX 12: Upstream and downstream models 138 APPENDIX 13: List of raw materials and consumables with transportation 146 APPENDIX 14: List of representative Volatile Organic Compounds (VOC) 147 ECOBILAN L016 MR5 April /156

5 APPENDIX 15: Critical Review Panel's Final Comments 148 GLOSSARY 152 ACRONYMS 156 ECOBILAN L016 MR5 April /156

6 Figures List Figure 1: Synthetic view of the process stages succession Figure 2: Representation of a Cradle to Grave system for Life Cycle Inventory (LCI) Figure 3: EUROFER study system Figure 4: Vertical and horizontal averaging methods Figure 5: Example of differences of results between vertical and horizontal averaging methods Figure 6: Theoretical example of a solid material distribution between different treatments and destinations Figure 7 Allocation of the recycling stage and of the raw material saving to the tube life cycle Figure 8: Grid electricity production model Figure 9: Project program Figure 10: Layout of the flow categories on the process stage flow diagrams Figure 11: Energy Inputs table for (2) Hot Rolling Mills Figure 12: Material Inputs table for (2) Hot Rolling Mills Figure 13: Product and Energy co-products table for (2) Hot Rolling Mills Figure 14: Air Emissions table for (2) Hot Rolling Mills Figure 15: Solid Waste and By-products table for (2) Hot Rolling Mills Figure 16: Boiler power plant representation Table List Table 1: List of grade, type and composition of the products covered by the study Table 2: List of the selected routes Table 3: Number of contributing sites per European country Table 4: Cement stabilisation requirements for 1 kg of waste Table 5: List of accounted air emissions Table 6: Optional air emissions (recorded when available) Table 7: List of accounted water emissions Table 8: Optional water emissions (recorded when available) Table 9: Cr, Ni and Mo contents of stainless steel scrap from external supply entered in the EAF Table 10: Bibliographical data used for ferro-alloys production (used only for the upstream and downstream impacts of stainless steel scrap) Table 11: Materials and energy whose consumption may be dependent on the produced grade in process stage Table 12: Electricity Grid European Union Table 13: Electricity grids used for sensitivity analysis Table 14: Results of sensitivity analysis on energy sources for electricity production Table 15: List of raw materials and consumables for which it was asked to fill in the distances of transportation Table 16: Deliverables Table 17: Inventory variability between routes for selected flows (inventory calculated for 304 2B coil) Table 18: Detailed list of studied sites per company for each process stage Table 19: Average emissions ratios (in kg/t of product) for accounted air emissions Table 20: Average emissions ratios (in kg/t of product) for accounted water emissions Table 21: Data quality statistics results Table 22: System expansion assumptions ECOBILAN L016 MR5 April /156

7 Table 23: Upstream models (5 pages) Table 24: Downstream process units (out of the sites) Table 25: Not tracked back inputs Table 26: List of raw materials and consumables for which transportation was asked to be accounted for Table 27: List of representative VOC ECOBILAN L016 MR5 April /156

8 1. Introduction Life Cycle Assessment (LCA) is a tool to assist with the quantification and evaluation of environmental burdens and impacts associated with product systems and activities. LCA is increasingly used by industries, governments and environmental groups to assist with decision making for environmental related strategies. The initial phase of LCA is the collection and calculation of Life Cycle Inventory (LCI) data that quantify the material, energy and emission data associated with a functional system. This stage precedes the Life Cycle Impact Assessment (LCIA) stage that involves classifying, characterisation and evaluating these data in relation to ecological impacts. A further possible stage is the interpretation of data and the potential for improvement through modification of the functional systems. ISO standard for LCI calculation was published in Meanwhile, LCIA and interpretation phase methodologies are under development with ISO standards expected at a later date. With these considerations in mind, EUROFER Stainless Steel Producers Group commissioned Ecobilan to develop a LCI database for a representative range of coiled flat stainless steel grades and product conditions manufactured in Europe. Stainless steel is one of the most versatile and widely used manufactured materials, and the stainless steel industry is anxious to have accurate LCI data for its own use and for its customers and other interests parties. Ecobilan was selected for the database development to provide independent world-class expertise on LCI methodology and to ensure that consistency and methodological rigour is maintained across all the companies participating in the data collection and results calculation phases of the study. As a further measure to ensure methodological rigour in the study and the quality of the data, the study has been reviewed by an external independent Critical Review Panel (CRP) in accordance with ISO 14040: clause This report describes the objectives, the organisation, and the methodology. It is intended to provide total transparency of goals, system inclusions and exclusions, decision rules and methodological sensitivity of the choices made. ECOBILAN L016 MR5 April /156

9 This is to ensure that compatibility with future studies, which may include downstream activities and systems, can be identified and to provide clear references for any changes in methodology required in future. 1.1 Objectives The project objectives were twofold: Produce a Life Cycle Inventory (LCI) from cradle to gate, covering a part of the life cycle of coiled flat stainless steel products, from raw material extraction up to coiled flat stainless steel products ready to leave the manufacturing sites, without taking into account downstream manufacturing into general products, their use, end of life or scrap recovery schemes. Enable each stainless steel company involved in the study to perform an environmental benchmark of their products and processes. The stainless steel products under study are used in industrial sectors including automotive, construction, packaging, chemicals plant and equipment, food and drinks storage and preparation, kitchen appliance and utensils. The objectives are consistent with the goals of the study, as described in section 2.2, and the database forms the basis of any broader LCIs involving the use and application of stainless steel industry products. 1.2 Phasing The project was carried out in four phases, as follows: Phase 0 Goals and scope definition Phase I Modelling of stainless steel production modelling of the stainless steel industry processes: breakdown into independent process units and definition of the inputs/outputs at the boundaries of each process unit, application of LCI methodology rules to stainless steel industry. Phase II questionnaire design, Database building and spreadsheets production ECOBILAN L016 MR5 April /156

10 data collection on site by the stainless steel technicians.0, entering and treatment of data on computer-based LCI/LCA software (TEAM 1 ), production of LCI spreadsheets for individual companies and products and the calculation of European averages. Phase III Software customising and hand-over simulation variables, development of user interface, software delivery and licensing. This methodology report documents the key points of Phases 0 to II including a description of the system model, the methodology rules and data requirements. The list of project deliverables is given in appendix 1, with the study program. 1.3 Project organisation A LCI Expert Group (called Expert Group in this document), composed of stainless steel industry representatives, was formed to provide technical support and to collaborate with Ecobilan in the steering of the project. EUROFER had a role of interfacing and facilitating contact between the two. Meetings were held at all critical stages to define methodology, formulate questionnaires for data collection and correct the database after analyses of the first results. The technical diversity of the participating sites was represented through the Expert Group, whose active participation was key to the achievement of the project objectives. For example, a full understanding of stainless steel industry nomenclature, the processes and material flows was established during meetings of this Group and common mailings. 1 TEAM TM : Tool for Environment Analysis Management is the LCA software tool developed by Ecobilan ECOBILAN L016 MR5 April /156

11 2. Goal and scope 2.1 General methodology and standardisation for Life Cycle Assessment (LCA) The evaluation of industrial systems is not a new subject. The first attempts to evaluate the environmental impacts of a product through its life cycle took place in the seventies and were generally limited to the energy aspects. A suite of International Standards, ISO , describes the principles and minimal requirements for conducting and reporting LCA studies. The ISO 14040, approved by the ISO member bodies in 1997, describes the general principles and framework for Life Cycle Assessment studies. The type of critical review is defined with reference to ISO clause 7. The four stages of LCA study identified in the ISO are: goal and scope definition, inventory analysis, impact assessment, interpretation. The methodology for this study benefited from a recently completed LCI study of carbon steels carried out in 1997 by Ecobilan on behalf of the International Iron and Steel Institute (IISI). In the first step, Goal and scope definition, are defined: the goal: the intended application, the reasons for carrying out the study and the intended audience (ISO 14040). The goal of the EUROFER study is defined in section 2.2. the scope: in brief words, what is studied (what activities) and how it is studied (level of accuracy, limitations, etc.). The scope of the present study is described in section 2.3. The goal and the scope of the study are defined at the early stage of the study before any data collection is initiated. The second step, Life cycle inventory analysis, quantifies the total material and energy flows into and out of the studied system. This requires the collection of numerical data associated with these ECOBILAN L016 MR5 April /156

12 flows and calculation procedures. The methodology for the Life cycle inventory analyses step has been approved in the related ISO standard. The collection of the stainless steel workshops data and their treatment are described in sections 3 and 4 respectively, while section 5 deals with the workshops upstream data (raw material production and transportation). The methodology for the third step, Impact Assessment, has not yet been clearly defined, although proposals emerging through ISO standardisation recommend three stages: - Classification: a mapping of items in the inventory to known environmental effects or impacts (e.g. global warming, acidification, resources depletion, etc.). - Characterisation: a calculation of scientifically-based indices; each index being an estimation of the potential impact of the inventory items contributing to a given environmental effect (e.g. global warming potential, acidification potential, resource depletion index, etc.). - Evaluation: the process of ranking or weighting various indices representing environmental impacts, in order to further 'aggregate' the parameters and aid decision making. Evaluation is a value-based process, not a scientific one. The fourth step, Interpretation, consists of the identification and evaluation of alternative system scenarios that could be used to reduce environmental impact. The EUROFER study includes the Life cycle inventory analysis step and does not at this stage cover the Impact Analysis nor the Interpretation phases. 2.2 Goals of the study The first intended audience of the results of the current project are EUROFER and the stainless steel companies. The second intended audience are the customers of stainless steel manufacturers involved in the study requesting LCI data about stainless steel. This second audience will receive information when the current project database will be ended up using metallic raw materials (chromium, nickel and molybdenum) LCI data. And also, the provider of the data (the stainless steel manufacturers involved in the study or EUROFER) will use this report and the flexibility of the delivered database to provide data for which methodology choices are explained, and as much as possible consistent with the current choices of the customer. The data provided in most of the cases for this second audience will be aggregated to ensure the confidentiality of the site data. Communication to the public is, at the present stage, not envisioned. ECOBILAN L016 MR5 April /156

13 The database is intended to provide EUROFER and stainless steel companies (first intended audience) with robust LCI data to assist with the evaluation of coiled flat stainless steel product environmental impacts and their applications by: benchmarking between the LCI of a given product (made by one company) and the straight average European LCIs of this product (over the studied routes), or between the LCI of a production process (on one plant) and the average LCI for this process (over the plants), assisting with prioritising environmental improvement programmes at process sites, assisting with investment decisions and the integration of environmental criteria to implement EUROFER members' industrial and investment policies, helping to identify strategic guidelines to react to the introduction of alternative materials. The project aims to build a database and develop a common European methodology for coiled flat stainless steel product Life Cycle Inventories (LCIs) across the stainless steel industry companies included in the study. This may subsequently form the basis for full LCAs across broader boundaries and complete product life cycles. The European average LCI values given in this report are based on data derived from the manufacture of coiled flat stainless steel products which represent about 65 % of the total European stainless steel production. Other than coiled flat products, the balance of European stainless steel production is mainly in the form of quarto plate and long products. Whenever LCI values are to be estimated for stainless steel quarto plate or long products on the basis of this study, it is necessary to evaluate the circumstances in each specific case and to be aware of the fact that the uncertainty will be greater than for flat stainless steel products. The methodology has been defined in compliance with the ISO on Life Cycle Assessment and the ISO standard on Life Cycle Inventory. The main methodological points are adapted to stainless steel manufacturing including: functional unit definition, system boundaries, inventory data collection, methodological treatment of by-products and recycling. To provide appropriate reassurance to users of this report, it has been subjected to a critical review by an independent expert panel in accordance with ISO clause ECOBILAN L016 MR5 April /156

14 2.3 Scope of the study System function and functional unit, products included in the study The system function and functional unit definitions are major elements of the LCI methodology. In defining the scope of an LCA study a clear statement on the specification of the performance characteristics (functions) of the product shall be made (ISO 14041). Within the scope of this study, the system function is: the production of a coiled flat stainless steel product at the factory gate 2 The functional unit, which enables the system performance to be quantified and the associated in/outputs to be normalised is: one kilogram of coiled flat stainless steel product at the factory gate 3 Interleaving paper is included in the inventory calculation as it is shipped with the coil to the customer. This means that an inventory calculated for 1000 kg of coiled flat stainless steel product includes the production of 1000 kg of stainless steel as well as the production of 4 to 8 kg of interleaving paper, depending on the producer. No other packaging than interleaving paper is taken into account in the present study. The products included in the study cover the main stainless steel applications: automotive, construction, packaging, chemicals plant and equipment, food and drinks storage and preparation, kitchen appliance and utensils. The list of products is given in Table 1. Their occurrence in the manufacturing route as well as the process stage nomenclature is represented in the route flow diagrams in appendix 3. The various intermediate products included in the present study refer to the five main grades produced by stainless steel companies: 304 (austenitic grade), 316 (austenitic grade), 2205 (duplex 4 grade), 2 i.e. ready to be shipped from the steelworks to the customer. 3 i.e. ready to be shipped from the steelworks to the customer. 4 Grade which is between ferritic and austenitic grades ECOBILAN L016 MR5 April /156

15 409 (ferritic grade), 430 (ferritic grade). These grades were selected because they are the most representative grades of the stainless steel products. Based on the data collected in the participating companies, these grades all together are representative of more than 85 % of the stainless steel products 5 (1997 data). These grades are defined by national and international standards in terms of their chemical composition and other characteristics. The main constituents are iron and chromium with optional additions of nickel, molybdenum etc., according to the grade requirements. Moreover, the process / surface finish combination of these products is defined by a characteristic code, which can be BHR (Black Hot Rolled), WHR (White Hot Rolled), 2B and BA (Bright Annealing). The BA and 2B surface finishing processes include skin passing process. Table 1 gives the precise list of the 18 products that are included in the study. Grades Alloy Types ferritic austenitic duplex 6 Main alloying elements Fe + Cr Fe + Cr Fe + Cr + Ni Content in weight-% 7 11 % Cr 17 % Cr 18 % Cr 8 % Ni Fe + Cr + Ni + Mo 16 % Cr 12 % Ni 2.5 % Mo Fe + Cr + Ni + Mo 22 % Cr 5 % Ni 3 % Mo Surface finishing of the product 8 (flat thin sheets) BHR WHR 2B BHR WHR 2B BA BHR WHR 2B BA BHR WHR 2B BA BHR WHR 2B Table 1: List of grade, type and composition of the products covered by the study 5 Products covered by this study can be sold as coil or cut sheet 6 Duplex grade is between austenitic and ferritic grades 7 These percentages are approximate: exact percentages will be used for each stainless steel production plant. 8 Hot or cold rolled strip, as coil or cut lengths (according to EN 10079) ECOBILAN L016 MR5 April /156

16 Therefore, the LCI will be calculated for 18 products and the results will be available for 18 types of products. 7 routes have been consequently defined. They are listed in Table 2 and detailed in appendix 3. A diagrammatic view of the process stages succession is displayed by Figure 1. N Name of the route Grades 1 Black Hot Rolled Route 304, 316, 409, 430, Continuous Annealing, Pickling, White Hot Rolled Route 3 Batch Annealing, Pickling, White Hot Rolled Route 304, 316, 2205, Continuous Annealing, Pickling, 2B 304, 316, 2205, Batch Annealing, Pickling, 2B Continuous Annealing, Pickling, BA 304, Batch Annealing, Pickling, BA 430 Table 2: List of the selected routes ECOBILAN L016 MR5 April /156

17 (1) EAF Stainless Steel Making (2) Hot Rolling Mill Black Hot Rolled Coils (BHR) (3) Continuous Annealing Pickling (4) Batch Annealing Pickling Continuous White Hot Rolled Coils (WHR) Batch White Hot Rolled Coils (WHR) (5) Cold Rolling Mill (6) Continuous Bright Annealing and Skin-Passing (7) Continuous Annealing Pickling and Skin-Passing BA Coils 2B Coils Remarks concerning packaging: Figure 1: Synthetic view of the process stages succession The packaging for the raw materials were not recorded in the electronic questionnaires, nor the packaging of the final stainless steel products, except for the interleaving paper. This latter is a special Kraft paper utilised for intermediate packaging of the stainless steel coils after cold rolling and final packaging of the coils after skin-passing step. Packaging of the raw materials Based on a sample of participating companies, the ratio packaging / raw materials (refractories, ferro-alloys ) is estimated to represent less than 1%. These packaging materials correspond to a mix of cardboard, wooden crates and pallets, plastic bulk packaging or paper bags. Assuming that all the raw materials are packed with a packaging representing 1% of their weight, then the total quantity of packaging materials per kg of final stainless steel product is about 1% x 1.38 kg = kg (case of the product 304 2B). Assuming that all packaging is made of plastic, which is a worst case assumption since among the current packaging materials, plastic has the highest primary energy (about 60 MJ/kg for polyethylene), then the total primary energy related to raw material packaging is at maximum: ECOBILAN L016 MR5 April /156

18 kg x 60 MJ = 0.83 MJ/kg of stainless steel coil. Therefore this value represents at maximum about 3 % of the total primary energy of this type of stainless steel coils (about 29 MJ/kg, see appendix 2). However, it should be outlined that this ratio is overestimated since, at this stage of the Stainless steel LCI database, upstream data concerning ferro-alloys are not integrated, which tends to under estimate the LCI results of the final stainless steel coils. Packaging of final stainless steel products Before the questionnaire designing, the issue regarding the packaging of the final stainless steel products was raised and information was asked to the participating companies, in order to state what should be recorded in the questionnaire. From their answers, it appeared that: the quantity of the interleaving paper used for the final shipment of the finished coils is not negligible. LCI results show that its quantity represents about 1 % of the weight of the stainless steel coil. Furthermore the technical specificity of this type of paper could mean, a priori, relatively high impact associated to its production; other final packaging (except wooden pallets), corresponding to plastic sheets, cardboard and metallic strips are minors, since representing a maximum of 0.5 % (w/w); wooden pallets used for the stainless steel shipments represent about 1.5 % (w/w) of the stainless steel coils. Hence, considering that the total primary energy of other packaging material (cardboard, plastic, steel) is at maximum 60 MJ/kg and that the total primary energy of wooden pallets is about 12 MJ/kg of pallet, then the total primary energy related to final packaging (interleaving paper excluded) is at maximum: 0.5% x 60 MJ + 1.5% x 12 MJ = 0.48 MJ per kg of stainless steel coil. Therefore, the total primary energy of other packaging represents at maximum 1.6 % of the total primary energy of this type of stainless steel coils (about 29 MJ/kg, see appendix 2). However, it should be outlined that this ratio is overestimated since, at this stage of the Stainless steel LCI database, upstream data concerning ferro-alloys are not integrated, which tends to under estimated the LCI results of the final stainless steel coils. ECOBILAN L016 MR5 April /156

19 2.3.2 System boundaries General principles of product system definition and aggregation A product system is the collection of materially and energetically connected unit processes which performs one or more defined functions (ISO 14040), a unit process being: the smallest portion of a product system for which data are collected when performing a life cycle assessment. Full earth to earth systems, also called cradle to grave systems, have only natural resource (earth) inputs and emissions to air, water, land with all the flows associated with production, use and end-of-life sub-systems contained within the system, as shown in Figure 2. In practice, some sub-systems have negligible contributions to the environmental inputs of the system and may justifiably be omitted from data collection (see section 5.5 and 5.6.3). The system LCI calculation consists of aggregating the inventories (inflows and outflows) of the modular components or sub-systems by summation, taking into account the contribution of each module to the functional unit (e.g. all flows calculated for one kilogram of stainless steel product). Aggregation calculations for this study were carried out using the TEAM TM software. ECOBILAN L016 MR5 April /156

20 Natural Resources Raw Materials Extraction Materials Production Intermediate Products Product Manufacturing Recycling Use Reuse End of Life Solid Waste Emissions to Air Emissions to Water Figure 2: Representation of a Cradle to Grave system for Life Cycle Inventory (LCI) Application to the EUROFER study The EUROFER study is a cradle to gate LCI study: it covers the production steps, from the raw materials in earth (the cradle) to the finished products ready to be shipped from the stainless steel plant (the gate). The manufacture of downstream products (e.g. assembly with other components in a car manufacturing plant), their use (e.g. car use) and end of life (e.g. car component reuse, recycling, incineration (with or without energy recovery) and/or landfilling) are not taken into account by this study. In addition, the collection of data concerning chromium, nickel and molybdenum ores extraction and their transformation into pure metals or ferro-alloys are not included in the scope of the present study. The flows corresponding to the ferro-alloys and the pure metals ECOBILAN L016 MR5 April /156

21 inputs (e.g. pure nickel) appear in the final delivered inventories in order to enable the future integration of those data to get the global cradle to gate LCI. The respective Nickel, Chromium and Molybdenum projects were launched and are carried out by the Ecobilan Group. The kick-off meetings were held, setting the scope for those projects. The methodologies chosen for the three projects are consistent between them and with the Stainless Steel project, including the flow classification and the databases format. Particular attention is paid to the LCI for products used by the stainless steel industry, in order to obtain consistent data. These projects will be finished at the end of 1999 at the earliest. The integration of these upstream data in the present stainless steel project will be considered in a separate project. This complementary work will result in the edition of the final cradle to gate results for the selected coiled flat stainless steel products. System boundaries Chromium, Nickel and Molybdenum pure metals Chromium, Nickel and Molybdenum ferro alloys Natural resources from earth Merchant scrap, other steelworks, etc. Raw material and energy production (including extraction) except for Cr, Ni and Mo Consumables production ferro alloys Transportation Scrap pure metals Site boundaries Stainless steel workshops Recovery processes minus Byproducts Stainless steel products Scrap Emissions to earth Saved operations By-product functions Figure 3: EUROFER study system As shown in Figure 3, the stainless steel product manufacturing system encompasses: the stainless steel plant system the boundaries of which are defined by the geographical boundaries of the site; ECOBILAN L016 MR5 April /156

22 the activities upstream of the stainless steel plant, i.e. the production and transportation of the raw materials, energy sources and consumables used by the steelworks. As already outlined, the extraction of chromium, nickel and molybdenum ores and their transformation into ferro-alloys and pure metals are excluded from the system due to lack of data at the present time. Once these data are available they will be used by EUROFER for final cradle to gate LCIs calculations, in the frame of a separate project. Certain upstream stages, which have a small contribution to the resulting LCI are excluded (see section and 5.6.3); the recovery processes outside the plants of the stainless steel by-products that are taken into account using the system expansion method recommended by the ISO to avoid partitioning, as explained in section 4.1; the operations saved by these stainless steel by-products, see same section 4.1 dealing with the by-product allocation procedures used in the study. The landfill sites which receive stainless steel workshops waste are not included in the study. The waste tonnages are recorded but the leachates and the air emissions emitted by the landfill sites are not accounted, neither the input/output related to the life cycle of the waste from the plant gate to the landfill site (transport and waste treatment) 9. The routes of scrap recovery and processing outside the steelworks are not included in the system, except their transport to the site (see section 5.2 and 5.6) Countries and sites included in the study, geographic coverage of LCI results The study covers 14 sites (see Table 3) located in the following 7 European countries: Belgium (2 sites), Finland (1 site), France (4 sites), Germany (3 sites), Italy (1 site), Sweden (2 sites), United Kingdom (1 site). 9 It is demonstrated in section that the treatment of the waste before landfilling has a negligible potential environmental impact. ECOBILAN L016 MR5 April /156

23 This coverage was selected to fulfil the goal of building a European database. A table detailing the studied sites for each process stage that was defined is included in appendix 4 (see Table 18). The contributing sites are among the largest belonging to major stainless steel companies from most of the principal European producer countries. The study is one of the largest undertaken in the field of LCI. Hence it is estimated that it covers about 65 % of the total European stainless steel production (i.e. all types of products included, i.e. hot-rolled flat products, cold rolled flat products, quarto plate and narrow strip). Moreover, when considering only the coiled products studied in this LCI database (i.e. hotrolled coiled flat products and cold-rolled coiled flat products), the participating companies deliveries correspond to about 85 % of total European deliveries of these products. At last, within the participating sites, the studied grades, i.e. 304, 316, 409, 430 and 2205, represent more than 87 % of the total production of these sites. ECOBILAN L016 MR5 April /156

24 Steel Making sites Hot Rolling sites Belgium Finland France Germany Italy Sweden United Kingdom 1 Genk 1 Carlam 1 Tornio 1 Tornio 2 L'Ardoise Isbergues 1 Fos 2 Bochum Krefeld 1 Bochum 1 Terni 1 Terni 1 Avesta 1 Avesta 1 Sheffield Total Annealing- Pickling sites 1 Genk 1 Tornio 2 Gueugnon Isbergues 2 Benrath Krefeld 1 Terni 1 Nyby 1 Sheffield 9 Cold Rolling sites 1 Genk 1 Tornio 2 Gueugnon Isbergues 2 Benrath Krefeld 1 Terni 1 Nyby 1 Sheffield 9 Surface treatment sites (BA and 2B) 1 Genk 1 Tornio 2 Gueugnon Isbergues 2 Benrath Krefeld 1 Terni 1 Nyby 1 Sheffield 9 Total of sites = Table 3: Number of contributing sites per European country At the end of phase II, each company that could finish the data collection received one spreadsheet per product manufactured in its plant(s), within the limit of the products covered by the study. Statistics were also delivered to companies. These statistics include average, minimum, maximum and standard deviation. Averages are calculated without weighting the contribution according to companies product tonnages and per product. LCIs are calculated first for each route involving the processes of the given route only with its given yields and the resulting LCIs are averaged across several routes. This averaging method (also called vertical average ) allows to carry out benchmark between companies on the same basis (1000 kg of a given stainless steel product). The vertical averaging method is different from the horizontal one, where each process is averaged first, and then the LCI is calculated (see Figure 4). ECOBILAN L016 MR5 April /156

25 HORIZONTAL AVERAGING METHOD Company 1 Company 2 Company 3 Operation 1 Operation 1 Operation 1 average for operation 1 Operation 2 Operation 2 Operation 2 average for operation 2 Operation 3 Operation 3 Operation 3 average for operation 3 Operation 4 Operation 4 Operation 4 Average calculated as weighted mean VERTICAL AVERAGING METHOD Company 1 Company 2 Company 3 Operation 1 Operation 1 Operation 1 Intermediate average at end of operation 1 Operation 2 Operation 2 Operation 2 Intermediate average at end of operation 2 Operation 3 Operation 3 Operation 3 Intermediate average at end of operation 3 Operation 4 Operation 4 Operation 4 Average calculated as weighted mean Figure 4: Vertical and horizontal averaging methods ECOBILAN L016 MR5 April /156

26 vertical average horizontal average route 1 route 2 Process 1 In: A: 110 Intermediate product B out:100 out: CO2: 5 Process 1 In A: 140 Intermediate product B out: 100 Out CO2: 30 Process 1 Averaged In A: 125 Intermediate product B out: 100 Out CO2: 17.5 Process 2 In: B: 105 Intermediate product C out:100 Out CO2: 10 Process 2 In B: 130 Intermediate product C: out: 100 Out CO2:20 Process 2 Averaged In B: Intermediate product C out: 100 Out CO2: 15 Process 3 In: C: 110 Final product D out:100 Out CO2: 5 Process 3 In C: 120 Final product D out: 100 Out: CO2: 30 Process 3 Averaged In C: 115 Final product D out: 100 Out CO2: 17.5 LCI results for product D (real) with route 1 In A: = 110*105/100*110/100 Out D: 100 Out CO2: = *110/ *110/100*105/100 LCI Results for product D (real) with route 2 In A: = 120*130/100*140/100 Out D: 100 Out CO2:100.8 = *120/ *120/100*130/100 50% 50% LCI Results of vertical averaging method: In A: Out D: 100 Out CO2: LCI Results of horizontal averagin method: In A: = 115*117.5/100*125/100 Out D: 100 Out CO2: 58.9 = *115/ *115/100*117.5/100 Figure 5: Example of differences of results between vertical and horizontal averaging methods ECOBILAN L016 MR5 April /156

27 When averages are calculated without weighting the contribution according to companies production, the resulting inventories from "horizontal" and "vertical" averages are slightly different as seen in Figure 5. By using "Vertical" averaging in the study, each company receives their own product s inventory. In the case of weighted calculation, there is no difference between vertical and horizontal LCI average. 2.4 Data collection The stainless steel workshops data were collected on site (see section 3) with rules and procedures which were established during the initial Expert Group meetings and clearly explained in the questionnaire user s guide sent to the sites. For most of the other process units encompassed in the system, i.e. upstream, by-product recovery processes and production of stainless steel by-product alternative materials, data were not collected during the study, but came from Ecobilan database and literature (see section 5). An exception is made for pure metals and ferro-alloys (chromium, nickel and molybdenum) upstream for which specific data are foreseen to be collected within the scope of separated studies. Therefore, the data of the process units for which non-specific data were used are likely to be different in quality, in terms of reliability, geographical relevance, methodology and completeness: reliability: some data are not collected on site but come from literature or from deductive calculation (emission calculated from element balance or fuel combustion ratio) ; geographical relevance: European average data were used for all upstream process units even significant differences may exist in the production between countries, and regions. An example of this is the combustion of fossil fuels in the grid electricity power plants: the efficiency and the emission ratio may change significantly from one country to another; This choice is made in order to improve the benchmark possibilities and to take into account the fact that the exchanges of products and energy is made at the European level. The individual databases (for private companies) will be delivered to companies (if ordered) with the country specific Electricity production model (and parameters to handle the possibility to take into account any other model, as the European one). methodology and completeness: some LCI models found in the literature are not complete, some significant air emissions can be missing. The methodology applied can be different from the EUROFER study methodology. For example, economical partition may have been used whereas it was never considered in the EUROFER study. The latter choice is due to the fact that the main the raw materials used in the stainless steel ECOBILAN L016 MR5 April /156

28 production (i.e. stainless steel scrap, nickel, chromium and molybdenum), are likely to undergo important prices variations due to the market versatility. In other cases, it may occur that the LCI results are poorly documented concerning how data were collected and treated so that it is not possible to identify possible data gaps and/or methodology issues. Moreover, upstream data can be easily updated in the TEAM TM EUROFER when better data become available. database delivered to All the upstream data that have been used for the LCI calculation are presented in Appendix 12. This appendix presents in details the sources of the data, their quality as well as their representativeness Data categories In LCI studies, the data are usually split into the following categories: inputs: raw materials, energy and consumables, water consumption recorded according to its origin (recorded in the inventory results as Water: Origin e.g. Water: Industrial Network ) emissions to air (recorded in the inventory results as (a) Name of the flow e.g. (a) Carbon Monoxide (CO) ), emissions to water (recorded in the inventory results as (w) Name of the flow e.g. (w) Chlorides (Cl - ) ), emissions to soil, waste, also called improperly emissions to land, and obviously, the products and co-products. The scope of the EUROFER study does not cover all the data categories and some emissions were not systematically completed: soil emissions (contaminated land) were outside of the study scope, a closed list of air and water emissions was systematically completed for all process units, calculated when missing in a site questionnaire. The list of these accounted emissions is given for respectively air and water emissions in section and Data from other emissions were not systematically completed, because not enough data were available both in the stainless steel workshops and in the upstream models. Examples of these non-accounted emissions are metallic emissions, organic compound air emissions such as VOC 10. When readily available, non-accounted emission data were recorded in 10 VOC: Volatile Organic Compounds. A list of representative VOC is presented in Appendix 14. ECOBILAN L016 MR5 April /156

29 the site questionnaires and integrated in the database. In the spreadsheets, LCI results for these non-accounted emissions are displayed as minimum and maximum values. Since there are some data gaps in the upstream models and/or parts of the stainless steel workshops, the average given for the non-accounted emissions may not be reliable. The main reasons to set a closed minimum list of accounted air and water emissions are the following: For establishing a minimum list: 1. To have the sites fill in the minimum amount of same information, in order to be consistent with ISO which recommends to gather data which should be consistent between sites if comparison are done. Here no comparison is done, but a benchmark, meaning that the sites may compare to an average. For the methodology of replacing a missing value for one flow by the average of those flows: 1. When one value is judged to be a base for the comparison, then it is important that the average has a representative order of magnitude. In order to get this representativeness, the missing values of the minimum list of flows is replaced by the average of existing ones. 2. When a value is not in the minimum list of flows, it should not be used for benchmark. For the selection of some flows: 1. To take into account the flows that are most often gathered during LCA (the current practice). 2. To take into account the flows that are specific to the stainless steel industry. 3. To be able to apply the main impact assessment methods to the LCI that is calculated during this project, with a good quality. 4. They are available on site, and they are of importance for the site experts. For the non-selection of some flows: 1. There is no measurement available on all sites, and they are judged to be not important by the site experts. 2. They have no known impact on the environment (from the point of view of impact assessment methods and/or a negligible one). 3. They are not currently used in LCA practice. ECOBILAN L016 MR5 April /156

30 3. Site Inventory 3.1 Site breakdown into modules In-order to facilitate data collection and evaluate the contribution of unit processes to the overall system inventory, the stainless steel manufacturing site processes are broken down into independent units or modules. Modules are defined to allow the discrete allocation of environmental impacts to the intermediate functional flows such as slabs, hot rolled coils or cold rolled coils. Each unit process is represented by a separate column in the spreadsheet so that its contribution to the cradle to gate LCI is readily available. Each module definition (boundaries, inputs and outputs) was carefully decided in agreement with the Expert Group in order to ensure consistency of methodology and data collection at the different sites. A module corresponds: either to one of the main industrial process stages and its associated ancillary workshops (e.g. skin pass mill + tension leveller + gas dedusting), or to a common utility to the site (e.g. a power plant producing steam or compressed air, a waste water treatment plant common to several workshops,...). 3.2 Process stage modules A process stage module is designated by the main element (e.g. Cold Rolling Mill) or its function (e.g. EAF Stainless Steel Making), preceded by a figure in brackets, following the chronological sequence of process stages in the manufacturing route. For example, (1) EAF Stainless Steel Making, is defined to include the electric arc furnace, converter (AOD or VOD), ladle metallurgy, continuous casting and grinding, by-products plants, effluent treatment plant and other dedicated ancillary plants. Another example is (3) Continuous Annealing Pickling that includes continuous annealing, descaling and pickling as well as ancillary units that are dedicated to these workshops. ECOBILAN L016 MR5 April /156

31 Stainless steel workshops systems include complex flow streams between unit processes, including the import and export of intermediate products and the recycling of by-products and waste. The resulting system flows are often non-linear and require a detailed description of system interactions so that the system can be adequately modelled. The process stage modules and their main connections are illustrated by the route flow diagrams (see appendix 3). Detailed flow diagrams are drawn for each process stage (see appendix 4). These flow diagrams describe the process stage make up and their associated inflows/outflows. This is intended to ensure the consistency of data collection and to avoid national and European variations regarding data entry and terminology. They have been established in cooperation with the participating members of EUROFER through the Expert Group. The latter agreed on the 7 following process stages: (1) EAF Stainless Steel Making (2) Hot Rolling Mills (3) Continuous Annealing - Pickling (4) Batch Annealing Pickling (5) Cold Rolling (6) Continuous Bright Annealing and Skin-Passing (7) Continuous Annealing Pickling and Skin-Passing Due to the European distribution of the contributing sites, the modules and process stage definition represents wide technology coverage. Indeed the aim is to have an exhaustive representation of each process stage. The process stage questionnaires created to assist site data collection are directly derived from the list of inflows/outflows established during the process stage flow diagram definition and modelling. Additional inflows and outflows unique to certain sites were also registered, when significant, directly by the site staff. ECOBILAN L016 MR5 April /156

32 3.3 Utility modules Energy production, solid waste treatment and effluent treatments are major elements that directly affect the LCIs, thus an inventory of the following utilities was made for each site: power plants (producing steam, hot water and compressed air), waste water treatment plants, The other utilities, for which site data or supplier data were incomplete, derived from other sources (e.g. companies suppliers) or standard LCI data. Typically this includes industrial gas plant (oxygen, nitrogen etc.), de-ionised water units... These European average data were provided either by Ecobilan database (DEAM TM 11 ) or collected from industrial sites or companies (see section 5). For instance data concerning the EAF dust treatment process were collected from several subcontractors of the participating companies. 3.4 Site data requirements Site inventory data were collected by the site staff and recorded in electronic spreadsheet questionnaires designed by Ecobilan and deriving from the site representation work previously described (section 3.2). All the inputs/outputs quantities correspond to one year plant operation (most often 1997 year) so that the data are representative and correspond to typical time averaged site production Inputs The term input covers raw materials, energy, consumables and semi-finished products that are utilised by the stainless steel workshops. As in any LCA study, not all of the site inputs of the stainless steel workshops were recorded: a lot of inputs are consumed in small quantities and their contribution to the final LCI is negligible. 11 DEAM stands for Data for Environment Analysis and Management ECOBILAN L016 MR5 April /156

33 As pointed out in the ISO 14041, there are several criteria used in LCA practice to decide which inputs will be studied including 1) mass, 2) energy, 3) environmental relevance. Studied means here that the input upstream (production and transportation) is taken into account in the LCI calculation. According to the ISO An appropriate decision rule, when using mass as a criteria, would be to require the inclusion in the study of all inputs that cumulatively contribute more than a defined percentage to the mass input of the product system being modelled. Practically, two levels of cut-off criteria were established with the Expert Group: 1. the first level of cut off criteria, described hereafter, defines the limitation of input data collected on site; 2. the second level of cut off criteria defines among data collected and recorded in the questionnaire, those whose upstream is taken into account. 1. Concerning the first level of cut-off rules, decision rules were established to restrict data collection to significant inputs as follows. Where data were available, the input flow was recorded. Where data are not readily available this material may be omitted with the following qualifications: All energetic inputs, including energy data, electricity, steam, water, compressed air are recorded, 99.9 % (w/w) of total process stage inputs must be recorded. As shown below, this rule strongly ensured that 99 % (w/w) of the mass-input are captured for the LCI system boundary. the omitted inputs are considered to not have an important environment burden. 2. Concerning the second level of cut-off rules, in order to get good quality results which might satisfy the goals of the study, the Expert Group decided that: 95% of the inputs in mass of the steelworks process stages, main input excluded 12 ; The appendix 9 presents the cut-off rules applied, for each process stage, for the upstream production processes selection. 99.9% of the inputs in mass, main input included, would be considered. 12 In process stage (1) EAF Stainless Steel Making, the main input corresponds to stainless steel scrap and carbon steel scrap, whereas in the other process stages, the main input corresponds to the intermediate stainless steel product (e.g. slab for process stage (2) Hot Rolling Mills). ECOBILAN L016 MR5 April /156

34 The sites were instructed to record at least 99.9 % of mass input in the questionnaires. Providing this has been respected, then the mass of omitted inputs calculated for each process stage is a maximum of 0.1% 13 of total mass input. Concerning the third cut-off criteria for data collection, it is difficult to know a priori, whether the environmental burden of a material has a significant contribution in the final LCI result. Typical criteria for the importance of environmental burden consider whether: - the material contains any highly toxic compounds: cyanide, phenol and hexavalent chromium or any other significant hazardous compounds generated in stainless steel making, - the financial cost is high since this can be used as an environmental burden indicator (high financial cost is often related to high energy value or to the scarcity of the resource) Waste and by-products Distinction between waste and by-product The distinction between waste and by-product is a preliminary step for any allocation procedure as stated in the ISO 14041: some outputs may be partly co-products and partly waste. In such case, it is necessary to identify the ratio between co-product and waste since the burdens shall be allocated to the co-products part only. This rule is particularly relevant in the case of stainless steel manufacturing since a material (scrap, slag, dust, sludge, etc.) generated by a process stage can have different destinations. As illustrated in Figure 6, the material can be: partly landfilled, partly recovered outside the site (e.g. EAF Converter slag is recovered for use in cement or in aggregates for embankment/landscaping), 13 Mass of recorded input = x total mass of input Mass of missing input = x total mass of input Hence: mass of missing input = 0.001/0.999 mass of recorded input 0.1 % w/w ECOBILAN L016 MR5 April /156

35 the remaining being recovered within the site (e.g. metal is recovered from EAF Converter slags and is consumed in the EAF). Site boundaries System boundaries Internal Landfill 5 t 1 t External Landfill (A) Stainless Steel Making EAF and Converor Slags: 20 t 10 t Internal recovery (e.g. metal content recovery) 10 t 4 t External recovery 4 t 2 t 2 t Cement Making Embakment/ Making Landscaping Figure 6: Theoretical example of a solid material distribution between different treatments and destinations The different final destinations at the site boundaries that are considered in the study are: - landfill (waste), possibly via a treatment, - external recovery (by-product). By-products that are stored in temporary landfills before being recycled are not recorded as waste. A waste is always assumed to be definitively landfilled. Incineration outside the site (with or without energy recovery) was considered. A model developed internally by Ecobilan for the incineration process (with or without energy recovery) was applied to calculate the environmental burdens associated to this operation. Flows of internally recovered materials were also recorded, the internal recovery tonnage is the difference between the total tonnage produced and the sum of the tonnage landfilled plus the tonnage recovered outside the site. ECOBILAN L016 MR5 April /156

36 Flows of internal recovered materials do not cross the site boundaries, so one could consider that the recording of these internal flows are not needed. They were however recorded because: the manufacturing route of a given product is only a portion of the global steelworks. A material generated and recovered on site may or may not cross the boundaries of a given product manufacturing system. When recovered material flows do not cross the product manufacturing system boundaries, no allocation procedure is applied since the flow is internal. When recovered material flows cross the product manufacturing system, they are treated as external recovery. the flows of internal recovered materials are also needed for the carbon and others elements (Fe, Cr, Ni, Mo) balance calculations. These balances were automatically calculated for each module in the electronic questionnaires. The carbon and other elements (Fe, Cr, Ni, Mo) balances were used to check the data consistency (see section 3.5) and calculate the emissions of carbon dioxide when non available on site (see section 3.4.3). In the example illustrated in Figure 6, the EAF slags outputs are recorded in the (1) EAF Stainless Steel Making module outputs as follows: Waste: EAF slags = 5 t + 1t Recovered Material: EAF slags = 4 t Internal Recovery: EAF slags = 10 t In the database, internal recovered materials were recorded with a minus sign in the inputs of the process stage where they are produced and as positive inputs to the process stage(s) where they are consumed. This gives a direct balance of internal recovered materials in the LCI results Additional information concerning waste Waste recording Because of regulation differences between countries concerning waste characterisation, the waste was not classified in categories (inert, industrial, hazardous, toxic, etc.) according to their potential environmental impact. The name of waste origin and its physical nature were kept in the database (e.g. EAF -> slab cutting dust, Hot Rolling Mill sludge, etc.) as a first indication of its potential environmental impact. Waste representing small tonnage, less than 1 % w/w of total waste tonnage for a given process stage, was not recorded unless treated outside the site. ECOBILAN L016 MR5 April /156

37 Waste treatment outside the site All waste treated outside the site was identified in the questionnaire and in the database. Within the scope of this study, the environmental burden relating to waste treatment outside the site were not included except for the following waste: - used oils which are incinerated (with or without energy recovery), - used grease which are incinerated (with or without energy recovery), - all the products produced by the steelworks that undergo an EAF or plasma process (i.e. EAF -> slab cutting dust, Shot blasting dust,..). These treatments were accounted for because they are high energy requiring. Within the stainless steel industry, other external treatments should be relatively minor and should make only a small contribution to the system inventory. Thus within the scope of this study, the environmental burden related to waste treatment outside the site (except for the ones presented above) have not been included. For an average final stainless steel product, the average quantity of waste treated outside the site per kg of final product is kg, corresponding to 304 2B coil. The sensitivity of excluding waste treatments is illustrated in Table 4 which gives the requirements for stabilising waste with cement, which is an extreme example since this treatment is considered as one of the most energy intensive treatment for waste. Requirement Quantity per kg of waste Primary energy equivalent (mostly fuel energy) Cement 0.5 kg 0.5 x 5 = 2.5 MJ Electricity kwh x 3.6 x 3 (max) = MJ Diesel oil (for landfilling) litre x 37.5 = 0.11 MJ Total 2.7 MJ Table 4: Cement stabilisation requirements for 1 kg of waste ECOBILAN L016 MR5 April /156

38 The primary energy requirement for the maximum level of waste treatment is in this case estimated to 2.7 MJ/ kg of waste x kg of waste = MJ per kg of final stainless steel coil, which is negligible (~0.2 %), as compared to the primary energy for this product manufacturing: in average 29 MJ per kg of 304 2B coil Waste landfilling Internal landfills (lagoons included) and external landfill were not distinguished. Although they are located inside the geographic site boundaries, internal landfills were considered outside the site boundaries because landfilling, in both internal and external landfill, is a flow to nature. Soil emissions (leachates) were not considered in this present study Air emissions An inventory of all known significant air emissions was drawn up for each module with the assistance of the Expert Group. The substances included in this list are hereafter called accounted emissions. The air emissions which are currently considered to have neutral environmental impacts such as oxygen (O 2 ), steam and nitrogen (N 2 ) were not recorded. Air emissions data were supplied by the sites, except CO 2 that was either provided by the sites or calculated with the carbon balance within the spreadsheet questionnaires during data collection. A code was devised, presented in appendix 5, which differentiates non-available, non relevant and real nil data in the questionnaires. This is particularly important for the air emissions where monitoring requirements differ from one country to another according to the local regulations. The list of accounted emissions is presented in Table 5. The reasons for the choice of the flows in the minimum list are given in section Whenever data were not supplied for these emissions, the values were based on calculation rules given in appendix 6. ECOBILAN L016 MR5 April /156

39 Name Carbon Dioxide (CO 2, fossil and mineral) 14 Carbon Monoxide (CO) Chromium Compounds (as Cr) Dioxins (unspecified) Molybdenum Compounds (as Mo) Nickel Compounds (as Ni) Nitrogen Oxides (NO x as NO 2 ) Particulates (unspecified) Sulphur Oxides (SO x as SO 2 ) Unit g g g g g g g g g Table 5: List of accounted air emissions Air emissions data relative to the metallic compounds (Cr, Ni, Mo) are calculated based on the amount of particulates emitted and on the metallic content of these particulates. All other air emissions were not completed for the European database when no data were supplied. Therefore, LCI results for these non-accounted emissions, namely HCl, HCN, other metals than Cr, Ni and Mo, etc., were included only for those sites and products where data are available. These optional emissions are listed in Table 6. Regarding VOC 15 emissions, for which numerous regulations currently exist, the Expert Group outlined that their definition is still vague. However, the possibility to add this flow in the minimum list was envisaged but due to the fact that not enough companies could fill in this information, this flow was not included in the minimum list of the accounted emissions. Moreover, based on the data collected, it appeared that dioxin emissions were systematically measured in the first process stage (1) EAF Stainless Steel Making and were estimated by the site staff to be null for the other process stages. Therefore, although this flow was first excluded from the minimum list of accounted emissions, it was finally added to it. On the other hand, the flow of methane, which was first included in the minimum list, was removed from it due to the fact that almost no site entered this data. 14 CO 2 (fossil) includes both CO 2 from fossil materials (e.g. CO 2 generated by heavy-fuel oil combustion) and CO 2 from other mineral materials, such as CO 2 generated by limestone (CaCO3) reduction to produce quick lime, (CaO). CO 2 from other origins like biomass is negligible in the case of stainless steel making. 15 VOC: Volatile Organic Compounds ECOBILAN L016 MR5 April /156

40 Name Cadmium Compounds (as Cd) Fluorides (as F-) Hydrocarbons (except methane) Hydrogen Chloride (HCl) Hydrogen Cyanide (HCN) Hydrogen Fluoride (HF) Hydrogen Sulphide (H2S) Hydroxides (as OH-) 16 Lead Compound (as Pb) Mercury Coumpounds (as Hg) Metallic Compounds (unspecified) Methane (CH4) Oils (unspecified) Tin Compounds (as Sn) VOC (Volatile Organic Compounds) Zinc Compounds (as Zn) Units g g g g g g g g g g g g g g g g Table 6: Optional air emissions (recorded when available) It is important to highlight that some substances presented in Table 5 and Table 6 are intersecting each other. For example, the category Metallic Compounds includes Chromium Compounds. In this case, the site staff was asked to record the data as precisely as possible and to strictly avoid double counting. Example: on a given site, a measure was only carried out for the total of metals, without distinguishing chromium, nickel and molybdenum compounds. In this case, a value was filled in only for the category Metallic Compounds. On the contrary if data are available on the detailed emissions, values were filled in for the categories Chromium Compounds, Nickel Compounds etc. and a zero was filled in the category Metallic Compounds. Air emissions do not vary according to the grade being produced. 16 Hydroxides quantity of which is expressed as OH- ECOBILAN L016 MR5 April /156

41 3.4.4 Water emissions As for air emissions, a list of accounted water emissions was established. This list is presented in Table 7. The reasons for the choice of the flows in the minimum list are given in section The sites were instructed to systematically record these emissions when data were available, that is to say all water emissions listed were considered as accounted. When data are missing, there were calculated using emissions factors obtained from the sites that have recorded these water emissions, as explained in appendix 11. These calculations rules are presented in appendix 6 and 11. The pollutants contained in the water consumed are recorded as inputs of the inventory. ECOBILAN L016 MR5 April /156

42 Name Acids (H + ) Aluminium (Al 3+ ) Ammonia (NH + 4, NH 3, as N) Cadmium (Cd ++ ) Chlorides (Cl - ) Chromium (Cr III, Cr VI) Chromium (Cr VI) COD (Chemical Oxygen Demand) Copper (Cu +, Cu ++ ) Fluorides (F - ) Hydrocarbons (unspecified) Iron (Fe ++, Fe 3+ ) Lead (Pb ++, Pb 4+ ) Manganese (Mn ++, Mn 4+ ) Molybdenum (Mo ++, Mo 3+,Mo 4+, Mo 5+, Mo 6+ ) Nickel (Ni ++, Ni 3+ ) Nitrates (NO - 3 ) Nitrogenous Matter (unspecified, as N) Phosphorous Matter (unspecified, as P) Silicates (Si ++ ) Sulphurated Matter (unspecified, as S) Suspended Matters (unspecified) Tin (Sn ++, Sn 4+ ) Zinc (Zn ++ ) Units g g g g g g g g g g g g g g g g g g g g g g g g Table 7: List of accounted water emissions After data collection completion, the first list that was established at the beginning of the project was revisited. Hence, the flows presented were excluded due to the fact that they had never been (or very rarely) filled in by the sites: ECOBILAN L016 MR5 April /156

43 Name Biochemical Oxygen Demand (BOD5) Cyanides (CN-) Degreasing Agents Metallic Ions (unspecified) Mercury (Hg+, Hg++) Chromium (Cr III, Cr VI) Units g g g g g g Table 8: Optional water emissions (recorded when available) Water emissions do not vary according to the grade being produced Data collection questionnaire presentation The questionnaires comprise listings of registered inputs/outputs at each module to be quantified by the site staff. The questionnaires were created using a spreadsheet software (Excel 7.0 or Excel 97) and include element (Carbon, Iron, Chromium, Nickel and Molybdenum) balance calculations to help site staff to avoid miscalculation, units or typing errors. An example of questionnaire is presented in appendix 5 of this report Data quality The questionnaires also require data quality information to be completed for each flow in coded form. Quality information deals with the source, the type and the time reference of the data. This information was used to assist statistical analyses on data. These results are presented in appendix Data source Three types of data sources were considered: Factory (standing for plant), i.e. site specific; Literature; Other: e.g. from other sites. ECOBILAN L016 MR5 April /156

44 Data type Five types were defined: Measured: the flow value comes from continuous measurement. For instance, the total electricity consumed is readily available from the electricity meters. The coal consumption is continuously measured with weighbridges or other forms of stock accountability. Averaged value: the annual flow is calculated from spot measurements. For instance, VOCs are measured three times a year during one day each time: from these values the annual value is calculated. Calculated: the flow value is calculated with ratio (emission factors), mass balance or other indirect methods. For instance, SO x emissions may have been measured for several years, an emission factor has been determined and used for the subsequent years; CO 2 air emission is calculated from the carbon balance. Estimated: the flow value estimation has been established based on approximations. For instance, the transportation distance of some raw materials may be estimated because of lack of better information. Unknown: This type is only available for data coming from a bibliographic source when the provided information is not sufficient to classify the data in the previous types Time reference for data collection Data collection relates to one-year operation. Questionnaires indicate the reference year specific to each data point. The majority of data derive from the most recent records, mainly 1997 data Data audit A data quality audit was performed after the data collection phase. Ecobilan requested a copy of the original documents that authenticate calculations and data that were entered in the questionnaires. Three items at least per type of process stage (stainless steel making, hot rolling mills, cold rolling mills, surface finishing) were selected on a random basis to ensure that sites could not predict which data would be required for justification. Not all process stage were audited: three process stages were selected per site. Out of 34 process stages audited (against a total of 53 process stages) included in the study, 25 audit files were filled out by the sites. The reason why some sites did not respond to the audit was usually because the person who had filled out the questionnaire was no longer working at the site. ECOBILAN L016 MR5 April /156

45 This data quality audit is another important element for the study credibility. 3.5 Questionnaire check procedures and treatment All returned questionnaires were checked by Ecobilan. Outstanding data and important missing information were detected by a list of established procedures presented below. The procedures consisted in automatic checking presented below. Whenever needed, data were checked by Ecobilan with the site, through adequate questions. For each site, a list of questions were established after reception and analysis of the electronic questionnaires in order to check, complete and correct the questionnaires. List of automatic checking: Identification and counting of missing figure data: all inflow/outflow should be filled either by a figure, a question mark or by "nr 17 ". Identification and counting of missing quality data (source and type): all data had to be characterised by two letters (one for characterising its source, one for characterising its type). Calculation of mass and water balances for each process stage under study. Calculation of iron, carbon, chromium, nickel and molybdenum balances for each process stage under study. Calculation of existing gaps between theoretical calculated CO 2, CO, SOx and NOx data and CO 2, CO, SOx and NOx provided by the sites (when any). For each process stage and for each boiler of the site, theoretical data were calculated for these air emissions considering the fuels inflows implemented in the questionnaire and literature emissions factors for each type of fuels. The theoretical calculated values were then compared to the data implemented by the site. List of manual checking: - Checking that energy, effluent, waste water treatment plant and transportation spreadsheets are filled in. - Calculation of the product yield for each process stage and analysis of its consistency. 17 nr = non relevant ECOBILAN L016 MR5 April /156

46 - Calculation of the efficiency for each steam/compressed air plant of the site and analysis of its consistency. - Checking with the site that there was no double-counting (e.g. checking that the water emission of chromium VI was not double counted by the categories Cr total and Cr 6+ ). - For effluent data, calculation of the concentrations of the substances in water and comparison of the obtained values to the French Regulation Calculation of data (inflows/outflows) per ton of main products for each process stage and comparison of these data with the average of the received data. After checking operations, the questionnaires were electronically downloaded in TEAM TM 19 for LCI calculation. To do this, an interface between the questionnaires and TEAM TM was specifically developed to avoid typing and miscalculation (e.g. unit conversion) errors. After the LCI calculation, extreme data revealed by the statistical results were checked. Finally, an energy balance was carried out (see section 6). 4. Calculation procedures used for site inventories Site inventory data as collected in the questionnaires were incorporated into an inventory model that takes account of all inter-modular process flows and system interactions in order to carry out the LCI calculation. For discrete analysis of multiple product systems, criteria are developed to allocate the resulting inventories to each stainless steel product. The criteria for allocation are a complex and heavily debated topic since it can have significant impacts on LCI results. Within any site, the stainless steel industry generates different stainless steel products (grades and surface finishing of the product) and by-products for which allocation rules have to be devised. The following sections describe the methodology applied within this study. 18 The French regulation used as a reference is the following: Arrêté du 02/02/98, JO du 03/0398 relative to all type of emissions from classified Installations for Environment Protection. 19 TEAM: Tool of Environmental Analysis and Management ECOBILAN L016 MR5 April /156

47 4.1 Allocation procedure applied for by-products (inputs and outputs) General principles ISO recommendation In clause 6.4.2: allocation procedure of the ISO 14041, ISO recommends the following stepwise procedure to be applied when real multi-functional systems have to be modelled into one or several mono-functional systems. First step: avoid allocation wherever possible, Wherever possible, allocation should be avoided by: dividing the unit process to be allocated into two or more sub-processes and collecting the input and output data related to these sub-processes, expanding the product system to include the additional functions related to the co-products taking into account the requirements of sub-clause (sub-clause deals with the system boundaries definition) Detailing the process is conceivable when the routes of the products are made up of separate process units. This approach can not be applied when one process unit is common to various products, such as stainless steel slabs and EAF/converter slags which are intimately generated within one single process unit: the EAF/converter. Second step: physical partition Where allocation cannot be avoided, the system inputs and outputs should be partitioned between its different products or functions in a way which reflects the underlying physical relationships between them.... Third step: economical partition (based on economic values) This method is avoided since economic parameters, particularly for a global scale study, can fluctuate (prices varying geographically and with time) Methodology choice in the Eurofer study for the Reference inventory calculation It is important to have in mind that it is not only a downstream problem (how to take into account the downstream impact of the co-product of this system), but also an upstream problem (how to ECOBILAN L016 MR5 April /156

48 take into account the upstream impact of the co-product from another system). Whatever be the methodology chosen for the downstream, the same has to be applied to the upstream, in order to be consistent, and to all the recovered matters. For all material co-products, the "scrap pile" approach was chosen. Ecobilan recommends this procedure in the scope of this project for two reasons: 1) to increase simplicity and transparency of study results, and 2) to assure that all life cycle burdens are properly accounted for. No upstream impact was taken into account when a co-product from a system was used by the product system, and no downstream impact was taken into account when a co-product was produced by the product system. This method boils down to applying the partitioning method with the following rule: the recovered matter does not take any of the burdens associated with the manufacturing stages (i.e. all of the manufacturing burdens are attributed to products and not, even partially, to the subsequent uses of the co-product). Recycled, or recovered material from another life cycle that are used are treated as free of any upstream production burdens except for regeneration of coproducts and transport of regenerated product to the site. This partitioning is illustrated in the Figure 7. The regeneration process as well as the raw material saving is attributed to the system where the co-product is recycled. ECOBILAN L016 MR5 April /156

49 Raw material saving recycling Chosen repartition Figure 7 Allocation of the recycling stage and of the raw material saving to the tube life cycle Note Recovered matter is not treated as a waste and does not appear as such in the inventory. The use of recycled materials does not carry any burden from their original production. The recovered matters used as raw materials in the Eurofer study are in fact the stainless and non stainless steel scrap. The regeneration process scrap is the scrap preparation (sorting, shredding, cleaning etc.). The environmental impacts at the scrap merchant's or in municipal facilities are not taken into account in the study. The scrap transportation from these facilities or from scrap producing sites (process scrap) is taken into account. Justification in the case of stainless and non stainless steel scrap It should be mentioned that stainless steel scrap has a high commercial value and consequently its recycling rate is high throughout the EU. The market for the stainless steel production is currently growing slowly. This situation is expected to continue for the next five years, at least. The use of stainless steel scrap and other steel scrap is widespread for the EAF process. Every furnace studied in this project uses scrap. ECOBILAN L016 MR5 April /156

50 The fact that the market for stainless steel is growing means that all available stainless steel scrap is used. And the fact that the growth is slow means that the amount of scrap used will not change very quickly. In this situation, it is justified to use the pile scrap approach as the reference case, because the scrap recycling into EAF is the current situation, and the scrap can not be said to replace a given virgin material. For all energy co-products, the "expansion of the system boundaries" method is applied for the calculation of the Reference inventory. 5 sites produce energy as steam or hot water through process stage 1, 2 and 7, this energy being effectively used by other systems. Two sites use recovered carbon monoxide (blast furnace gas, EAF off gas) as fuel. Two sites use steam from process stages not included in the study. It is assumed in the Reference scenario that the production of energy by the system under study saves the production of energy using conventional fuels. In order to be consistent, the impacts of conventional fuel production are added to the inventory if recovered energy from other systems is used. As shown in Appendix 8, the production of 1 MJ of hot water and steam saves the production and combustion of 1 MJ of natural gas. In the database however, for future use by Eurofer, the choice of other fuel is given. As far as the upstream burdens of CO gas are concerned, the impacts of natural gas production are added as upstream burdens (see Appendix 12) Sensitivity analysis to these choices Three sensitivity analysis were carried out: Applying the expansion of system boundaries instead of the pile scrap approach to material by-products other than scrap Applying the expansion of system boundaries instead of the pile scrap approach to stainless steel scrap Applying the partitioning method to the energy by-products ECOBILAN L016 MR5 April /156

51 Application of the expansion of the system boundaries to stainless and non stainless steel scrap Stainless steel scrap Using sites information: if the quantity of stainless steel scrap decreases, the use of ferro-alloys and pure metals increases. As the content in Cr, Ni and Mo of ferro-alloys and pure metals is higher than that of stainless steel scrap, then carbon steel scrap will also be used to make up the required iron content of the stainless steel. In fact sites prefer to use stainless steel scrap instead of carbon steel scrap and ferro-alloys. For austenitic grades production, any stainless steel grade is used. For ferritic grades, as nickel and molybdenum are not wanted in the composition of the stainless steel, only ferritic scrap can be used. As ferritic scrap tonnage is small, sites use carbon steel scrap and ferro-alloys. For upstream impacts, the use of stainless steel scrap is given the same impact as the use of ferroalloys and carbon steel scrap. For downstream, the same approach has been used: stainless steel scrap outputs use save ferro-alloys and carbon steel scrap. The average Cr, Ni and Mo contents of the stainless steel scrap used by the site were calculated. 7 out of 9 sites having an EAF gave the Cr, Ni and Mo contents of their scrap and the average was calculated giving each site the same weight. Table 9 presents these different contents. Cr content Ni content Mo content kg/tonne kg/tonne kg/tonne Stainless Steel Scrap 304 (from external supply) Stainless Steel Scrap 316 (from external supply) Stainless Steel Scrap 430 (from external supply) Stainless Steel Scrap 409 (from external supply) Table 9: Cr, Ni and Mo contents of stainless steel scrap from external supply entered in the EAF 1 tonne of stainless steel scrap 304 contains 166 kilograms of chromium. If this tonne of scrap had not been available, the site would have used ferro-chromium to match these 166 kilograms. The ferro-alloys production data used were taken from bibliographical sources and do not correspond to ferro-alloys production but rather to pure metal production in the case of chromium and nickel. ECOBILAN L016 MR5 April /156

52 The production data are given for 1 kg of chromium (respectively nickel and molybdenum) contained in the ferro-chrome. The same functional unit was used for the on-going projects on ferro-alloys production. Once the data are available, they will replace the ferro-alloys bibliographical data used here (see Table 10). As far as the downstream impact of stainless steel scrap is concerned, the same approach was taken. As the stainless scrap produced by the site and recovered externally are the same grade as the main product, the Cr, Ni and Mo content of the scrap produced vary according to the grade produced. 274 Chromium (Cr): Production Laboratorium für Energiesysteme ETH, Zurïch, 1996 Page: 58 Primary source:mori G., persönliche Mitteilung, Institut für Technologie und Hüttenkunde der Nichtseisenmetalle. 274 Molybdenum (Mo): Production Source: IDEMAT (Delft University of Technology, 1995) Primary source: 1) Metal Resources and Energy, Chapman, London ) Meatls and Minerals, section Minerals, edition ) Energiebewust ontwerpen, Kemme Nickel (Ni): Production Laboratorium fur Energiesysteme ETH, Zurïch, 1996 Page: Table 10: Bibliographical data used for ferro-alloys production (used only for the upstream and downstream impacts of stainless steel scrap) Non stainless steel scrap Carbon steel scrap and low alloyed steel scrap are said to be in sufficient quantity in Europe for the years to come. Furthermore, no other sources of iron can be easily used in replacement. In the sensitivity analysis, the pile scrap approach is kept for non stainless steel scrap. Calculations In the project, the set of reference inventories was calculated using the pile scrap approach. Then, a second set of inventories was calculated using expansion of system boundaries (applying what is presented above) in order to make a sensitivity analysis and see what is the relative variation of the LCI to this methodological choice, using the method and values indicated above. ECOBILAN L016 MR5 April /156

53 In order to calculate this second set of inventories, data for metallic raw materials extracted from the bibliography were used. Their quality is average, but it gives a good first indication about the relative importance of this methodology choice. Those data were used to model the upstream of about 500 kg of stainless steel scrap for 1000 kg of 304 2B coil produced, which is around 30% of the metal inputs. The results of the sensitivity to the methodological choice show that this choice has a very high importance on the overall results: the change for CO2, NO X, SO X and primary energy is more than 80%. This figure may be exaggerated as bibliographical data for chromium and nickel ferroalloys correspond in fact to pure metal production, increasing the upstream impacts of stainless steel scrap Application of the expansion of the system boundaries to material by-products other than scrap The principle of system expansion is based on the fact that the by-product saves or avoids another product with equivalent function (e.g. EAF slags save clinkers). The expanded system then includes the recovery processes of the by-product and the route of the product(s) that is replaced by the by-product. For the sensitivity analysis, in this study, the inventory of the operations saved by the stainless steel by-products were subtracted as a credit from the inventory of the stainless steel manufacturing routes. Clearly the utilisation of this type of credit relies upon judgements, e.g. do EAF slags replace exactly clinkers in terms of functionality? A substitution ratio of 0.9 was applied in the EUROFER study to take into account the fact that the hydraulic binding power of the EAF slags is lower than that of clinker (based on the French norm AFNOR P ). In other words, it was consider that 1 kg of EAF slag recovered in cement making saves 0.9 kg of clinkers. Hence the LCI of a mass m of clinkers which is considered as equivalent to a mass m of EAF slag multiplied by 0.9, was subtracted from the stainless steel product LCI. The system expansion method is described further in Appendix 8 and a specific example is illustrated. This method is optional and can be used through a set of variables in the interface delivered to EUROFER. ECOBILAN L016 MR5 April /156

54 Only by-products fulfilling the following criteria were taken into account in the sensitivity analysis: 1. the main solid by-products in terms of tonnage, 2. the by-products for which recycling chain data are available. Their list is given in appendix 8. Taking into account more by-products may quickly lead to an explosion of the system complexity, and the data collection task, whereas the impacts on stainless steel LCI is likely to be small. Their total tonnage is on average small: 0.11 kg/kg of product (for the 304 2B coils). However, the by-products not taken into account in the calculations are listed in the stainless steel product LCI results for any possible future integration of their potential environmental effects, if desired. The consequences of applying system expansion method to the material by-products (slags, used interleaving paper, used refractories ) were studied. The LCI results are very little sensitive to these choices, since the variation is in general less than 5% (this is in particular the case for the fuel energy and total primary energy indicators, the CO2 emissions and other accounted air and water emissions) Application of partitioning to energy by-products For all energy co-products, the "expansion of the system boundaries" method is applied for the calculation of the Reference inventory. For the sensitivity analysis, the partitioning method was applied with the following rule: no upstream impact was taken into account when a co-product from a system was used by the product system, and no downstream impact was taken into account when a co-product was produced by the product system. The consequences of the partitioning method to the steam and hot water recovered on site as well as recovered energy inputs such as carbon monoxide was analysed on the coils 304 2B. It appears that the LCI results are very little sensitive to this choice since the variations are almost always inferior to 5 % (this is in particular the case for the fuel energy and total primary energy indicators, the CO2 emissions and other accounted air and water emissions). 4.2 Allocation of the process stage inventories between several stainless steel products Ideally, each module should yield one stainless steel product. However, this is never the case because: ECOBILAN L016 MR5 April /156

55 1. stainless steel products which are not included in the study are also processed in the considered process stage (e.g. ingots in (1) EAF Stainless Steel Making ). 2. different grades of stainless steel products exit the same process stage (e.g.: ferritic and austenitic slabs in (1) EAF Stainless Steel Making ). The method used to eliminate the flows associated with stainless steel products not included in the study includes: not recording inputs/outputs specific to the product(s) not included in the study, e.g. mineral additives added to the ingots only, recording the inputs/outputs specific to the studied products, without partitioning. partitioning the other inputs/outputs using a partitioning parameter (composition, produced tonnage..). Considering the products selected in the study, different stainless steel grades can go through the same process stage. For instance, austenitic and ferritic slabs go out of the first process stage (1) EAF Stainless Steel Making. Therefore special allocation rules have to be defined in order to achieve LCI results for each grade. for metallic raw consumptions in the first process stage (1) EAF Stainless Steel Making, as the content in chromium, nickel and molybdenum is different for each grade, the metallic raw inputs have been partitioned between these grades, as a function of the composition of the studied grade, based on the data transmitted by the sites for a given grade. For other raw materials (e.g. argon, oxygen,..) and energy consumptions for which consumed quantities vary in function of the studied grade in (1) EAF Stainless Steel Making, allocations rules were applied, based on the data transmitted by the sites for a given grade. All these flows are listed in Table 11. for raw materials and energy consumption for which quantities are independent of the stainless steel grades, a mass partitioning was applied. ECOBILAN L016 MR5 April /156

56 Raw metallic materials Carbon Steel Scrap Stainless Steel Scrap Ferro-Aluminium Ferro-Chromium (high carbon content) Ferro-Chromium (low carbon content) Ferro-Manganese Ferro-Silico-Manganese Ferro-Molybdenum Ferro-Nickel Ferro-Silicium Ferro-Titanium Aluminium (pure) Chromium (pure) Iron (pure) Manganese (pure) Nickel (pure, electrolytic) Nickel (pure, briquette) Molybdenum Oxide (MoO2) Nickel Oxides (NiO) Other materials and energy Electricity Steam Light Fuel Oil Natural Gas CO gas Propane Argon (Ar, gas) Argon (Ar, industrial grade, liquid) Argon (Ar, pure grade, liquid) Nitrogen (N2, liquid) Nitrogen (N2, gas) Oxygen (O2) Table 11: Materials and energy whose consumption may be dependent on the produced grade in process stage Allocation of utility inventories Energy allocation The inventory of the production of steam, hot water, and compressed air on site were systematically calculated for each site. Moreover, electricity was always grid electricity, whatever the site. ECOBILAN L016 MR5 April /156

57 Compressed air Regarding compressed air production, data come either from DEAM TM database or from site input. DEAM TM data have been used for compressed air when: all compressed air is produced with electricity only, not with steam, the compressor yield average is close to kwh/m3 of compressed air (average yield for 5 bar compressed air). In other words, compressed air production inventory data were provided by the sites when the standard compressed air production model was not valid for the given site. Steam and hot water For steam and hot water, in most cases their sources are a site power plant and the process stage heat recovery systems. Steam and hot water flows were recorded in energy units (MJ). As it is impossible to differentiate the source of steam/hot water for each process stage, the breakdown of steam/hot water was an average of both sources and this average breakdown was used for all the workshops located on the site. At the process stage level, this method may introduce a gap between the modelling and the reality because the process stage is only part of the global system. The potential effect is a shift of the environmental burdens from one process stage to another one. At a site level, the reality is fully respected. The site power plants producing steam and hot-water are multi-functional systems. Their inventories have thus to be allocated between the different products. The allocation method applied to these integrated power plants is explained in appendix Effluent allocation The water networks in stainless steel plants are usually rather complex: water treatment plants may combine effluents from many process stages, water (and its related pollution) is recirculated from one process stage to another before going to the water treatment plant. Water flows and water emissions released in the environment and coming from a common waste water treatment plant (called CWWTP in the text) or from a given process stage after recirculation had to be allocated between the different site process stages and utilities to fulfil the modular construction. ECOBILAN L016 MR5 April /156

58 The CWWTP consumption (energy and reagents) and solid waste (sewage sludge, waste oil,...) had also to be allocated between each process stage and utility from which the treated effluents come from. The steps of the allocation procedure applied to the effluent and the common waste water treatment plant were: 1. the site staff allocated the flows according to their information, experience and knowledge whenever possible; 2. Ecobilan allocated the flows that were not allocated by the site staff with methods based on physical relationships. Appendix 11 describes the methods used by Ecobilan to allocate the CWWTP inventories between different process stages. 4.4 On-site transportation Consumptions and emissions from site vehicles were not included in the study. Indeed it is very difficult to accurately estimate all round trips between workshops. However, in order to check that this on site transportation doesn t represent much compared to the final LCI results, the contribution of this transportation step was evaluated. Based on the data collected in the frame of the IISI study and considering that the transportation step on stainless steel workshops can be approximated by the one on steel workshops, the fuel consumption dedicated to the on-site transportation can be estimated to approximately litre / kg of stainless steel product, corresponding to about 0.03 MJ of fuel energy per kg of product. This latter value can be compared to the total fuel energy necessary to produce 1 kg of stainless steel product (LCI result), equal to about 29 MJ. Therefore, excluding on site transportation from the system is acceptable. ECOBILAN L016 MR5 April /156

59 5. Site upstream and downstream inventory The site upstream sub-systems refer to the production and the transportation of the site inputs: raw materials, consumables and energy (see Figure 3). As described below, the site input production inventory data are either customised to each site or standardised for all sites using data from DEAM TM20 depending on their assumed contribution to the LCI. Detailed information concerning all upstream models are given in appendix 12. Quality assessment of these data is included in this appendix. 5.1 Electricity Electricity Model As grid electricity contribution in the stainless steel product LCI is very important (due to the electric arc furnace), the choice of electricity model is essential. Methodologically speaking, two solutions are possible. Either, the model is country specific and reflects the proportions of each electrical route for each country (e. g. 50 % natural gas, 25 % nuclear, 15 % coal and 10 % hydroelectricity). Or the same European average model is used for all the participating countries to avoid hiding the site/company differences by the differences on the electricity model. The Expert Group agreed to implement the European electricity model in the final interface that will be delivered to EUROFER. For the site data delivery, the model of grid electricity production will be customised to each country as illustrated in Figure 8. The part of the energy sources used in the production of grid electricity (% of nuclear, % of hydro, % of thermal energy fuel by fuel) was adapted to European situation and to each country 20 DEAM: Database for Environmental Analysis and Management- Ecobilan's LCI database ECOBILAN L016 MR5 April /156

60 from Energy Statistics of OECD Countries and is representative of the year 1996 (data published in 1997). All models of electricity production per type of primary energy (e.g. coal extraction, transport and combustion in electricity power plant) are DEAM TM models (sources: ETH-Zürich and Electricité de France 1996) and standard for all sites (see appendix 12). These approximations may affect the LCI results knowing that the inventory of electricity production per type of primary energy can be significantly different from one country to another. Electricity production from natural gas (standard for all countries) (including gas extraction, refining, transport and combustion in power plant) Electricity production from coal (standard for all countries) (including coal extraction, transport and combustion in power plant) Electricity production from hydropower (standard for all countries) Electricity Breakdown (Specific to each country) u % of electricity from coal v % of electricity from lignite w % of electricity from heavy fuel oil x % of electricity from natural gas y % of electricity from nuclear z % of hydropower electricity Electricity production from lignite (standard for all countries) (including lignite extraction, transport and combustion in power plant) Electricity production from heavy fuel oil (standard for all countries) (including oil extraction, refining, transport and combustion in power plant) Electricity production from nuclear energy (standard for all countries) (including uranium ore extraction, transport, refining, nuclear power plant) Figure 8: Grid electricity production model The European energy grid for 1996 is presented Table ETH Zürich : Eidgenössische Technische Hoschule, Zürich. The Federal Office for energy is based in Zürich (Switzerland). This is an independent organisation originating from the Ecole Polytechnique of Zürich that publishes numerous reports listing LCI (Life Cycle Inventories). ECOBILAN L016 MR5 April /156

61 European Efficiencies Union 1996 Coal 22.27% 39.30% Lignite 8.8% 32.30% Fuel Oil 8.55% 38.50% Natural Gas 9.15% 33% Nuclear 34.93% 33% Hydro 13.93% 90% Process Gas 0.88% Free electricity 1.35% Distribution losses 6.3% Table 12: Electricity Grid European Union Sensitivity analysis on electricity grid A sensitivity analysis was carried out on the electricity grid. The inventory for the production of 1 tonne of 3042B was calculated with the Italian energy grid and with the French energy grid. These grids represent two extreme situations in Europe: a grid with a high percentage of fossil fuel energy sources and a grid with a high percentage of nuclear energy source. Italy France European Efficiencies Union 1996 Coal 9.92% 5% 22.27% 39.30% Lignite 0.06% 0.4% 8.8% 32.30% Fuel Oil 50.02% 1.1% 8.55% 38.50% Natural Gas 19.42% 0.8% 9.15% 33% Nuclear 0% 77.8% 34.93% 33% Hydro 17.5% 14.2% 13.93% 90% Process Gas 1.43% 0.7% 0.88% Free electricity 1.6% 0.6% 1.35% Distribution losses 7.3% 2% 6.3% Table 13: Electricity grids used for sensitivity analysis ECOBILAN L016 MR5 April /156

62 The results are presented in Table 14. Variation with Italian Grid Variation with French Grid (r) Natural Gas (in ground) 19% -22% (r) Oil (in ground) 135% -26% (r) Coal (in ground) -50% -75% (r) Uranium (U, ore) -95% 104% (r) Lignite (in ground) -98% -99% (a) Carbon Dioxide (CO2, fossil and mineral) 20% -41% (a) Carbon Monoxide (CO) -40% -43% (a) Nitrogen Oxides (NOx as NO2) 16% -38% (a) Particulates (unspecified) 1% -2% (a) Sulphur Oxides (SOx as SO2) 106% -55% (a) Hydrocarbons (except methane) 110% -27% (a) Hydrocarbons (unspecified) -3% -2% (a) Methane (CH4) 102% -63% Waste (total) -15% -22% Waste: Highly Radioactive (class C) -97% 106% Table 14: Results of sensitivity analysis on energy sources for electricity production The main inventory parameters can vary by plus or minus 100% due to electricity grid change. ECOBILAN L016 MR5 April /156

63 5.2 Stainless steel and non stainless steel scrap Sites use different categories of stainless steel scrap in the EAF depending on the grade of stainless steel being produced. In the questionnaire, stainless steel scrap grades were sorted in 5 categories, with no further indication. For the final inventories edition, the stainless steel scrap was divided in four categories: - Stainless steel scrap (304, from external supply), scrap with the approximate carbon, nickel, molybdenum and chromium content corresponding to the 304 grade; - Stainless steel scrap (316, from external supply), scrap with the approximate carbon, nickel, molybdenum and chromium content corresponding to the 316 grade; - Stainless steel scrap (409, from external supply), scrap with the approximate carbon and chromium content corresponding to the 409 grade; - Stainless steel scrap (430, from external supply), scrap with the approximate carbon and chromium content corresponding to the 430 grade; Non stainless steel scrap are divided in two categories: - Alloy steel scrap (low alloyed, from external supply), scrap with small amount of nickel, chromium and molybdenum; - carbon steel scrap; In the reference inventory calculation, the scrap pile approach was used as seen in section 4.1. No upstream impact is taken into account when scrap is used by the product system, and no downstream impact is taken into account when scrap is produced by the product system. A sensitivity analysis was performed on this methodological choice (see section ). ECOBILAN L016 MR5 April /156

64 5.3 Ferro-alloys and pure metals (chromium, nickel and molybdenum) As already stated in section ( system boundaries ), the collection of data concerning chromium, nickel and molybdenum ores extraction and their transformation into pure metals or ferro-alloys was not included in the scope of the present study. Therefore the flows corresponding to the ferro-alloys and the pure metals inputs (e.g. pure nickel) appear in the final delivered inventories in order to enable the future integration of those data to get the global cradle to gate LCI. The integration of these upstream data in the present stainless steel project will be considered in a separate project, once the respective projects on chromium, nickel and molybdenum will be finished. This complementary work will result in the edition of the final cradle to gate results for the selected coiled flat stainless steel products. 5.4 Industrial gases (O 2, N 2, Argon) The data used for these industrial gases production were collected specifically for this study (argon gas and liquid argon) or come from literature sources (nitrogen and oxygen). The sources and quality of these data are detailed in appendix Other raw materials production For all other raw materials and consumables, standard production models were used for all sites. Not all inputs of their production were taken into account. Omissions are described below Cut-off criteria for raw material production All the inputs that correspond to energy and water were always taken into account. Regarding other inputs (for which mass is the relevant unit), a single and relevant list of raw materials for which upstream production is taken into account, was established for all sites (see ECOBILAN L016 MR5 April /156

65 Appendix 12). This list was assumed to cover at least 95 % of the stainless steel workshops total inputs, main product 22, fuels and water excluded. Appendix 9 presents the selected cut-off rules applied for raw material production for each process stage Production models There are several sources of raw material production models, as shown in appendix 12. The major one, DEAM TM23, is a series of data sets compiled by Ecobilan covering most industrial fields. Each data set contains data for either a process step or a set of steps, comprising all useful information. Data derive mainly from existing industrial processes. During the data generation, highest attention is paid with respect to present and future transparency standards. Data are checked in three ways: order of magnitude standards, mass and/or element conservation and cross-referencing. Other source models were used for the few site inputs which are not covered by DEAM TM database, namely: refractories (alumina based, magnesite based, dolomite based, chromium magnesite based); metallic addition : ferro-silicium, ferro-manganese, silico-manganese; interleaving paper (paper which is used in the last process stages to protect the surface of the stainless steel coils); steel rolls used in the cold rolling stage (5) and (2); shot blasting products. 5.6 Site external transportation DEAM TM transportation models were used to calculate the transportation inventories based on distance and means indication (trucks, rail, sea, etc.) recorded in the questionnaires. 22 In process stage (1) EAF Stainless Steel Making, the main input corresponds to stainless steel scrap and carbon steel scrap, whereas in the other process stages, the main input corresponds to the stainless steel product (e.g. slab for process stage (2) Hot Rolling Mills) 23 DEAM: Data for Environment Analyses and Management ECOBILAN L016 MR5 April /156

66 5.6.1 Transportation models The transportation used for conveying raw materials and semi-finished stainless steel products are: rail (electricity and diesel engine distinguished), road, shipping by barge, shipping by freighter. The models of rail and barge transportation come from ETH-Zurich 24 (1996). The model of freighter transport and road transport use BUWAL 25 data. The functional unit of all transport models except road transport is the transport of 1 kg of material over a distance of 1 km. The transport distance per means was asked in the questionnaires. For road, the functional unit is 1 litre of diesel oil consumed in a truck engine and includes water, lubricant and tire consumption. It is usually assumed that trucks consume in average 38 litres of diesel oil per 100 km with full charge. The consumption of diesel oil for 1 kg of material with no empty return is estimated as follows (calculation carried out by the Questionnaires/TEAM TM interface). Diesel oil quantity in litre / kg transported = x truck capacity x1000 x distance Both truck capacity and transportation distance were asked in the questionnaires. When the trucks return empty (information was also asked in the questionnaire), the environmental burden of the truck return is taken into account. The empty truck consumption is estimated as 25.3 litre per 100 km (2/3 of the full charge truck consumption). 24 ETH-Z: Eidgenössische Technische Hochschule von Zürich (Federal Technical Institute of Zürich). 25 BUWAL : Bundesamt für Umwelt, Wald und Landschaft : Swiss Federal Office of Environment, Forests and Landscape. ECOBILAN L016 MR5 April /156

67 5.6.2 Semi-finished product transportation Transportation of semi-finished products from one site to another, i.e. when several sites are involved in the manufacturing route, was taken into account Raw materials and consumables transportation The Expert Group decided to take into account the transportation for the raw materials and consumables listed in Table 15. These materials were selected in order to take into account, for each process stage, the material for which the quantity weight x distance is not negligible (i.e. representing more than 5 %) compared to: - the quantity weight x distance for scrap in process stage (1) EAF Stainless Steel Making, - the quantity weight x distance for the semi-finished product in the other process stages. The site staff was then asked to record transportation information for this list of materials. ECOBILAN L016 MR5 April /156

68 Carbon Steel Scrap Sodium Sulphate (Na 2 SO 4 ) Stainless Steel Scrap Sodium Hydroxide (NaOH) Ferro-Aluminium Lime (slaked, Ca(OH) 2 ) Ferro-Chromium (high carbon content) Ammonium (NH 3 ) Ferro-Chromium (low carbon content) Sodium Metasulphite (Na 2 S 2 O 5 ) Ferro-Manganese Sulphuric Acid (H 2 SO 4 ) Ferro-Silico-Manganese Hydrochloric Acid (HCl) Ferro-Molybdenum Hydrofluorhydric Acid (HF) Ferro-Nickel Nitric Acid (HNO 3 ) Ferro-Silicium Argon (gas) Ferro-Titanium Argon (liquid, industrial grade) Lime (quick, CaO) Argon (liquid, pure grade) Lime (slaked, Ca(OH) 2 ) Limestone (CaCO 3 ) Fluorspar (CaF 2 ) Coal EAF Electrodes Ladle Electrodes Nickel (pure, electrolytic) Nickel (pure, briquette) Molybdenum Oxide (MoO 2 ) Interleaving Paper Table 15: List of raw materials and consumables for which it was asked to fill in the distances of transportation Transportation of other raw materials than those listed above was not taken into account. 5.7 Site downstream processes Oily waste incineration The stainless steel workshops generate oily waste. The oily waste may be regenerated or incinerated on site or outside the site (with or without heat recovery). On the sites covered by this study, oily waste, if incinerated, was always incinerated outside the site. The impacts of this incineration are taken into account (see Appendix 8). If the incinerator recovers energy, the production of the equivalent amount of energy is subtracted from the inventory, if the expansion of the system boundaries is chosen for energy by-products (as it is the case for the Reference inventory). ECOBILAN L016 MR5 April /156

69 Oily waste recycling process considered in this study is a distillation process using clay Dust treatment Data on dust treatment were gathered at four dust treatment plants (cf. Appendix 12). Appendix 8 describes the becoming of the outputs of these plants and the methodological choices that were chosen. ECOBILAN L016 MR5 April /156

70 6. Energy indicators 6.1 Definition and use Besides the physical inputs/outputs LCI calculation, energy indicators were calculated. These energy indicators, detailed below, measure the energy performance of the studied systems. Because the energy indicators are not physical inputs/outputs, they can be considered as outside the scope of an LCI, therefore outside the scope of the EUROFER study. ISO has not yet clearly defined how and when the energy indicators shall be calculated. Within the study, the energy indicators aims at: 1. giving a first approximation of the total primary energy used for stainless steel manufacturing, 2. making a rough energy balance: what part of the energy drawn from the environment is spent in the system, for data checking purposes (see section 6.2.2). The energy indicators that were calculated are the following: total primary energy, non renewable energy, renewable energy, fuel energy and feedstock energy. These indicators are defined in the following paragraphs. Total Primary Energy: sum of all the energy sources which are directly drawn from the earth such as natural gas, oil, coal, uranium ore, biomass, or hydropower energy. The total primary energy is split into non-renewable and renewable energy on one hand, into fuel energy and feedstock energy on the other hand. Therefore, the following relationships link these indicators: Total Primary Energy = Non-renewable energy + Renewable energy = Fuel energy: + Feedstock energy ECOBILAN L016 MR5 April /156

71 Non-renewable energy: includes all fossil and mineral primary energy sources, for example, oil, natural gas, coal and nuclear energy. Renewable energy: includes all other primary energy sources: mostly hydropower and biomass. Fuel energy: corresponds to the part of primary energy entering the system which is consumed by the processes in the studied system, Feedstock energy: corresponds to the part of primary energy entering the system which is not used as fuel energy, that is generally the calorific energy of the outputs (the products, the byproducts, the waste) and the fuel losses. The term fuel energy covers all the energy that is spent for the process purposes, either to produce heat, mechanical energy or to enable endothermic chemical reaction. In fact, the term fuel energy, commonly used by LCI practitioners, is confusing. Fuel energy for LCI practitioners is a synonym of net energy requirement or process energy. ISO standard gives a definition only for feedstock and process energy which are different from those given above. Feedstock energy is defined as combustion heat of raw material inputs, which are not used as an energy source, to a product system, expressed in terms of higher heating value or lower heating value. Process energy is defined as energy input required for a unit process to operate the process or equipment within the process excluding production and delivery energy. A major disadvantage of these definitions is that they do not refer to primary energy, and consequently they do not reflect the principle of energy conservation along the product life cycle. In other words the sum of process energy and feedstock energy is not constant along the product life cycle and equal to the primary energy. 6.2 Calculation The energy indicators are calculated for each module and then aggregated in the same way as in/outputs. ECOBILAN L016 MR5 April /156

72 6.2.1 Primary energy Primary energy (renewable and non-renewable) is the energy drawn from the environment, therefore the calculation of primary energy indicators applies to the raw material extraction steps only. At the steelworks level, these indicators are nil since raw materials are not extracted from the earth on site. The definition of the energy drawn from the environment is subject to discussion: for example shall the free enthalpy of the natural resources be accounted for? Within the study, the primary energy calculation is based on the energy used by the industrial systems. net caloric values (NCV) for fossil fuels and biomass materials, gravitational energy for hydropower: (1.11 MJ of gravity energy yields 1 MJ of electricity, BUWAL 26 data), burn-up rate for uranium ore: ( g of enriched uranium yields 1 MJ of electricity, BUWAL data). The other natural resources of energy (wind, sun, etc.) are negligible for the study Fuel and feedstock The fuel energy balance is calculated for each process unit, as the difference of primary energy between inputs and outputs. For the stainless steel workshops process units, only the net calorific values of the materials are considered in the primary energy. Therefore, for a stainless steel workshop process unit, the fuel energy balance is calculated according to the formula: Fuel energy mass (kg) x NCV (MJ / kg) - mass (kg) x NCV (MJ / kg) = inputs With NCV: Net Calorific Value According to the principle of energy conservation: outputs Feedstock Energy = Total Primary Energy - Fuel Energy 26 BUWAL: Bundesamt für Umwelt, Wald und Landschaft (Swiss Federal Office of Environment, Forest and Landscape). ECOBILAN L016 MR5 April /156

73 For a stainless steel workshop process unit, the primary energy is nil. Therefore, for a these process units, the feedstock energy balance is calculated according to the formula: Feedstock Energy - Fuel Energy = mass (kg) x NCV (MJ / kg) - mass (kg) x NCV (MJ / kg) = outputs inputs All energetic material with significant heat values: namely natural gas, light fuel oil, heavy fuel oil, liquefied petroleum gas, CO gas were directly recorded in MJ in the electronic questionnaires. By definition, electricity is not taken into account in the fuel energy indicators of a process stage since electricity is a secondary energy (i.e. produced by industrial systems from primary energy). The primary energy spent (i.e. the fuel energy) to produce electricity is accounted for at the level of the process unit that produces that electricity, namely the grid electricity production model. The same remark applies to all other secondary energy flows, namely: steam, hot water and compressed air. Fuel losses, not recorded in the questionnaires and waste calorific energy are assumed to be negligible and are not taken into account. For the stainless steel product manufacturing systems, the feedstock is small as compared to the primary energy since: the calorific energy of the carbon content of stainless steel is small and therefore not taken into account in the calculations, all calorific by-products are taken into account with the system expansion method. The feedstock result was used as a checking tool: high absolute values of feedstock (higher than 2 MJ/kg of product) highlight mistakes of data collection or data treatment Remarks concerning the use of net caloric value (NCV) NCV was used throughout the EUROFER Study. NCV or Low Heat Value is a measure of the energy recoverable by a system where combustion products are not redeemed to ambient temperature and pressure. Hence water vapour is not condensed. GCV (Gross calorific Value) or High Heat Value is the total quantity of energy that is released by the combustion of unit quantity of fuel, as in a bomb calorimeter. ECOBILAN L016 MR5 April /156

74 The difference between GCV and NCV is the latent heat of vaporisation of the water contained in the exhaust gases plus the temperature difference between ambient and vapour. Whilst the energy indicators must be regarded as separate definitions for clarity and reporting only NVC is of practical use to industry owing to thermodynamic limitations the same way when given in NCV or in GCV. In fact GCV is a physical unit which generally does not reflect the industrial reality but an ideal case. NCV is an industrial unit that is adapted to quantify the energy balance of the industrial systems, but on the counterpart omits potential improvements which could be achieved with sophisticated technologies. Ecobilan generally uses NCV that is closer to the industrial realities. For instance if a product containing a lot of water is considered, its energy feedstock expressed in GCV will have no sense because most of this energy cannot be usually recovered. More generally, NCV value is also more pertinent to compare the energy efficiencies of two industrial systems. ECOBILAN L016 MR5 April /156

75 7. Conclusion An external independent Critical Review Panel (CRP) has reviewed the study. The panel reviewed and gave advice on the methodology throughout the project. The final conclusion of the Critical Review Panel as well as Ecobilan's responses to these conclusions are added to this report in Appendix 15. Apart from this methodology report, major deliverables of this study are the database where the production of flat coiled stainless steel is modelled and the life cycle inventory results for flat coiled stainless steel products and processes. Relevant database features The completeness and reliability of the database must be underlined. All participating European stainless steel producers defined the process flow charts in order to ensure the consistency of data collection. It is estimated that 99.9% w/w of steelworks material inputs and 100% of energy inputs have been recorded. The upstream production of these inputs has been taken into account for more than 95% w/w of these inputs excluding the intermediate product such as slab or white hot rolled coil. The data were implemented in an Excel questionnaire by the sites. Ecobilan s engineers checked the result of the implementation by automatic and "manual" procedures. Thereafter, the data contained in the Excel questionnaire were downloaded automatically to avoid typing mistakes, so the data in the final database are exactly what the sites implemented. A data audit was performed at the end of data collection: sites were to send original documents for flows randomly selected by Ecobilan. This audit is an important element for the credibility of the study. The results of the audit were checked manually. Another important feature of the database is its transparency. The data collected by the sites are directly available in the different modules of the database as well as information on each flow. Upstream and downstream data have been documented as much as possible, depending on available information. ECOBILAN L016 MR5 April /156

76 How to use the results: Life cycle inventory results have been calculated for 18 stainless steel products and for each individual process stage. Each stainless steel producer was given its own inventory results as well as the average, minimum and maximum values of the other participating companies. The results can be used by the stainless steel producers to perform an environmental benchmarking of their process stages as well as their products. Process stage benchmarking can lead to process improvement. For instance, reduction of electricity consumption in the first process stage can lead to major environmental and cost improvement. Product benchmarking can pinpoint the process stage contributing the most to the environmental impact and show where significant reduction of impact can be made. The sites must bear in mind, however, that some flows have variations over 100% from one site to another (alloying element air and water emissions, for instance as seen in Appendix 2) so it would not be appropriate to draw conclusions that would lead to investment decisions based on these flows. For Eurofer, the database represents an environmental picture of stainless steel flat coiled products in 1997, since data were gathered for These data can be used to represent the production of such stainless steel products for 5 to 10 years. Data can be easily updated by the association and show improvements over time. As the methodology and tools are now defined, companies that did not participate in this project can easily join and provide data. Data publication: Eurofer can use the database and the user-friendly interface to obtain life cycle inventory results representative of the production of flat coiled stainless steel products. Increasingly, government agencies and local communities are encouraged to buy products with minimum impacts on the environment. This practice, called green purchasing, is often based on Life Cycle Assessment results. The stainless steel industry must be certain that accurate life cycle inventory data are used to evaluate the environmental impact of products containing stainless steel. Life cycle assessment is also a method used in eco-labelling programmes. Environmental statements and communication brochures can be published based on this study. The following points should be acknowledged before publication: Data publication should be done only when the inventory results for the production of ferrochrome, ferro-nickel and ferro-molybdenum are integrated in the database. ECOBILAN L016 MR5 April /156

77 The inventory results should be calculated based on the production of each participating company in order to represent flat coiled stainless steel products produced in Europe. This can be done using the interface and the variables implemented in the database. The results are very sensitive to the chosen electricity model as well as the allocation rules defined for the stainless steel scrap. Thanks to the interface, the results can be calculated for a different set of options regarding these methodological choices. Care must be taken to clearly mention the choices made for the data publication. Further work This study is a starting point for Eurofer and the participating companies. Eurofer as well as participating companies can perform impact assessment on the inventory results to evaluate the potential impacts of their products and processes on the environment. If needed, Eurofer and participating companies can focus their efforts on selected impact categories such as greenhouse gas effect. The study can help participating companies to define an environmental strategy to reduce their impacts and improve their processes. The results show where to concentrate their efforts to get the highest Results/Effort ratio. Companies can then check the results of the application of such a strategy by benchmarking inventory results between years. Eurofer can initiate, with the help of the companies, an update of the database. Once the database is updated, Eurofer can send the new averages to participating companies as well as communicate on the new regional averages. Eurofer can also add new products and new process stages such as long products processing by using the same methodology. The study can also be part of an Environmental Management System (e.g. ISO series) for individual companies. The database and the TEAM TM software can store and centralise environmental data and provide figures and graphs for environmental reports. Both Eurofer and individual companies can use the inventory results as a marketing tool for stainless steel in new applications. Displaying the results to potential customers helps answering the customer's environment related questions while asserting the stainless steel industry's environmental stewardship. Presentation of the study results can be made to current and potential customers. Brochures can be distributed to the public and to specialised magazines. Taking action, as an industry, to record and in the future possibly reduce its environmental burden, demonstrates the stainless steel industry's commitment towards the environment. ECOBILAN L016 MR5 April /156

78 APPENDIX 1: Project deliverables Time Deliverable Form Completion of Phase I 1) Draft Methodology Report for stainless steel production: During Phase II methodology rules, route flowchart, process flowchart for each process stage, including all in/output flows. 1) Questionnaire distribution 1) Written report 2) Word and PowerPoint files 1) Excel 97 or 7.0 Questionnaire 2) Training to support the collection of data by stainless steel companies 2) Technical training session 3) Written guidelines to accompany data collection questionnaire Completion of Phase II 3) Spreadsheets distribution 4) Excel 97 or 7.0 files ( 3½ disk or ) Completion of Phase III (end of the project) 1) Final Methodology Report (project methodology, results and conclusions) 1) Written final report 2) Software tool TEAM composed of the model of stainless steel production, the European LCI database and a user interface to carry out simulations agreed by Ecobilan and EUROFER 3) User guide and reference to support the use of the software tool by EUROFER and individual stainless steel companies 4) Training to support the use of the software tool 2) Copy of final report on computer diskette 3) Software tool for EUROFER and option for companies to license the software 4) Written user guide and reference for the software tool Each Working Group meeting (four meetings) 5) Spreadsheets for each company covering its covered products and plants 1) Meeting report 1) Written report 5) Spreadsheets in both written and software form (Excel 97 or 7.0 files) Table 16: Deliverables ECOBILAN L016 MR5 April /156

79 Figure 9: Project program ECOBILAN - L016 MR5 April /156