Optimal Recycling of Steel Scrap and Alloying Elements (PDF)

This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any...

This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes. Policy Analysis pubs.acs.org/est Optimal Recycling of Steel Scrap and Alloying Elements: Input- Output based Linear Programming Method with Its Application to End-of-Life Vehicles in Japan Hajime Ohno,*,† Kazuyo Matsubae,‡ Kenichi Nakajima,§ Yasushi Kondo,∥ Shinichiro Nakamura,∥ Yasuhiro Fukushima,† and Tetsuya Nagasaka⊥ † Department of Chemical Engineering, Graduate School of Engineering, Tohoku University, 6-6-07 Aramaki Aza Aoba, Aoba-ku, Sendai, Miyagi 980-8579, Japan ‡ Department of Environmental Study for Advanced Society, Graduate School of Environmental Studies, Tohoku University, 468-1 See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. Aramaki Aza Aoba, Aoba-ku, Sendai Miyagi 980-0845, Japan § Center for Material Cycles and Waste Management, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305-8506, Japan ∥ Faculty of Political Science and Economics, Waseda University, 1-6-1 Nishi-waseda, Shinjuku-ku, Tokyo 169-8050, Japan Downloaded via 190.97.251.72 on October 5, 2024 at 01:36:01 (UTC). ⊥ Department of Metallurgy, Graduate School of Engineering, Tohoku University, 6-6-02 Aramaki Aza Aoba, Aoba-ku, Sendai, Miyagi 980-8579, Japan * S Supporting Information ABSTRACT: Importance of end-of-life vehicles (ELVs) as an urban mine is expected to grow, as more people in developing countries are experiencing increased standards of living, while the automobiles are increasingly made using high-quality materials to meet stricter environmental and safety require- ments. While most materials in ELVs, particularly steel, have been recycled at high rates, quality issues have not been adequately addressed due to the complex use of automobile materials, leading to considerable losses of valuable alloying elements. This study highlights the maximal potential of quality-oriented recycling of ELV steel, by exploring the utilization methods of scrap, sorted by parts, to produce electric-arc-furnace-based crude alloy steel with minimal losses of alloying elements. Using linear programming on the case of Japanese economy in 2005, we found that adoption of parts-based scrap sorting could result in the recovery of around 94−98% of the alloying elements occurring in parts scrap (manganese, chromium, nickel, and molybdenum), which may replace 10% of the virgin sources in electric arc furnace-based crude alloy steel production. 1. INTRODUCTION (i.e., alloyed) combined forms.7 Not only alloyed metals, but Urban mining is a key concept for the sustainable use of also mechanically combined metals become practically metals.1 However, the recovery and/or recycling of metals from inseparable during the remelting stage in the recycling process, end-of-life (EoL) products (i.e., urban mines) has become unless they are separated in advance.5,9 Consequently, closed- increasingly complicated due to the complex structure and loop recycling of metals maintaining required quality tends to composition of recent functional products.2 Potential issues be infeasible. Currently open-loop recycling is the mainstream such as unintentional alloying, contamination, and dissipation recycling strategy.10,11 may occur, induced by this complication in some metals with In open-loop metal recycling the way in which scrap is sorted high recycling ratios3 as those metals are typically not used and allocated to the production of secondary materials independently, but are combined and/or alloyed.4−6 The determine recycling ratio of the metal.2,6,12,13 The importance efficient recovery of metals from EoL products requires of considering subprimary metal contents in scrap allocation to analyzing how metals are contained in EoL products and determining the appropriate recycling methods for them.2,7 Received: August 31, 2017 Closed-loop recycling would be an ideal way of recycling Revised: October 20, 2017 metals sustainably.8 However, it is seldom applied, because Accepted: October 24, 2017 metals are mostly utilized in mechanically and/or physically Published: November 7, 2017 © 2017 American Chemical Society 13086 DOI: 10.1021/acs.est.7b04477 Environ. Sci. Technol. 2017, 51, 13086−13094 Environmental Science & Technology Policy Analysis secondary material productions was emphasized in studies 2. MATERIALS AND METHODS regarding aluminum scrap recycling, since the reduction of 2.1. Contributions of Linear Programing to Waste impurity accumulations in recycled aluminum is an important Recycling. In waste recycling, various types of waste are subject.13−16 On the other hand, studies focusing on treated as secondary resources. Additionally, various technol- subprimary metal contents in steel recycling are rare, except ogies are available for waste recovery and use. Consequently, for those concerned with contamination by copper,4,17 there are many possible combinations of waste, recovery, and presumably because of the small share of steel materials recycling technologies. For sustainable resource management, it containing alloying elements (AEs) compared with mass- is important to choose an optimal combination of the available produced carbon steel.18,19 The recent increase in automobile alternatives that maximizes recycling benefits and minimizes production20 and its increasing use of alloy steel21 made losses. LP has been used as a practical approach to identify a sustainable management of both primary (iron) and subprimary system that minimizes the environmental impacts of various (AEs) metals in scrap,6,12,22,23 important. In fact, Daigo et al.24 production and recycling systems, such as plastic materials found that the dissipation of AEs into carbon steel is causing a production,30 energy supply,31 oil refining,32 and solid waste gradual increase in the concentration of AEs (regarded as recycling.33 Those studies use linearized process inventories for contaminants) in steel scrap,25 emphasizing the need for quality formulation of a linear program. In addition, LP has also been in recycling steel for sustainable use. applied with input-output table to various areas in industrial In our previous studies,6,12,22 the masses of the AEs ecology.34−38 Kondo and Nakamura developed the waste input- contained in automobiles are estimated by waste input-output output-LP model (WIO-LP),39 which can be used to optimize material flow analysis (WIO-MFA)26,27 with confirmations of the selection of waste treatment and recycling technology, consistency with chemically analyzed composition28 and along with goods and service-production technologies, under discussions on the uncertainty included in the results.12 the objective of environmental load minimization based on the Furthermore, the potential benefits of scrap sorting before IO model extended for waste (WIO).40 Lin applied WIO-LP to shredding and melting are estimated by a scenario-based wastewater treatment to optimize technological choices for approach comparing the AE contents in ELV-derived steel minimizing environmental loads.41 Studies based on LP are also scrap (ES) sorted by parts (hereafter, called “parts scrap” to available for ELV recycling.42,43 However, there are few LP- distinguish it from ES in generic terms) with industrial based studies focusing on the quality of scrap recycling in terms standards of steel grades.6,12,22 In addition, the fate of AEs of AEs. An exception is Løvik et al., who applied LP to optimize embodied in each of products over multiple life cycles of ELV-derived aluminum scrap recycling to reduce scrap surplus, products are tracked by MaTrace-alloy model by giving scrap considering AE concentrations in sorted aluminum scrap.13 allocation scenarios to alternative refining processes exoge- Løvik et al.’s approach is similar to the present study in its nously.29 However, the exploration of allocation scenarios may consideration of quality aspects. However, using the IO have been insufficient in scope and in variations. To discuss approach in this study enables the evaluation of the influence how effective practical scenarios of interests are, and to of optimized recycling on the whole national economy, whereas systematically synthesize near-ideal allocation scenarios, max- Løvik et al.’s study only focused on aluminum used in the imal potential benefits needs to be addressed.22 automobile industry. In this paper, we reveal the optimal allocation of parts scrap Gaustad et al.15 identified three scrap allocation models: to the production of EAF steel, and evaluate potential benefits pooled scrap allocation, pseudoclosed-loop scrap allocation, of parts scrap utilization as a secondary source of AEs on the and market-based scrap allocation. In the pooled scrap allocation, scrap is collected and allocated without any reduction of the embodied greenhouse gas (GHG) emissions distinction for either origin or usage. In the pseudoclosed- and costs. Building upon previous works, we use input-output loop allocation, scrap derived from an EoL product is directly based linear programing (LP) to identify an optimal way of allocated to produce the same product again. In reality, using secondary AE sources, that is, parts scraps having specific however, scrap is purchased by secondary material producers AE compositions. The secondary and virgin AE sources are first, and secondary material producers then decide the usage of used together to produce electric-arc-furnace (EAF) steels that scrap for secondary material production (i.e., market-based are subject to different AE requirements. Two alternative allocation).15 Since mass-produced metals, such as aluminum objective functions are considered: (1) minimizing greenhouse and steel, are widely utilized in various industries, market-based gas (GHG) emissions embodied in virgin AE sources, and (2) allocation can simulate the behavior of secondary producers minimizing the cost of purchasing virgin AE sources in EAF that have several options regarding scrap use and secondary steelmaking. In addition to the allocation of secondary AE material production to meet the demand from various sources, the uses of virgin AE sources were also optimized industries. This study follows the market-based allocation based on the objective functions. Constraints for the linear approach by applying the IO database, developed for the WIO- program are given by the economy-wide supply demand MFA (hereafter, called WIO-MFA table),6,12,22 based on the balance obtained from the IO database developed for Japanese IO table for 200544 as a reference to AE demand in simultaneous AE flow analysis with WIO-MFA.6,12 This enables EAF steelmaking and demand for EAF steel materials. us to understand the influences of optimized parts scrap 2.2. Linear Program. The equality constraints of our utilization on both steelmaking-related and other sectors. optimization problem are classified into three groups. The first In addition, we explore strategies for reducing the embodied group represents the supply demand balance regarding goods GHG emissions to be consistent with the cost reduction by and services. The equality takes the form x = Ax + y,45 the studying the differences between the results from the two standard supply demand balance equation in input-output objective functions. Finally, we discuss the technological and analysis. The second group represents the supply demand political implications of our work. balance regarding scrap. The third group of constraints refers to 13087 DOI: 10.1021/acs.est.7b04477 Environ. Sci. Technol. 2017, 51, 13086−13094 Environmental Science & Technology Policy Analysis the composition of metals that are specific to alloy steels and is result of WIO-MFA22 with the AEs demands for EAF determined based on technical conditions, whereas the first and steelmaking indicated in the WIO-MFA table,6 rather than second groups of constraints refer to market balance against industrial standards. Since Løvik et al. pointed out the conditions. We consider non-negativity constraints on variables presence of several industrial standards for alloys, defining the and two alternative objective functions to minimize: GHG representative allowable ranges of AE concentration is difficult, emission and production cost. limiting the usefulness of the allocation model approach.13 The production recipes, that is, the combinations of raw Addressing this challenge, we choose to employ the WIO-MFA material inputs, for producing various types of alloy steel are table as a reference of AE demands in EAF steelmaking, instead the design variables in this linear program. In general, there are of various industrial standards. The database comprises the two methods to implement this optimization of production weighted average of AE inputs in the production of each grade recipes in input-output analysis. One is to introduce additional of EAF steel, considering both industrial standards and their columns and variables, which represent alternative production share in the production of each steel grade.6,12 Accordingly, the technologies to produce alloy steel; this results in the input supply and demand information for AEs in the WIO-MFA table coefficient matrix A having more columns than rows.35,36 This is well represented, reflecting the practical conditions of EAF method is suitable when “extreme” production recipes are steelmaking. For simplicity, this study avoids the issues of known and any feasible recipe can be expressed as a weighted contamination by nonferrous elements, other than AEs, by average, or convex combination, of extreme alternatives. assuming that parts scrap is completely separated and sorted However, this is not the case in our study. The other method from the mechanically combined (i.e., nonalloyed) nonferrous that we employed is to treat input coefficients as variables. Our elements, which are relatively easily separated by ELV problem can be modeled using a linear program, coping with treatments such as magnetic and/or eddy current separation.46 the nonlinearity introduced by this method along the lines of As in our previous study, we estimate the AE composition of an Kondo and Nakamura.39 It should be noted that the third set of ELV generated in 2005 with the WIO-MFA table of that year, group constraints, showing the mass balance of each sector and thus neglect any change in the composition that could have regarding metal elements, plays the role of keeping variable taken place over the life of the ELV.22 These are strong input coefficients valid in the sense that requirements for the assumptions that disregard any possible variability in material elemental composition of produced alloy steel are fulfilled. composition and the AE contents of automobiles over time. Because parts scrap contains AEs, it can be a secondary AE Given the limited data availability, however, we find it a source and replace some fractions of AEs derived from virgin reasonable approach to be adopted for obtaining a benchmark. AE sources. The model would enable us to quantify the extent Furthermore, we ignored the varieties of AE contents in the of this substitution under the objectives. parts originated from the different models of automobiles by We introduced additional inequality constraints to our linear referring to the national average composition of automobiles program to rule out meaningless solutions. See Supporting obtained by WIO-MFA instead. In other words, our model Information (SI) for further details. simulates the situation that there are heaps of ES sorted by 2.3. Data. In this study, the WIO-MFA table, developed based on the Japanese IO Table for 2005 in our previous parts regardless of the original model of automobiles. This studies,6,22 provides the data required to formulate constraints assumption is much closer to the real situation of practical ELV for the optimization of parts scrap use. The WIO-MFA table is recycling rather than identifying and sorting the scrap based on an extended IO table for flow analysis in the level of AEs by models or types of automobiles.47 disaggregating both raw material sectors (e.g., Ferroalloys) and As virgin AE sources, ferroalloys (e.g., ferromanganese, crude-steel production sector in the original IO table based on silicomanganese, ferrochromium, ferronickel, and ferromolyb- production reports obtained by Japanese main steelmakers, denum), metallic AEs (e.g., metallic manganese and metallic industrial standards and their share among the production of nickel), and other sources (e.g., molybdenum briquettes) were each steel grade.6 For example, the Ferroalloys sector, which considered. For the grade of steel to be produced by EAF, we had been aggregated into only one sector for all metal species in define carbon steel and the 19 grades of alloy steel separately. the original IO table, was disaggregated into nine sectors such Alloy steel in this study represents a steel that requires contents as Ferromanganese, Ferrochromium and so on corresponding of AEs and/or special treatments to obtain some desired to each metal species.6 Crude-steel production sector was properties (i.e., “specialty steel” in Japanese categorization21). disaggregated into 58 sectors on the basis of two production Accordingly, carbon steel is steel that is not categorized as alloy methods (i.e., basic oxygen furnace [BOF] and EAF) and 29 steel in this study. For detailed definitions on steel grades, see steel grades.6 The WIO-MFA26,27 was conducted based on the SI. The embodied GHG emissions and price intensity of each table, and enabled us to obtain material compositions and AE virgin AE source, were separately set for domestically produced contents in approximately 400 kinds of products.6,22 The virgin AE sources and imported ones, based on several obtained composition and AE contents of automobile can be references.44,48−50 The intensity for embodied GHG emissions regarded as national average confirmed to be in the same range was referenced from the LCI database IDEA ver 1.040 for with a detailed chemical analysis.28 Besides, AE contents in domestically produced virgin AE sources, and from Ecoinvent49 parts scrap were estimated based on the result of WIO-MFA for for imported virgin AE sources. Regarding the virgin AE automobiles by considering current ELV treatments such as sources’ purchase costs, the unit price of each virgin AE source dismantling of engine and/or reusable parts, pressing, and was obtained from an appendix of the 2005 Japanese IO table44 shredding.6,22 For more details on the development of WIO- and Japanese trade statistics50 for domestic and imported virgin MFA table and the results of the analysis, see our previous AE sources, respectively. For the numerical values regarding the works.6,22 sets and constants, see SI. To optimize parts scrap use, in terms of its AE content, we We conducted a sensitivity analysis with regard to the compare the AE content in parts scrap estimated based on the influence of variations of data on the results by using the dual 13088 DOI: 10.1021/acs.est.7b04477 Environ. Sci. Technol. 2017, 51, 13086−13094 Environmental Science & Technology Policy Analysis Figure 1. Optimized flows of scrap mass (upper half) and alloying elements (bottom half) under two objective functions (left half and right half) with satisfying demands for alloying elements in EAF steel production in Japan in 2005. The left-hand side and right-hand side of each Sankey diagram respectively depict separated parts scrap to be used for steel production by EAF, and alloy steel grades chosen to be produced by using parts scrap from 19 grades as well as carbon steel. Steel grades not appearing in the figure were unable to use parts scrap in the production due to certain discordances between contents and requirements of alloying elements in steel grades and parts scrap, respectively. properties of the LP. The results from the sensitivity analysis objective functions, parts scrap was almost entirely consumed are explained in the SI. in alloy steel production, with only a small fraction ending up in carbon steel. This implies that most AEs were efficiently utilized 3. RESULTS AND DISCUSSION as a secondary source of AE in EAF steelmaking: approximately 10 200 t (t) of manganese, 19 900 t of chromium, 5500 t of 3.1. Optimized Scrap Usage. Figure 1 shows the nickel, and 570 t of molybdenum were utilized as AE sources, optimized flows of scrap mass and alloying element under the respectively. These masses correspond to 94%, 99%, 98%, and two objective functions. It turned out that of the 19 grades of 98% of manganese, chromium, nickel, and molybdenum alloy steel, only 9 grades were assigned for parts scrap usage with carbon steel. The requirements for AEs in the other 10 occurring in parts scrap, respectively. In total, 97% of AEs in grades do not match with the AE content of any parts scrap. parts scrap can be effectively utilized as AE sources in EAF For these 10 grades of alloy steel, parts scrap cannot be used steelmaking. According to our previous study, 93% of ES was without the dilutive input of iron, or intended removal by remelted without sorting in EAF steelmaking to produce oxidation. For example, Cr stainless steel (i.e., ferritic stainless carbon steel in Japan in 2005,6 implying that 93% of AEs steel) was not assigned as a parts scrap application, because this contained in ES dissipated into carbon steel and steelmaking alloy steel does not require nickel input, but requires significant slag. Furthermore, the ratio of AE recovery from parts scrap amounts of chromium, some manganese, and molybdenum. was estimated as 78% by a scenario-based approach.22 Our This demand pattern for AEs in Cr stainless steel production results thus demonstrate significant benefits of sorting ES into did not allow for the input of any parts scrap, as all parts scrap parts scrap and appropriately utilizing them on the recovery of contains traces of nickel. The scrap flows toward carbon steel AEs contained in ES. imply the loss of AEs into carbon steel as the nonfunctional 3.2. Benefits of Parts Scrap Utilization. Benefits were inclusion. quantified by referring to the base state of Japan in 2005 as On a scrap mass basis, automobile body parts dominated the described in the WIO-MFA table (hereafter, referred to as BS). flows, as they have the largest mass among the parts scraps,22 Figure 2 shows that the optimal use of parts scrap can reduce whereas on an AE mass basis, exhaust parts dominated the the input of AEs from virgin sources by 10% of BS for both the flows of chromium, nickel, and molybdenum. Under both the objective functions. Figure 2 also shows that the extent of 13089 DOI: 10.1021/acs.est.7b04477 Environ. Sci. Technol. 2017, 51, 13086−13094 Environmental Science & Technology Policy Analysis virgin sources consumed in EAF steelmaking at BS. Compared with the results obtained under cost minimization, this reduction is 1.5 times larger. Regarding the total purchase cost of virgin sources, 31.6 billion Japanese yen (JPY) (ca. 287 million USD at 1 USD = 110 JPY) could be optimally reduced, which corresponds to 15.2% of the value of virgin sources consumed in EAF steelmaking at BS. For both GHG emission and costs, the magnitudes of reduction achieved under the optimization exceeded those of our previous research, which shows 318 to 609 kt-CO2eq and 13 billion JPY for the reduction of GHG emission and costs, respectively without employing any optimization algorithms, 22 showing the usefulness of applying optimization to parts scrap utilization for the maximization of benefits. 3.3. Choice of Virgin Sources of AEs. Figure 3 compares the amounts of consumption of virgin AE sources in 2005 with those obtained under each of the objective functions. For chromium, the two objectives resulted in the same results with regard to the choice of virgin chromium sources. Of its four Figure 2. Reduction achievements obtained by optimal parts scrap virgin AE sources, optimization resulted in the entire utilization by referring to base state in Japan in 2005 (BS). Reduction elimination of domestic and imported “Other Cr sources” achievements in GHG and cost are different depending on the and “FeCr,” whereas for “FeCr (Imp)” the reduction was kept objective functions, whereas almost the same mass of AEs derived at a modest level. For the other AEs, the results differed from virgin sources are reduced. depending on the choice of the objective function. For manganese, “Metallic Mn (Imp)” was reduced under cost reduction in GHG emissions and the purchase cost of virgin AE minimization, while “SiMn” was reduced under GHG emission sources differ depending on the objective function. Regarding minimization. For nickel, domestic/imported “other Ni GHG minimization, 749 kt-CO2eq could be reduced, which sources” were reduced under cost minimization, whereas corresponds to 28.3% of the GHG emissions embodied in “FeNi (Imp)” and “Metallic Ni” were reduced under GHG Figure 3. Consumption of virgin AE sources in EAF steelmaking observed in base state in Japan in 2005 (BS) and obtained by the optimization under each of the objective functions. Details for virgin AE sources indicated by simplified labels in the graph are shown in the SI. 13090 DOI: 10.1021/acs.est.7b04477 Environ. Sci. Technol. 2017, 51, 13086−13094 Environmental Science & Technology Policy Analysis Figure 4. Frontiers maximizing the benefit of parts scrap utilization (a) and the breakdown of the frontier by AEs (b). Points A to E represent the points where the frontier is bent and correspond to the ranges of certain weights between two objective functions. For example, Point C represents the reduction ratios for the cost reduction and the GHG reduction at the weight (0.98, 0.02) to (0.87, 0.13). Sums of reduction ratios of points in (b) are consistent with the ratios in (a). For example, the sum of GHG reduction ratios of Point B for four AEs and iron comes to about 24% of the same sum for Point B in (a). emission minimization. For molybdenum, domestic/imported weights. By combining with Figure 3, Point C, which would be “FeMo” was reduced under both objective functions, but with a the best balance between the objectives, can be achieved by slight quantitative difference. Under cost minimization, the choosing “SiMn” for manganese, and “FeNi (Imp)” and demand for AE virgin sources with higher prices would be “Metallic Ni” for nickel instead of virgin AE sources chosen reduced, whereas under GHG minimization those with larger under the objective function minimizing the costs (for detailed embodied GHG emission intensities would be reduced. methodology and results, see SI). Consequently, if a high-price virgin AE source happened to 3.5. Implications and Future Directions. This study have low GHG emission intensity, a trade-off relationship demonstrated that appropriate utilization of parts scrap as would emerge between the solutions of the objective functions secondary AE sources reduced both GHG emissions and the (see SI for more details on the results for individual virgin AE costs of virgin AE sources. The proposed ES recycling strategy sources). sorting ES into just eight kinds of parts scrap and utilizing them 3.4. Multiobjective Optimization. The trade-off relation- appropriately in EAF steelmaking could make parts scrap an ship can be visualized as a frontier as shown in Figure 4 by indispensable source of urban mining. It implies that more conducting a multiobjective optimization. This procedure precise sorting of ES by grades of steel will bring better benefits enables us to find weights balancing both the reductions of on AE saving in ES recycling. In current ES recycling, however, GHG emissions and of purchase costs for virgin AE sources. ES is usually regarded as an “iron source,” with its AE content The result of this multiobjective optimization can support mostly neglected,6 even in Japan and EU, which have decision making regarding parts scrap utilization. See the SI for established automobile recycling laws under 3R (Reduce, methodological details of our multiobjective optimization. Reuse, and Recycling) policies51 and ELV directive, 52 In Figure 4 (a), every point on the frontier represents the respectively.53 The laws only regulate the mass-based recycling maximum benefits of both objective functions. The points, A, B, rate of materials from ELVs,54 and do not control the use of ES C, D, and E on the frontier correspond to the benefits for both in EAF steelmaking. While discussions on the concept and objective functions at a certain weight ratio between the definition of the Circular Economy is still ongoing,55,56 minimization of costs and GHG emissions. Points A and E importance of the quality of recycling and subsequent represent the benefits when we consider either only costs preservation of materials are pointed out as one of its essential minimization or embodied GHG minimization, respectively. A components.57 This study provides a theoretical benchmark for slight consideration for embodied GHG reduction has a large implementation of this important strategy for steel material impact on the benefits. Point A corresponds to the maximum used in automotive sector. reduction in costs. Point C indicates, however, that if the extent Toward the development of a quality Circular Economy, of cost reduction was kept at a slightly higher level, 14.4% amendment of current policies is required to engage the instead of 15.2%, a reduction in embodied GHG of 27.6% industries utilizing the recycled materials as well as the recyclers could be achieved, which is very close to its maximum and producers of the products subject to recycling. The current reduction. This implies the possibility of a slight sacrifice in cost policies only regulate activities of ELV recyclers in terms of reduction resulting in a significant reduction in GHG emissions. ES.53 A higher level of engagement of EAF steelmakers would Figure 4(b) decomposes the points on the frontier by be an important target of the amended policies. For the elements. By observing correspondences of the points between purpose of exploiting the maximal benefits of AE content- Figure 4(a) and (b), we can understand which AEs contributed oriented ES recycling in grade level, ELV recyclers are expected to the shift of the points. The shift from Points A to B was to play larger roles than EAF steelmakers. Investment made by the change in choice in the virgin manganese sources; requirements for ELV recyclers, such as the installation of Points B to C, and Points C to D were caused by virgin nickel additional processes and/or equipment, are likely to discourage sources; and Points D to E were due to virgin molybdenum them from committing to detailed disassembling and sorting of sources. Besides these changes, chromium and iron contributed ES, unless compensated for by the proper pricing of ES instead to the benefits of both reductions with constant choices at any of the current pricing one on the shapes of scrap alone,22 and/ 13091 DOI: 10.1021/acs.est.7b04477 Environ. Sci. Technol. 2017, 51, 13086−13094 Environmental Science & Technology Policy Analysis or economic incentives supported by policies to encourage EAF ORCID steelmaker to purchase and utilize ES. In addition to ELV Hajime Ohno: 0000-0002-8826-3854 recyclers, developers of scrap sorting technologies accurately Shinichiro Nakamura: 0000-0002-7735-024X detecting elements and sorting steel scrap according to AE Yasuhiro Fukushima: 0000-0002-1525-7242 contents are also important players in the system for the Notes exploitation of potential benefits. Recently, technologies that The authors declare no competing financial interest. rapidly determine AE composition in steel material by applying X-ray58,59 or laser60,61 have been developed. The development ACKNOWLEDGMENTS of automatic separation equipment and/or process applying these technologies is expected to be the next step. Financial This research was supported by the Japan Society for the support for the technology developers is another important role Promotion of Science (JSPS) (KAKENHI 23686131, of policies. Development of better sorting technologies allow 26281059, 15K12265, and Grant-in-Aid for JSPS Research ELV recyclers to sort ES precisely according to AE content and Fellow 258801), the Iron and Steel Institute of Japan (a provide useful secondary sources of iron and AEs for EAF research group for recycling automobiles from the perspective steelmakers. Not only for ELV recycling but for all kinds of of the material industry), and the Research Institute of Science waste, policies covering all the stakeholders in waste treatment and Technology for Society of the Japan Science and Technology Agency (JST-RISTEX). systems are necessary toward the development of a quality Circular Economy. REFERENCES Since these benefits were estimated based on the prices of virgin AE sources, prices for each type of parts scrap may also (1) United Nations Environment Programme (UNEP). Metal recycling: Opportunities, limits, infrastructure2013. be decided based on the AE content in each scrap. Under cost (2) Reck, B. K.; Graedel, T. E. Challenges in metal recycling. Science minimization, the optimum pattern of the use of parts scrap 2012, 337 (6095), 690−5. would vary with the prices of virgin AE sources. AE value- (3) Graedel, T. E.; Allwood, J.; Birat, J.-P.; Buchert, M.; Hagelüken, oriented optimization and AE content-oriented ES price C.; Reck, B. K.; Sibley, S. F.; Sonnemann, G. What do we know about estimation will be a future direction for this study. This study metal recycling rates? J. Ind. Ecol. 2011, 15 (3), 355−366. considered two objective functions. Given the availability of (4) Nakamura, S.; Kondo, Y.; Matsubae, K.; Nakajima, K.; Tasaki, T.; relevant inventory data, a larger set of objective functions can Nagasaka, T. Quality- and dilution losses in the recycling of ferrous be considered. This would help us identify more practical materials from end-of-life passenger cars: Input-output analysis under information relevant for a wide range of stakeholders. Another explicit consideration of scrap quality. Environ. Sci. Technol. 2012, 46 important direction for future research would be to extend its (17), 9266−73. (5) Nakajima, K.; Takeda, O.; Miki, T.; Matsubae, K.; Nagasaka, T. scopeJapan, for this studyon a global scale. Japan has both Thermodynamic analysis for the controllability of elements in the EAF steelmakers and ELV recyclers as stakeholders, but other recycling process of metals. Environ. Sci. Technol. 2011, 45 (11), 4929− countries and regions may not. Since ELVs are generated 36. worldwide,62,63 this difference in the existence of stakeholders (6) Ohno, H.; Matsubae, K.; Nakajima, K.; Nakamura, S.; Nagasaka, could be included in the discussion on global scrap price T. Unintentional flow of alloying elements in steel during recycling of decisions. International price decisions on ES, based on AE end-of-life vehicles. J. Ind. Ecol. 2014, 18 (2), 242−253. content, will affect the attitudes of countries regarding ELV (7) Ohno, H.; Nuss, P.; Chen, W. Q.; Graedel, T. E. Deriving the treatment and activate the trade for ES as a useful urban mine metal and alloy networks of modern technology. Environ. Sci. Technol. of AE. In facing a rapidly motorizing world, an efficient 2016, 50 (7), 4082−90. recycling system for ELV will be required not only at the (8) Gleich, A. v.; Ayres, R. U.; Gössling-Reisemann, S. Sustainable metals management: Securing our future-steps towards a closed loop stakeholder level in a region, but also at a global level.64 economy; Springer: Dordrecht, 2006; p xvi, 607 p. Furthermore, we can apply this model to every type of steel (9) Hiraki, T.; Takeda, O.; Nakajima, K.; Matsubae, K.; Nakamura, scrap. Along with ES, quality-oriented recycling of mass- S.; Nagasaka, T. Thermodynamic criteria for the removal of impurities generated construction steel scrap, industrial machinery, and from end-of-life magnesium alloys by evaporation and flux treatment. other types of steel scrap can be evaluated by determining their Sci. Technol. Adv. Mater. 2011, 12 (3), 035003. representative AE compositions. Implementing steel scrap (10) Nakamura, S.; Kondo, Y.; Kagawa, S.; Matsubae, K.; Nakajima, recycling on a larger scale, we can develop an efficient steel K.; Nagasaka, T. MaTrace: Tracing the fate of materials over time and and AE urban mining society. across products in open-loop recycling. Environ. Sci. Technol. 2014, 48 (13), 7207−14. ASSOCIATED CONTENT (11) Pauliuk, S.; Kondo, Y.; Nakamura, S.; Nakajima, K. Regional distribution and losses of end-of-life steel throughout multiple product * Supporting Information S life cycles-insights from the global multiregional MaTrace model. The Supporting Information is available free of charge on the Resour. Conserv. Recycl. 2017, 116, 84−93. ACS Publications website at DOI: 10.1021/acs.est.7b04477. (12) Nakajima, K.; Ohno, H.; Kondo, Y.; Matsubae, K.; Takeda, O.; Miki, T.; Nakamura, S.; Nagasaka, T. Simultaneous material flow The detailed methodology for the derivation of the linear analysis of nickel, chromium, and molybdenum used in alloy steel by program, multiobjective optimization, the background means of input-output analysis. Environ. Sci. Technol. 2013, 47 (9), data for the program, and detailed results on the choice 4653−60. of virgin AE sources as well as multiobjective (13) Lovik, A. N.; Modaresi, R.; Muller, D. B. Long-term strategies optimization, and sensitivity analysis (PDF) for increased recycling of automotive aluminum and its alloying elements. Environ. Sci. Technol. 2014, 48 (8), 4257−65. AUTHOR INFORMATION (14) Yamada, H.; Hatayama, H.; Daigo, I.; Matsuno, Y.; Adachi, Y. Optimization method for metal recycling system and its application to Corresponding Author aluminum recycling flow in Japan. Nippon Kinzoku Gakkaishi 2006, 70 *Phone +81-22-795-5869; e-mail: [email protected]. (12), 995−1001. 13092 DOI: 10.1021/acs.est.7b04477 Environ. Sci. Technol. 2017, 51, 13086−13094 Environmental Science & Technology Policy Analysis (15) Gaustad, G.; Olivetti, E.; Kirchain, R. Toward sustainable products may rely on several technologies simultaneously. J. Econ. material usage: Evaluating the importance of market motivated agency Struct. 2012, 1 (1), 3. in modeling material flows. Environ. Sci. Technol. 2011, 45 (9), 4110− (37) Springer, N. P.; Duchin, F. Feeding nine billion people 7. sustainably: Conserving land and water through shifting diets and (16) Olivetti, E. A.; Gaustad, G. G.; Field, F. R.; Kirchain, R. E. changes in technologies. Environ. Sci. Technol. 2014, 48 (8), 4444−51. Increasing secondary and renewable material use: A chance con- (38) López-Morales, C. A.; Duchin, F. Economic implications of strained modeling approach to manage feedstock quality variation. policy restrictions on water withdrawals from surface and underground Environ. Sci. Technol. 2011, 45 (9), 4118−26. sources. Econ. Systems Res. 2015, 27 (2), 154−171. (17) Daigo, I.; Fujimaki, D.; Matsuno, Y.; Adachi, Y. Development of (39) Kondo, Y.; Nakamura, S. Waste input−output linear a dynamic model for assessing environmental impact associated with programming model with its application to eco-efficiency analysis. cyclic use of steel. Tetsu to Hagane 2005, 91 (1), 171−178. Econ. Systems Res. 2005, 17 (4), 393−408. (18) U.S. Geological Survey (USGS). 2011 minerals yearbook. (40) Nakamura, S.; Kondo, Y. Input-output analysis of waste USGS: Virginia, 2011.10.15385/yb.miracle.2011 management. J. Ind. Ecol. 2002, 6 (1), 39−63. (19) The Japan Iron and Steel Federation, Japanese steel production. (41) Lin, C. Identifying lowest-emission choices and environmental In The Japan Iron and Steel Federation,: Tokyo, 2013. pareto frontiers for wastewater treatment wastewater treatment input- (20) International Organization of Motor Vehicle Manufactures output model based linear programming. J. Ind. Ecol. 2011, 15 (3), (OICA), World motor vehicle production. (accessed April 16, 2013, 367−380. 2013). (42) Farel, R.; Yannou, B.; Bertoluci, G. Finding best practices for (21) The Japan Iron and Steel Federation. Order booked of steel automotive glazing recycling: A network optimization model. J. Cleaner products. http://www.jisf.or.jp/en/statistics/order/TimeSeries.html Prod. 2013, 52 (0), 446−461. (accessed November 22, 2013). (43) Simic, V. Fuzzy risk explicit interval linear programming model (22) Ohno, H.; Matsubae, K.; Nakajima, K.; Kondo, Y.; Nakamura, for end-of-life vehicle recycling planning in the eu. Waste Manage. S.; Nagasaka, T. Toward the efficient recycling of alloying elements 2015, 35 (0), 265−82. from end of life vehicle steel scrap. Resour. Conserv. Recycl. 2015, 100, (44) Ministry of Industrial Affairs and Communications (Japan). 11−20. 2005 input-output table for Japan. Ministy of Industrial Affairs and (23) Reijnders, L. Conserving functionality of relatively rare metals Communications; Tokyo, 2009. associated with steel life cycles: A review. J. Cleaner Prod. 2016, 131, (45) Miller, R. E.; Blair, P. D. Input-output analysis: Foundations and 76−96. extensions. 2nd ed.; Cambridge University Press: Cambridge England ; (24) Daigo, I.; Goto, Y. Comparison of tramp elements compositions New York, 2009; p xxxii, 750 p. in steel bars between Japan and China. Tetsu to Hagane 2014, 100 (6), (46) Gaustad, G.; Olivetti, E.; Kirchain, R. Improving aluminum recycling: A survey of sorting and impurity removal technologies. 756−760. Resour. Conserv. Recycl. 2012, 58, 79−87. (25) Oda, T.; Daigo, I.; Matsuno, Y.; Adachi, Y. Substance flow and (47) Ministry of Environment (Japan). The advanced utilization of the stock of chromium associated with cyclic use of steel in Japan. Tetsu to end of life vehicle originated scraps: As the technical report of ministry of Hagane 2009, 95 (10), A720−A729. environment, Japan funded project ″ advanced metal recycling technology (26) Nakamura, S.; Nakajima, K. Waste input-output material flow development″. Tohoku University: Sendai, 2012. analysis of metals in the Japanese economy. Mater. Trans. 2005, 46 (48) National Institute of Advanced Industrial Science and (12), 2550−2553. Technology (AIST), Japan Environmental Management Association (27) Nakamura, S.; Nakajima, K.; Kondo, Y.; Nagasaka, T. The waste for Industry (JEMAI). LCI database IDEA ver 1.0. In 2012. input-output approach to materials flow analysis - concepts and (49) Ecoinvent. Ecoinvent life cycle inventory database v2.2. 2010. application to base metals. J. Ind. Ecol. 2007, 11 (4), 50−63. (50) Ministry of Finance (Japan). Trade statistics of Japan. Ministry of (28) Japan Environmental Sanitation Center. Report for implemen- Finance Japan: Japan, Tokyo, 2005. tation of efficient and rational treatment for end of life vehicles. Japan (51) Ministry of Economy, Trade and Industry (Japan). End-of-life Environmental Sanitation Center: Japan, 2009. vehicle recycling law. http://www.meti.go.jp/policy/recycle/main/ (29) Nakamura, S.; Kondo, Y.; Nakajima, K.; Ohno, H.; Pauliuk, S. english/law/end.html (accessed July 6, 2017). Quantifying recycling and losses of Cr and Ni in steel throughout (52) European Commission. Directive 2000/53/ec of the european multiple life cycles using MaTrace-alloy. Environ. Sci. Technol. 2017, 51 pariament and of the council of 18 september 2000 on end-of-life (17), 9469−9476. vehicles. 2000. (30) Azapagic, A.; Clift, R. Life-cycle assessment and linear- (53) Sakai, S.-i.; Yoshida, H.; Hiratsuka, J.; Vandecasteele, C.; programming - environmental optimization of product system. Kohlmeyer, R.; Rotter, V. S.; Passarini, F.; Santini, A.; Peeler, M.; Li, J.; Comput. Chem. Eng. 1995, 19, S229−S234. Oh, G.-J.; Chi, N. K.; Bastian, L.; Moore, S.; Kajiwara, N.; Takigami, (31) Saner, D.; Vadenbo, C.; Steubing, B.; Hellweg, S. Regionalized H.; Itai, T.; Takahashi, S.; Tanabe, S.; Tomoda, K.; Hirakawa, T.; LCA-based optimization of building energy supply: Method and case Hirai, Y.; Asari, M.; Yano, J. An international comparative study of study for a Swiss municipality. Environ. Sci. Technol. 2014, 48 (13), end-of-life vehicle (elv) recycling systems. J. Mater. Cycles Waste 7651−9. Manage. 2014, 16 (1), 1−20. (32) Hirshfeld, D. S.; Kolb, J. A. Analysis of energy use and co2 (54) Andersson, M.; Ljunggren Söderman, M.; A. Sandén, B. Lessons emissions in the u.S. Refining sector, with projections for 2025. from a century of innovating car recycling value chains. Environ. Innov. Environ. Sci. Technol. 2012, 46 (7), 3697−704. Soc. Transit. 2017, DOI: 10.1016/j.eist.2017.03.001. (33) Anghinolfi, D.; Paolucci, M.; Robba, M.; Taramasso, A. C. A (55) Kirchherr, J.; Reike, D.; Hekkert, M. Conceptualizing the dynamic optimization model for solid waste recycling. Waste Manage. circular economy: An analysis of 114 definitions. Resour. Conserv. 2013, 33 (2), 287−96. Recycl. 2017, 127, 221−232. (34) Duchin, F. A world trade model based on comparative (56) Korhonen, J.; Honkasalo, A.; Seppälä, J. Circular economy: The advantage withmregions,ngoods, andkfactors. Econ. Systems Res. 2005, concept and its limitations. Ecol. Econ. 2018, 143, 37−46. 17 (2), 141−162. (57) European Commission. Communication from the commission (35) Duchin, F.; Levine, S. H. Sectors may use multiple technologies to the european parliament, the council, the european economic and simultaneously: The rectangular choice-of-technology model with social committee and the committee of the regions: Closing the loop - binding factor constraints. Econ. Systems Res. 2011, 23 (3), 281−302. an EU action plan for the circular economy. 2015. (36) Duchin, F.; Levine, S. The rectangular sector-by-technology (58) Matsubae, K.; Iizuka, Y.; Ohno, H.; Hiraki, T.; Miki, T.; model: Not every economy produces every product and some Nakajima, K.; Nagasaka, T. Distribution analysis on steel alloying 13093 DOI: 10.1021/acs.est.7b04477 Environ. Sci. Technol. 2017, 51, 13086−13094 Environmental Science & Technology Policy Analysis elements in the end of life vehicle scrap recycling process. Tetsu to Hagane 2014, 100 (6), 788−793. (59) Daigo, I.; Goto, Y. Comparison of tramp element contents of steel bars from Japan and China. ISIJ Int. 2015, 55 (9), 2027−2032. (60) Kashiwakura, S.; Wagatsuma, K. Rapid sorting of stainless steels by open-air laser-induced breakdown spectroscopy with detecting chromium, nickel, and molybdenum. ISIJ Int. 2015, 55 (11), 2391− 2396. (61) Li, K.; Guo, L.; Li, C.; Li, X.; Shen, M.; Zheng, Z.; Yu, Y.; Hao, R.; Hao, Z.; Zeng, Q.; Lu, Y.; Zeng, X. Analytical-performance improvement of laser-induced breakdown spectroscopy for steel using multi-spectral-line calibration with an artificial neural network. J. Anal. At. Spectrom. 2015, 30 (7), 1623−1628. (62) Simic, V.; Dimitrijevic, B. Risk explicit interval linear programming model for long-term planning of vehicle recycling in the eu legislative context under uncertainty. Resour. Conserv. Recycl. 2013, 73 (0), 197−210. (63) Tian, J.; Chen, M. Sustainable design for automotive products: Dismantling and recycling of end-of-life vehicles. Waste Manage. 2014, 34 (2), 458−67. (64) Restrepo, E.; Lovik, A. N.; Wager, P.; Widmer, R.; Lonka, R.; Muller, D. B. Stocks, flows, and distribution of critical metals in embedded electronics in passenger vehicles. Environ. Sci. Technol. 2017, 51 (3), 1129−1139. 13094 DOI: 10.1021/acs.est.7b04477 Environ. Sci. Technol. 2017, 51, 13086−13094