Metrics for End-of-Life Strategies (ELSEIM)

Catherine M. Rose, Ab Stevels 1 Design for Sustainability Research Group Delft University of Technology Jaffalaan 9 2628 BX Delft, The Netherlands +31 15 278 1524, fax: +31 15 278 2956 rose@cdr.stanford.edu, ab.stevels@philips.com
Abstract - The environmental impact of a products end-of-life is modeled, using the End-of-Life Strategy Environmental Impact Model (ELSEIM). The model is based on data from case studies provided by Philips Consumer Electronics. The model calculates the environmental impact across all possible end-of-life strategies (reuse, service, remanufacture, recycle, and disposal). The environmental impacts include the following: manufacturing or extraction, transportation, energy, end-of-life, and packaging. Results from case studies from Philips Consumer Electronics show that the popular belief that reuse is the best end-of-life strategy to reduce environmental impact is only partly true: For television, the reuse strategy instead of the current recycling with disassembly strategy brings little gains. If future energy consumption of televisions decreases further the gain will turn into a loss. Cellular phones are assessed as good candidates for reuse and should be redesigned accordingly. dealing with the product at the end-of-life. The endof-life strategy hierarchy is based on decreasing environmental impact [1]: Reuse: Reuse is the second hand trading of product for use as originally designed (through yard sale, ebay.com, donation, etc). Service: Servicing the product is another way of extending the life of a durable product or component parts by repairing or rebuilding the product using service parts at the location where the product is being used. Remanufacture: Remanufacturing is a process in which reasonably large quantities of similar products are brought into a central facility and disassembled. Parts from a specific product are not kept with the product but instead they are collected by part type, cleaned, inspected for possible repair and reuse. Remanufactured products are then reassembled on an assembly line using those recovered parts and new parts where necessary. Recycle (with disassembly): Recycling reclaims material streams useful for application in products. Disassembly into material fractions increases the value of the materials recycled by removing material contaminants, hazardous materials, or high value components. The components are separated mostly by manual disassembly methods. Recycle (without disassembly): Using mechanical shredding technologies reduces material size which then is separated using methods based on magnetic, density or other properties of the materials. Disposal: This end-of-life strategy is to landfill or incinerate the product with or without energy recovery.

I. INTRODUCTION OF END- OF-LIFE STAGE A product being designed can be optimized for individual life cycle phases, ranging from material extraction, manufacture, assembly, use and end-oflife. Life cycle engineering, or the life cycle perspective, aims to optimize these stages together, instead of separately. Trade-offs are necessary in this process to develop the optimal product that balances the gains and losses in the focal areas: energy usage, material usage, packaging, chemical content and endof-life. The concern for product end-of-life grows as consumer interest and market activity moves for greater producer responsibility. Market forces, especially in business-to-business activities, are encouraging companies to examine more closely the treatment of the product at end-of-life. End-of-life strategies describe the strategy associated with
1

II. PREVIOUS RESEARCH IN ENVIRONMENTAL IMPACTS OF END- OF-LIFE STAGE

The end-of-life analysis of a product must account for three major variables: cost, environmental impact and

Ab Stevels is also Senior Environmental Advisor at Philips Consumer Electronics Environmental Competence Centre.

product characteristics. Previous work in using product characteristics to determine the end-of-life strategy is well documented [1-3]. The product characteristics dictate what is allowable at the end-oflife. The cost analysis of product end-of-life is performed by researchers at Delft University of Technology [4]. The analysis of the actual environmental impact of products at end-of-life has been overlooked. Previous analyses include end-of-life as a subsection of a larger analysis on the entire product life cycle. These approaches often use a simplified view of the product end-of-life - accounting for only two end-oflife strategies: recycle or disposal (through incineration or landfill). These studies do not include the relevant end-of-life strategies such as reuse, service or remanufacture as possible treatments at the end-of-life. Additionally, these analyses rely on the MRE, material recycling efficiency, which assumes eliminating pounds of product weight corresponds to reductions in environmental impact. Current research seeks to quantify product end-of-life environmental impact by managing environmentally relevant materials through Environmentally Weighted Recycling Quotes (EWRQ), which weigh every material fraction according to its real environmental impact [5]. The advantages of their approach over conventional weight based recycling metrics are greater ability to maximize the control of enviornmental impact at minimum cost and characterizes true environmental performance of products in their end-of-life [6].

extraction and manufacturing processes. For some materials the extraction environmental impact is much higher than the manufacturing processing. The environmental impact quantifies the impact of usage conditions (batteries or grid electricity), as well. Packaging environmental impact accounts for the impact of plastic, cardboard and paper used in packing materials. In the current calculation, the endof-life impact is only accounted for by the disposal through incineration. The Ecoscan results are given in millipoints, quantifying the environmental impact of the product and higher values denote greater environmental impact. The following table gives the environmental impacts (in millipoints) for selected Philips Consumer Electronics products (for example [9]). Table 1. Life Cycle Phase Environmental Impact for Selected Philips Consumer Electronics products
Products Cell Phone VCR CRT Monitor LCD Monitor Portable CD Player Audio Product TV Manufact uring 51 183 723 1404 58 Usage 18 215 577 269 1266 Packaging 0.3 8 57 30 3 Endof-Life 3 14 52 12 536 Total 72.3 420 1409 1715 1863

535 396

1393 1611

10 12

22 40

1960 2059

III. CASE STUDIES Philips Consumer Electronics Environmental Competence Centre performs extensive case studies that examine energy usage, environmentally relevant materials, end-of-life, material composition, and packaging. In order to quantify the products environmental performance, the products are disassembled and the material content, manufacturing process and component weight are recorded. Philips researchers have performed this environmental benchmarking on approximately seventy consumer electronic products. Philips Environmental Competence Centre then uses this information to perform a life cycle analysis on the product using Ecoscan [7]. The version of Ecoscan used is based on Ecoindicator 95 [8]. Ecoscan, like most Life Cycle Assessment (LCA) tools, examines the entire life cycle of the product, from extraction, manufacture, packaging, usage to end-of-life. The manufacturing environmental impact includes

IV. END- OF-LIFE STRATEGY ENVIRONMENTAL IMPACT MODEL

The environmental impact of a products end-of-life is modeled using data from these case studies, and is abbreviated ELSEIM. The environmental impact analysis results are crucial for determining the environmental impact of different product end-of-life strategies. To consistently perform an environmental impact analysis across all possible end-of-life strategies, it is necessary to determine a reference point. This model assumes that a product is introduced to the market, used for 10 years and then reaches end-of-life, and then calculates how much is the environmental impact to deliver a product in resellable condition. For example, a product that will be reused, does not need any new manufacturing or end-of-life processing before it is ready for resell. On the other hand, if the product has the end-of-life strategy of recycling this requires materials to

manufacture a new product. Since the original product is being recycled, there will be a reduction in the overall environmental impact based on the percentages of recovered material. The environmental impacts include some or all of the following manufacturing and extraction, energy, transportation, disposal, and packaging, depending on the end-of-life strategy. The environmental bonus is based on percentages of materials recovered from recycling. The units of the resulting equations are millipoints. EI stands for environmental impact and LCA represents the values directly derived from the Ecoscan values.
EIlife cycle = EImanufactur + EItransporta + EI packaging e tion + EIusage + EIdisposal + EBbonus
EI manufactur e = (1 + x ) LCA manufactur e

are highest for metals, close to 100%, whereas for wood is 0%. These assumptions combined with the life cycle analysis data collected on the Philips products, yield environmental impact estimates for various end-of-life strategies. The following table displays the values from the ELSEIM analysis on televisions. Table 2. Metrics for end-of-life scenarios for Mainstream Televisions (17")
Environmental Impact Reuse Service Remanufacture Recycle with disassembly Recycle without disassembly 792 2.2 1611 1611 12 2.9 -76.9 3954 Disposal

(1) (2)

Manufacturing Transportation Energy (first life) Energy (second life) Packaging Disposal Bonus Total

396 0 1611 1611 0 0 0 3618

435.6 0.4 1611 1611 0 0 0 3658

554.4 2.2 1611 1611 12 0 -116.5 3674

792 2.2 1611 1611 12 2.9 -291.3 3740

792 2.2 1611 1611 12 17 0 4045

The value x in equation 2 is the percentage of the product that must be manufactured for the second life. The values range from 0% (for reuse), 10% (for service), 40% (for remanufacture), 100% (for recycle and disposal). EI transporta tion = 1 . 131 y w (3) In Equation (3), the transportation required is the distance (y) between the end-user and the recycling or recovery facility and ranges from 20 miles for reuse and 100 miles for disposal. The variable w is the weight of the product in kilograms. The value, 1.131, has units millipoints per mile-kilogram (derived from the Philips Eco-indicator database). EI packaging = LCA packaging (4)
EI energy = LCAenergy(1 st life) + LCA energy(2 nd life)

(5)

ELSEIM accounts for the energy consumption during the first and second life of the product. The LCA values are based on Philips internal data on the consumer behavior and years of operation.
EI disposal = 2 . 1 w electronic s + 2 . 0 w metals + 1 . 8 w plastics + 0 . 8 w misc / glass + 0 . 1 w wood

V. RESULTS OF ELSEIM Performing these calculations for other products reveals interesting conclusions. As shown in Figure 1, the environmental impact is lowest for reuse and highest for disposal. In some cases, such as LCD monitor, the environmental impact varies from approximately 1900 to 3300 millipoints. For other products, such as the cellular phone, the environmental impact of reuse is 88 millipoints and for disposal is 140 millipoints. The range for the portable CD player is 62 millipoints. In all of these calculations, however, the energy consumption was assumed constant over the first and second life of the product.
Figure 1. Environmental impact of End-of-Life Strategies
4500 4000 Environmental Impact (millipoints) 3500 3000 2500 2000 1500 1000 500 0 Cell Phone VCR CRT Monitor LCD Monitor Portable CD Player Audio Product Mainstream TV

(6)

Equation 6 assesses the environmental impact of the product disposal, as a weighted sum of the weights of the materials. The multipliers are derived from the Philips database and are given in units of millipoints per kilogram. The variable, w, is the weight in kilograms of the material fraction.
EBbonus = ( 0. 8 LCAelectronic + 1 .0 LCAmetals s + 0.8 LCA plastics + 0 .8 LCAmisc / glass + 0. 0 LCAwood )

(7)

Recycling of the product yields an environmental bonus rather than impact; hence the value is negative, reducing the overall environmental impact. The recovered material is reapplied, preventing the need to extract more raw materials from the environment. The values are given as the recycling yields, assumed from a disassembly operation. The recycling yields

Example Products Reuse Recycle with Disassembly Service Recycle without Disassembly Remanufacture Disposal

For consumer electronics, energy consumption dominates. Thus, the difference between the best

(reuse) and the worst (disposal) strategies may be less than commonly perceived. The environmental impact of reuse compared to disposal for cell phone 63%, for VCRs 76%, for televisions 89%, and the highest for portable CD player 98%. ELSEIM is dynamic, allowing for changes in the product design, particularly in the reduction of energy consumption of product during its 'second life'. ELSEIM calculates the environmental impact of the products using the product energy consumption of the product, based on the consumer behavior and years of use. For example, portable CD players have a high environmental impact due to their reliance on batteries for power. If the energy consumed by portable CD players is reduced, then the environmental impact of the product is lower. Figure 2 compares the environmental impact of the end-of-life strategies with no reduction in energy consumption and with a possible 20% reduction in energy consumption (when the second product is released). Since the end-of-life strategy of reuse does not take into account the purchase of a new and updated product, there is no change in the environmental impact. However, the recycling and disposal end-of-life strategies have reductions in the environmental impact because there was a reduction in the power consumed by the updated product. The 20% reduction in energy consumption converts to an 11% reduction in the overall environmental impact of the end-of-life strategies (recycle and disposal). Figure 2. Environmental Impact for Portable CD player (Comparison of Energy Reduction)
3000 Environmental Impact 2500 2000 1500 1000 500 0
Se rv ic e Re m an uf ac tu re w ith Di Re sa cy ss cl em e bl w ith y ou tD is as se m bl y Di sp os al Re us e

consumption, where possible technologically, would reduce the environmental impact of the other products as well. However, for some products (such as cellular phone and LCD monitor), reductions in the energy consumption in the second life only have 3% and 1% reductions, respectively, in overall environmental impact. Therefore, for this case, the efforts to reduce the energy consumption (to reduce the environmental impact) are overwhelmed by other life cycle stages. Especially for cellular phones and LCD monitors, a large percentage of the environmental impact occurs in the manufacturing phase rather than usage. Therefore a reuse strategy followed for cell phone and LCD monitors provides relatively high environmental gain. On the other hand, portable CD players are good candidates for a recycling end-of-life strategy, as improvements to energy consumption can be gleaned.

Energy Reduction

No Energy Reduction

The environmental impact of other products, such as the audio product and television, would reduce approximately 8% if the energy consumption were reduced 20%. Obviously, reductions in the energy

VI. PRODUCT DESIGN AVENUES Based on these results, the focus of environmental improvements for these products depend on the areas should be placed on the areas of greatest benefit. For portable CD players and televisions, efforts should be placed on reducing the energy consumption of the products. For cellular phones and LCD monitors, efforts should be expended on reducing the manufacturing environmental impact. Figure 3 shows the environmental impact reduction resulting from energy consumption. The horizontal axis represents the possible energy consumption reduction from 0% to 50% in comparison to the previous model. The vertical axis gives the percentage change in overall environmental impact. For the average energy consumption reduction of 20%, the cell phone's overall environmental impact is reduced 5% whereas for the portable CD player, it is reduced 24%. An additional analysis examining the percentage change in overall environmental impact if the manufacture of the product reduced in impact was performed. As suspected, cellular phones and LCD monitors reaped the most benefit from improvements in the manufacturing environmental impact. For a reduction of 20% in the manufacturing impact, the overall environmental impact of a cellular phone reduced by 17% and for LCD monitors the results were 20%. Whereas, for the portable CD players, a reduction of 20% in manufacturing impact only lowered the total environmental impact by 0.9%.

Re cy cl e

Figure 3. Percentage change in environmental impact resulting from energy consumption reduction
100

Percentage Change in Overall Environmental Impact

90 80 70 60 50 40 30 20 10 0 0 5 10 15 20 25 30 35 40 45 50

Candidates for energy reduction

Candidates for manufacturing reduction

Percentage Energy Consumption Reduction
Cell phone Television VCR Portable CD player CRT Monitor Audio System LCD Monitor

To date, there has been substantial research to reduce the environmental impact of manufacturing and energy stages of a product's life. Research at Delft University of Technology continues in the field of human power, focusing on small electronic products. The effort seeks to identify alternatives that have lower environmental burden than current, batterypowered sources [10, 11]. This work promises opportunities to reduce the environmental footprint of small electronic products, with energy consumption up to 150 mWatt [10]. For portable devices, such as portable CD player and cellular phone, this research could allow for reductions of 1266 millipoints and 18.3 millipoints, respectively. For larger electronic products, Philips Consumer Electronics works to develop solutions reducing the environmental impact from materials, weight and energy consumption. By eliminating superfluous components due to design history, integrating electronics, and minimizing cable and wiring, Philips Consumer Electronics was able to reduce energy consumption of the mainstream television by 39% [12]. Philips Semiconductors, through innovations such as GreenChip, reduces the typical power consumption in stand-by mode from between 5-10 Watts to 1-2 Watts [13]. These developments are relevant, as the U.S. Department of Energy estimates that Americans spend $1 billion annually powering TVs and VCRs in stand-by [13].

VII. RELEVANCE IN OVERALL END- OF-LIFE STRATEGY DEVELOPMENT

These calculations may seem extraneous to the quantification and analysis of the environmental impact of the particular end-of-life strategies. On the

contrary, these calculations are crucial for determining if a consumer should continue to use the product or replace it with a newer model. This becomes a question of whether the energy consumption has reduced enough between the model years and if the impact of manufacturing a new product outweighs those gains. Hence, these results relate directly to the end-of-life strategies appropriate for these consumer electronics products. As mentioned in Section II, there are currently three available metrics or parameters to assess a product at the end-of-life: cost, environmental impact and product characteristics. Table 3 compares the recommended end-of-life strategy based on ELSEIM and ELDA. Additional data and analysis is necessary to determine the recommended end-of-life strategies appropriate for the products listed with 'to be decided.' For these products, there are opportunities to improve energy consumption and manufacturing environmental impact. ELSEIM sometimes recommends a lower strategy than allowable by the product characteristics (ELDA). According to ELSEIM (Figure 1), there are environmental gains to move from a lower end-of-life strategy to a higher end-of-life strategy (for example, moving from recycle to remanufacture). In some cases, the gains are so moderate that the required organization and company strategy change may not be warranted (for discussion on realigning the Environmental Value Chain, see [14]). The ELSEIM results shown are preliminary and are based on small subset of electronics products, from one manufacturer. However, it is important to note that for televisions the recommended strategy from both methods is recycling, in contrast to the popular opinion that televisions should be reused.

Table 3. Recommendations for end-of-life strategies

Products Portable CD Player Audio System (FW870) CD-Recordable (FW-R8) CD-Recordable (CDR 785) DECT Cordless Phone Cellular Phone VCR DVD Upmarket Television Mainstream TV LCD Monitor CRT Monitor Fax Product characteristics allow for: Remanufacture Remanufacture Remanufacture Recycle (without disassembly) Remanufacture Remanufacture Recycle (without disassembly) Remanufacture Recycle (with disassembly) Recycle (with disassembly) Remanufacture Remanufacture Remanufacture ELSEIM recommends: Recycle To be decided Recycle To be decided To be decided Reuse To be decided To be decided Recycle Recycle Reuse To be decided To be decided

VIII. CONCLUSIONS This work presents the first calculation of product end-of-life strategy environmental impact. The results reveal that previous perceptions about the importance of end-of-life are true in some cases but not true in all cases. There are more opportunities to focus efforts on energy consumption of the portable CD player rather than on improving the end-of-life environmental impact. As research provides more information on the environmental impact of additional products, the ELSEIM will be stronger. ELSEIM, in combination with product characteristics and cost analysis, will be a valuable tool for determining what is appropriate to do with electronic products at the end-of-life. Continuing research will focus on understanding the basic assumptions of this model and assessing their validity. IX. ACKNOWLEDGMENTS
Funding for the research comes from the Delft University of Technology in the Netherlands. The authors acknowledge Philips Consumer Electronics Environmental Competence Centre for providing access to their extensive case studies.

2. Rose, C.M., K.A. Beiter, and K. Ishii. (1999) 'Determining of End-of-Life Strategies as a part of Product Definition,' in IEEE ISEE. Danvers, MA. 3. Rose, C.M. and K. Ishii (1999) 'Product End-of-Life Strategy Categorization Design Tool,' Journal of Electronics Manufacturing. Vol. 9, No. 1: p. 41-51. 4. Boks, C.B. and A.L.N. Stevels. (2001) 'Ranking Ecodesign Priorities from Quantitative Uncertainty Assessment for End-of-Life Scenarios,' in IEEE ISEE. Denver, CO: 5. Huisman, J., C. Boks, and A.L.N. Stevels. (2000) 'Environmentally Weighted Recycling Quotes Better Justifiable and Scientifically More Correct,' in IEEE ISEE. San Francisco, CA. 6. Huisman, J., C. Boks, and A.L.N. Stevels. (2000) 'Application and Implications of Using Environmentally Weighted Recycling Quotes in Assessing Environmental Effects of the End-of-life of Consumer Electronics,' in Electronics Goes Green. Berlin, Germany. 7. ten Houten, M., de Kok, R. (2000) 'EcoScan 3.0: A Powerful Tool for Ecodesign,' TNO: Delft. 8. Goedkoop, M., M. Demmeers, and M. Collignon. (1996) 'The Eco-indicator 95: Weighting method for environmental effects that damage ecosystems or human health on a European scale.'. Amsterdam: TNO. 39. 9. Reijnen, F. (1999) 'Environmental Benchmark Mainstream Television,' Eindhoven: Philips Consumer Electronics Environmental Competence Centre. 10. Timmers, G. and A.J. Jansen. (2000) 'Human powered energy systems,' in Electronics Goes Green. Berlin, Germany. 11. Stevels, A.L.N. and A.J. Jansen. (1998) 'Renewable energy in portable radios: an environmental benchmarking study,' Journal for Sustainable Product Design, (Issue 4, Jan. 1998). 12. Stevels, A.L.N. (1999) 'Ecodesign for Competitive Advantage,' in Stanford University Instructional Television Network. Stanford, CA. 13. van Knippenberg, P. and H. Meinders. (2000) 'Greening your Business'. Eindhoven, NL: Royal Philips Electronics. 80. 14. Rose, C.M. and A.L.N. Stevels. (2000) 'Applying Environmental Value Chain Analysis to Product Take-Back Systems,' in 7th CIRP Life Cycle Engineering. Tokyo, Japan.

X. REFERENCES 1. Rose, C. (2000) 'Design for Environment: A Method for Formulating Product End-of-Life Strategies.' Stanford University.