Economic and environmental evaluation of design for active disassembly

Journal of Cleaner Production xxx (2016) 1e12

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Economic and environmental evaluation of design for active disassembly Jef R. Peeters a, *, Paul Vanegas a, b, Wim Dewulf a, Joost R. Duflou a a b

KU Leuven, Department of Mechanical Engineering, Celestijnenlaan 300A, Box 2422, 3001, Leuven, Belgium Center for Environmental Studies, University of Cuenca, Campus Quinta Balzay, Av. Victor Manuel Albornoz, Cuenca, Ecuador

a r t i c l e i n f o

a b s t r a c t

Article history: Received 26 January 2016 Received in revised form 10 October 2016 Accepted 11 October 2016 Available online xxx

Prior research has demonstrated that a disassembly based end-of-life (EoL) treatment for electronic products is characterized by the highest recovery rates for precious metals (PMs) and non-commodity plastics, such as flame retardant plastics. Nonetheless, EoL electronic products are nowadays also commonly recycled without disassembly in different types of size-reduction based treatments or in an integrated PM smelter-refinery. This disparity of recycling processes adopted worldwide resulted in a high uncertainty on the EoL treatment processes that will be adopted for discarded electronic products. As a result, governments, original equipment manufacturers and recycling companies struggle to determine the economic and environmental value of design for disassembly. For this reason, a methodology is presented to calculate the Composite Rate of Return (CRR) on investing in design for disassembly and the resulting environmental impacts. This methodology is applied to evaluate the economic and environmental benefits of implementing three types of active fasteners for eleven electronic products which are available in both a product service system (PSS) and a traditional sales oriented business model. The performed analyses demonstrate that the preferred EoL treatment, as well as the economic and environmental benefits of implementing design for active disassembly, strongly depends on several product properties and boundary conditions. Based on the performed sensitivity analysis, the application of active pressure and temperature sensitive fasteners is expected to be only economically viable for products placed on the market in a PSS context, in which they will be separately collected with a high collection rate. Furthermore, impulse sensitive elastomer based fasteners are characterized with the highest rate of return and considered to be suited for both products sold in a traditional sales oriented business model and for products used in a PSS. © 2016 Elsevier Ltd. All rights reserved.

Keywords: Design for disassembly Active disassembly Demanufacturing Ecodesign Recycling Waste electrical and electronic equipment (WEEE)

1. Introduction Developing and emerging economies are expected to legitimately aspire to achieve the welfare of industrialized countries. This is predicted to result in a growth of the global middle class from 1.8 billion people in 2009 to 4.9 billion by 2030 (Pezzini, 2012). With today's consumption patterns this will result in a significant increase of the pressure on resources, as well as an increase in the amount of waste generated and the related environmental burdens. Therefore, a substantial increase in resource-efficiency is essential to guarantee a high quality of life for present and future generations and to achieve sustainable growth (EU Commision, 2011a, b).

* Corresponding author. E-mail address: [email protected] (J.R. Peeters).

For these reasons and because of financial incentives of producer responsibility organizations, customer demands for green products, personal values of managers and designers and legal obligations, original equipment manufacturers (OEMs) are striving to increase their resource efficiency (Dubois and Peeters, 2015). At the same time, new recycling processes have been developed in response to legislative developments, as well the increase in resource prices over the past decades. These developments enable to increase the recovery of, among others, PMs (Ghosh et al., 2015) and plastics (Peeters et al., 2014). However, best available recycling technologies are not consistently deployed worldwide, as a result of significant differences in labor costs, legislative requirements, volumes of waste and access to markets for recyclates (Peeters et al., 2013). At the same time, numerous guidelines have been developed to design product to facilitate disassembly, repair, remanufacturing

http://dx.doi.org/10.1016/j.jclepro.2016.10.043 0959-6526/© 2016 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

2

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

and recycling (Boothroyd, 2002; Telenko et al., 2015). Many of these guidelines have the objective of facilitating the breaking down of products into their individual components for the purpose of repair, refurbishing, remanufacturing cannibalization for component reuse, and/or recycling of the complete product or individual components, also defined as demanufacturing (Duflou et al., 2008). In general, the efficiency of breaking down products into their individual components or demanufacturing can be increased by improving the product architecture and/or the applied fasteners (Harjula, 1996). An improved product structure facilitates most types of demanufacturing processes. However, the selection of the appropriate fasteners is more complex, since most fasteners will only improve a specific demanufacturing process and impede others. For example, the implementation fracture lines developed by Balkenende et al. (2014) facilitate the separation of PWBs from LED lamps in a shredder process, but do not facilitate a repair, refurbishing or remanufacturing process,. Accordingly, good knowledge of the demanufacturing processes that will be adopted is required to determine which of the sometimes contradicting guidelines should be implemented. However, OEMs face today high uncertainty on the demanufacturing processes that will be adopted in consequence of the disparity of recycling processes adopted worldwide. As a result, OEMs also face high uncertainty on the influence and value of the implementation of design for demanufacturing. EoL products that were sold in industrialized countries in a traditional sales oriented business models are commonly collected by joint collection schemes. These collection schemes mostly charge a fixed contribution fee per product. Consequently, the enforcement of legislative requirements and/or the differentiation in contribution fee charged by the collection scheme are required to make producers bear the actual EoL treatment cost. In contrast, OEMs which sell products in a product-service system (PSS) business model and retain the ownership of the product over the complete product lifetime or regain ownership at the EoL stage, have direct financial incentive of implementing design for demanufacturing (Beuren et al., 2013; Reim et al., 2015; Van Ostaeyen et al., 2013). However, clear economic and/or environmental benefits are a prerequisite for both OEMs to invest in Design for demanufacturing and for governments to include design for demanufacturing targets in their legislation or to make the differentiation in contribution fees to collection schemes mandatory. Several methodologies have been presented in recent research to evaluate different recycling processes (Ravi, 2012) and disassembly options (Cheung et al., 2015; Go et al., 2011), as well the economic, environmental and/ or social benefits of design for disassembly (Ma and Okudan Kremer, 2015; Sabaghi et al., 2016; Ziout et al., 2014). However, these methodologies do not take into account that a broad variety of distinct EoL treatments can nowadays be used for electronic products. In addition, these methodologies cannot be used to determine the Rate-of-Return (ROR) of investing in design for demanufacturing. Therefore, a methodology is presented to determine the environmental and economic performances of a variety of commonly adopted EoL treatment options for WEEE and to determine the ROR of investing in design for demanufacturing, such as active disassembly. To apply this methodology, the product's material composition is first analyzed. Thereafter, the efficiency of commonly adopted recycling processes, as well as the efficiency increase that can be obtained by the implementation of design for demanufacturing is assessed. The proposed methodology is demonstrated by means of a case study for eleven electronic products, in which the economic and environmental benefits of design for active disassembly for these products is evaluated. Design for active disassembly requires the

implementation of active fasteners which can be simultaneously unfastened without direct, individual, physical contact between a disassembly tool and every individual fastener (Duflou et al., 2008). The presented case study focus on the evaluation of active fasteners, since prior research predicted that active disassembly has the potential to shift an EoL treatment with systematic disassembly from a cost factor to a profit generating activity (Duflou et al., 2006; Willems et al., 2005). 2. Materials The following electronic products, which are available both in a sales oriented business model and in PSSs, were selected for the presented case study because of their distinct material compositions, product structures and lifetime distributions to allow the investigation of the influence of these product parameters on the economic viability and environmental benefits of implementing design for active disassembly: a Worldline Yomani payment terminal, a Philips 42 inch Econova LED TV (LCD with sided LEDs, model number 42PFL6805H), a Barco Coronis Fusion 10 MP medical monitor (MDCG-10130), an Asus Nexus 7 tablet (2012), an Apple IPad 2 tablet (2011), a Sagem B-Box 2 modem and a Scientific Atlanta V3 setup-box (IPP 430 MC). In addition, an average laptop, LCD monitor and LCD TV is determined in this research by the analysis of a broad range of representative products. In total, 153 EoL LCD TVs with Cold Cathode Fluorescent Lamps (CCFLs) were analyzed in 2011 with an average weight of 13.54 kg (s: 6.5 kg), an average screen size of 30.72 inch (s: 6.3 inch) and an average production year of 2005 (s: 2.2 year), 51 EoL LCD monitors with CCFLs were analyzed in 2014 with an average weight of 4,56 kg (s: 1.46 kg), an average screen size of 17.34 inch (s: 1.68 inch) and an average production year of 2006 (s: 1.8 year) and 32 EoL laptops analyzed in 2015 with an average weight of 2.65 kg (s: 582 g) and an average production year of 2005 (s: 3.3 year). All these caste study products were dismantled in this research and both the weight and material type were registered for the dismantled components. Results of these material composition analyses are shown in Fig. 1. Previous studies have pointed out that plastics are often mismarked (Xiuli et al., 2006). Therefore, the plastic and flame retardants (FRs) were identified in this research by means of a combination of density measurements, sliding-spark spectroscopy, laser-induced breakdown spectrometry (LIBS) and/or Fourier transform infrared (FTIR) analysis. In addition, both the easily accessible high grade printed wiring boards (PWBs) that can be separated in a partial disassembly process and the high grade PWBs that can only be separated in an indepth disassembly process, later on referred to as first level and second level PWBs, were gathered from the case study products. Of these PWBs the PMs and copper concentrations were analyzed by inductively coupled plasma and spark optical emission spectroscopy. To determine the PM concentration for an average LCD TV 18 kg of main board and 3 kg of the PWBs of the LCD module were analyzed from TVs produced between 2004 and 2007. For the PM concentration of an average LCD monitor the results obtained for one Sony SDM-S73a LCD monitor from 2004 were used as representative values. The PM concentration of an average laptop is based on the analysis of the RAM PWB, main PWB with CPU and hard disk PWB of a Dell Inspiron of 2003 and the DVD PWB of a Fujitsu Primergy Server (RX300 S3). To assess the variation in PM concentration between similar laptops of different years, also the main board of an older Dell Latitude from 1998 was analyzed. Since a significantly higher PM concentration was found for the older laptop, two average case study products were used which only differ in the PM concentration of the main board. Net values for the high grade PWBs were provided by an integrated PM smelter-

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

3

Material composiƟon Modem Digicorder Apple Tablet Nexus Tablet Econova LED TV Average LCD TV Laptop lo PM Laptop high PM Medical monitor Average Monitor

Payment terminal 0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Weight%

PWBs high grade

PWBs Low grade + cables

PlasƟcs

Steel

Aluminum

BaƩeries

Magnet

Others

PotenƟal recycling value Modem Digicorder Apple Tablet Nexus Tablet Econova LED TV Average LCD TV Laptop lo PM Laptop high PM Medical monitor Average Monitor Payment terminal 0

2

4

6

8

10

12

Euro/product

First level PWBs ABS No FR PC/ABS No FR Aluminum

Second level PWBs HIPS No FR FR PlasƟcs

Power supply PWBs + cables PMMA Steel

Fig. 1. Material composition and potential recycling value of the different case study products.

refiner based on the determined metal contents and prices of July 2015, as shown in Fig. 3. The recycling values of the cables and low grade PWBs, which do not contain large integrated circuits, are calculated using a fixed value of 500 euro/tonne, which is based on discussions with Belgian recycling facilities. The lifetime distributions analyzed by Wang et al. (2013) are used in the presented case study. No data could be retrieved for the lifetime distribution of payment terminals and medical displays. Therefore, the following values, as estimated by experts of Worldline and Barco, were used for the payment terminals and medical displays respectively: 1% and 2% first year additional failure rate, 20% and 3% constant failure rate and average lifetimes of 7 years and 5 years with a standard deviation of both 1 year. For these products, 90% of the first year additional failures and constant failures are assumed to be repaired.

3. Processes In the presented case study, six distinct EoL treatment scenarios are considered, as shown in Fig. 2. These six scenarios were

considered to be the most realistic EoL-treatment options for the selected case study products by the consulted Belgian recycling companies. The analyzed material separation efficiencies of these EoL treatment options are described in this chapter.

3.1. Integrated smelter-refinery (scenario 1) Integrated PM smelter-refineries process a wide variety of complex waste streams containing precious and other non-ferrous metals. These processes are characterized with an overall recycling efficiency of approximately 95% for gold, silver, palladium and copper (Chancerel et al., 2008, 2009; Meskers et al., 2009). For the most part, integrated PM smelter-refineries do not treat entire devices, but rather PM containing fractions from WEEE that were either separated in a disassembly or size-reduction based process, such as PWBs and connectors. However, in some cases it is economically preferable to treat electronic products directly in an integrated PM smelting-refining plant without pre-processing to increase the overall recycling efficiency for PMs and to reduce processing costs. For example, small

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

4

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

Separation/sorting processes

Liberation/demanufacturing processes

Scenario’s

1

2

3

4

5

6

Complete disassembly Partial disassembly Non-intensive size-reduction

Integrated smelter-refinery

Hand picking Intensive size-reduction Sieve

Incineration (dust)

Blower

Incineration (foils)

Magnet

Steel recycling

Eddy current PWB optical sorting Shaking table

Aluminum recycling Integrated smelter-refinery Copper smelter

Automated Optical sorting Manual Optical sorting

Plastic recycling

Density separation Incineration Fig. 2. Distinct EoL treatment scenarios for electronic products.

Fig. 3. Separation cost for manual disassembly, post-shredder hand picking and automated color sorting and value recovered from PWBs from LCD TVs.

portable electronic products, such as mobile phones, are commonly treated directly after battery removal in an integrated PM smelterrefinery. The drawback of directly processing electronics in an integrated PM smelter-refinery is that plastics and structural metals (aluminium and iron) cannot be recovered as plastics and metals in the process. On the other hand, it should be considered that the plastics are used as an energy source for the smelting process, typically replacing fossil fuels. Of all case study products only the Yomani payment terminal and the B-Box modem, which are plastic rich products, are considered to be suited for direct treatment in an integrated PM smelter-refinery. For the other case study products direct treatment in an integrated PM smelter-refinery is economically unattractive with existing installations, because of technical and operational impacts caused by their high content of structural metals.

To account for the environmental impact of an integrated PM smelter-refinery process the impact generated for the production of secondary PMs as determined for the Boliden plant by Ecoinvent 3.0 is used. When plastics are treated in an integrated PM smelterrefinery both the impact of the incineration process and the fuel offset by the housing plastics in the smelter are accounted for. For this environmental analysis a similar impact as municipal incineration and a reduction of the use of crude oil in proportion to the caloric value of plastics of on average 33 MJ/kg is assumed. 3.2. Disassembly (scenario 2 and 3) The PM concentrations and value of all PWBs were analyzed after manual disassembly. Therefore, this value is expected to be recovered in both a manual and an active disassembly process.

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

Furthermore, all FR plastics are assumed to be recycled after disassembly, since Peeters et al. (2014) have demonstrated that it is technically feasible to recycle FR plastics in a closed loop system after manual disassembly and sorting based on LIBS or FTIR analysis. 3.2.1. Manual disassembly For a manual disassembly process the treatment cost is mainly related to the amount of time required to disassemble a product and the related labor cost. Therefore, the manual disassembly times were analyzed in depth for the different case study products. For the presented average case study products the disassembly times are based on time measurements performed at different recycling facilities in Belgium during the period 2010e2014 with EoL products collected in Belgium. The time measurements were performed with multiple employees with experience in disassembly activities. The disassemblers were allowed to take a break of 15 min after 2 h of work, and 30 min break for lunch. Disassemblers had a pneumatic screw driver and a set of bits at their disposal. Despite the fact that the workers had experience disassembling EEE, a short training was given to them, in which the case study products were disassembled together with the researchers. Following the procedure previously described, the disassembly time of 73 LCD TVs, 48 laptops and 27 monitors was analyzed. The product specific disassembly times of the payment terminal, medical monitor, modem and setup-box are based on different experiments in which the case study products were manually disassembled with the use of electrical screw drivers by the researchers. All experiments were recorded with a camcorder, which allowed a subsequent detailed analysis of the time required for every disassembly step. The disassembly process was split up in: (i) localizing the fasteners and trying to remove the targeted component while it is still attached, and (ii) positioning the tool on the fasteners, disconnecting fasteners and removing the targeted component and placing it in a bin which is located close to the disassembly station. This split up enables to determine the disassembly time for the case study products when either the disassembler has no product knowledge during the recycling of jointly collected products (i þ ii) or good product knowledge during repair operations or recycling of a large number of the same products (ii). The Econova TV was designed for disassembly with easily identifiable screws, which all have the same screw heads and can all be unscrewed in the same direction (Umeda et al., 2012). Therefore, the disassembly time for this case study product is calculated by multiplying the number of screws with the average time of 2.4 s needed per fastener, as analyzed based on the recorded LCD TV disassembly experiments. As for the two tablets, the disassembly times are based on the analysis of on-line disassembly instruction videos, since these products cannot be opened with standard tools (DirectFix, 2012a, b). Of these videos only the disassembly times needed to remove the main PWBs and batteries were measured. 3.2.2. Active disassembly Due to the labor intensity of manual disassembly of electronic products and the high labor cost in industrialized regions, manual disassembly in these regions is generally characterized by a low to negative profitability (Duflou et al., 2008). The most timeconsuming step of a manual disassembly process is the localization of fasteners, which accounts for one to two thirds of the manual disassembly time depending on the product category (Duflou et al., 2006; Peeters et al., 2015c). However, facilitating the visual localization of fasteners is inconsistent with both the trend of miniaturization and the increasing aesthetics requirements for electronic products. Therefore, several generic active fasteners have

5

been developed by the authors and in prior research, as shown in Table 1. The most advanced research on fastening techniques for Active Disassembly has focused on the development of temperature sensitive fasteners. (Chiodo et al., 2002; Nick Jones et al., 2003). Nowadays, the only commercially available temperature sensitive fasteners are tapes containing thermoplastic expandable microspheres (Bain and Manfre, 2006; Kawaguchi, 2004). These microspheres encapsulate a hydrocarbon liquid which gasifies and expands when exposed to heat. The price of the tape is estimated to be 30 euro/m2 or 1.5 euro/TV, 1 euro/monitor, setup-box or laptop and 0.3 V for the other case study products. The amount of energy required for heating up electronic products with 115  C and the related cost are calculated by using the weight and caloric value of the main components and an efficiency of 75%. Furthermore, the cost of 2 employees is accounted for separating the actively disassembled components and the investment and maintenance costs are assumed to be similar to those of a manual disassembly process. Furthermore, pressure sensitive fasteners which are equipped with a cavity that contracts when the ambient air pressure increases were patented in 1990 (Kott, 2003; Pugliesi-Conti and Girardiere, 1990). The embodiments of these fasteners were optimized by means of 2D topology optimization in collaboration with Philips (Willems, 2007a; Willems et al., 2007a, b). More recently, a second generation of pressure sensitive active fasteners, which make use of a closed cell elastomer foam and which are significantly smaller, were developed by the authors (Peeters et al., 2015a). These fasteners have also been demonstrated to be sufficiently robust for the application in electronic equipment (Peeters et al., 2015a). The cost for implementing these fasteners was estimated to approximate 10 eurocents per pressure sensitive fastener or 1.5 euro/TV, 1 euro/monitor, setup box or laptop and 0.3 V for the other case study products. Furthermore, the investment cost for a 1 m3 pressure chamber is assumed to be 11.000 V (Willems, 2007b) and the total investment cost for a pressure triggered disassembly process is assumed to be 40.000 V. Based on the power consumption of a 7 kW compressor the energy cost to increase the air pressure in this chamber with 2 bar is calculated to be 3.99 eurocents/m3 (Willems, 2007b). Since electronic products are in most cases not centralized at EoL and it is not realistic to expect that all recyclers worldwide will invest in dedicated setups to actively disassemble products, also low-cost elastomer fasteners have been developed by the authors. These fasteners can be simultaneously released with standard disassembly tools by applying a sufficiently high force over a limited period of time, typically around 1 s (Van den Bossche et al., 2014). These fasteners have been demonstrated by means of practical experiments to allow reducing the disassembly time by 70%e90% for the housing of LCD TVs, without compromising product robustness (Peeters et al., 2015c). The concept behind these elastomer-based fasteners is that they will not release but only deform to a limited extent during a product drop, since the elastomer will absorb the released energy. When a force is applied over a longer period of time, commonly referred to as an impulse, for example by placing a lever between the attached components, the elastomer-based fasteners will deform and subsequently release (Peeters et al., 2015c). Domain experts estimated an additional production costs of approximately 0.10 V for the implementation of these elastomer-based fasteners for the complete back cover of LCD TVs. Therefore, a cost of 0.3 euro/TV, 0.1 euro/monitor, setup box or laptop and 0.05 V for the other case study products is assumed. Furthermore, taking into account the disassembly time reduction with at least 70%, the operational cost is assumed to be 30% of the manual disassembly cost, since no dedicated setup is required for triggering these impulse sensitive fasteners.

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

6

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

Table 1 Working principles and embodiments of fasteners for active disassembly. Trigger

Temperature

Pressure (Peeters et al., 2015a)

Imulse (Peeters et al., 2015c)

Working principle Example and embodiment

phase change materials

Compression of closed cell foam

Damping by elastomer

1. 2.

3.

Component A

1.

Component A

Component B

Component B

Microspheres

Closed cell foam

Double sided tape

2.

1.

Component A Component B

2.

Snap-fit

Elastomer Screw

3.

Required installation

Oven

For all active disassembly processes the cost of sorting the components after active disassembly is taken into account. The material separation and sorting efficiencies and costs of these active disassembly processes are assumed to be similar to those of a manual disassembly process. Furthermore, the environmental impact of the implementation of these fasteners in a (I)TV, (II) monitor, setup box or laptop and (III) other case study products is calculated to be I: 3.8, II: 2.5, III: 0.8 mPt for temperature sensitive fasteners based on the impact of the production of adhesives for metal, I:11.7, II:7.8, III: 2.3 mPt for pressure sensitive fasteners based on the impact of polystyrene production, and I:13.0, II:4.3, III:2.2 mPt for impulse sensitive fasteners for the production of polybutadiene rubber. 3.3. Size-reduction (scenario 4,5 and 6) The first step in an automated size-reduction based EoL treatment for electronic products is to liberate the different materials allowing subsequent material separation. For this purpose a shredder process is commonly adopted, which is in some cases preceded by a smashing operation to open products for the purpose of depollution, for example for battery removal, or to separate valuable components, such as PWBs. Post-shredder, the ferrous metals are commonly separated by a magnet and the non-ferrous metals are typically separated with an eddy current separator. The efficiency of these separation processes is known to depend on the product architecture, material composition and adopted fasteners. However, due to the complexity to analyze effect of product parameters on the post-shredder steel and aluminum separation efficiency an average yield of 99% for steel and 83% for aluminum is used in the present research (Peeters et al., 2013). Thereafter, different technologies can be used to separate the plastics based on plastic type for the purpose of recycling and for the separation of cables and PWBs for further processing by either a copper smelter for copper recovery or an integrated PM smelterrefinery for PM and copper recovery. In the presented case studies three distinct scenarios involving size-reduction are considered. In the standard shredder based recycling scenario (6), only a water-shaking table and a sink-float process are adopted post-shredder. With a sink-float process only the non-FR plastics HIPS and ABS can be separated from the case study products for recycling with an assumed efficiency of 90%. Other plastics, such as thermoplastic Poly(methyl methacrylate) (PMMA) and FR plastics cannot be separated with this process, since prior research has demonstrated that these plastics have overlapping densities (Peeters et al., 2014). In the advanced shredder based recycling scenario (5) also an automated optical sorter is adopted to separate the different types of plastics. Prior research has demonstrated that Near InfraRed (NIR) sorters are characterized by a low separation efficiency for

Pressure room

Manual disassembly station

plastics from electronics due to identification problems for black plastics (Peeters et al., 2014). As a result, only the transparent PMMA, which is commonly used in monitor and LED TV LCD modules, can be separated with the automated NIR sorter. To evaluate the separation efficiency of these optical sorters for PMMA, an experiment was set up in which a plastic mix of 6.46 kg with a concentration of 54.6% PMMA originating from LCD TVs and monitors was separated by an optical sorter equipped with both a NIR and color camera. By means of manually sorting the separated fractions a yield of 96% and a purity of the PMMA of 97% were concluded. These efficiencies are used in Scenario 5, whereas the effective efficiency is expected to be lower due to the significantly lower concentration of PMMA in most case study products. Furthermore, to evaluate the efficiency of post-shredder separation of PWBs two experiments were set up with in total 1480 kg and 418 kg respectively of only LCD TVs, which were processed in a knife shredder with a 50 mm sieve. Afterwards, the mercury was removed and the steel was separated with a magnet from the postshredder residue. In the first experiment, 11.4 and 7.4 kg of PWBs were separated by means of in-line hand picking by two consecutive operators. In the second experiment, the non-ferrous material was separated by means of an eddy current separator. Afterwards, the PWBs were either separated from the non-ferrous fraction of the eddy current based on the green color of the PWBs (5.28 kg of PWBs) or from the plastic fraction of the eddy current separator based on induction (18.7 kg of PWBs). The PWBs separated by every operator and by the automated sorters were weighed and analyzed by inductively coupled plasma and spark optical emission spectroscopy to determine their PM and Cu concentration. Based hereon, the recovered value was calculated, shown in blue at the right side in Fig. 3, and compared with the separation and sorting costs including depreciation costs, which are marked in red at the left side of Fig. 3. Thereafter, the profit per tonne of processed LCD TVs and the share of the value that can be recovered is compared with manually disassembling the PWBs. The results of these analyses indicated that with a combination of automated color and induction sorting approximately 19% of the value can be recovered during pre-processing from the total amount of incoming PWBs. Since a combination of color and induction sorting is characterized by the highest recovery efficiency, the use of this set up is assumed in Scenario 5. In the smasher based scenario (4) the PWBs are assumed to be separated by means of hand picking by 4 employees directly after breaking open the product in a chain smasher or in a rotating drum installation. Since intensive size-reduction is needed to assure that all mercury contained in the CCFLs can be washed or evaporated, it is not possible to treat products with CCFLs backlighting in this scenario. The overall separation efficiency for PMs of this scenario strongly depends on the product design and the ease at which the product can be opened by smashing and the extent to which the

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

PWB can be removed. In the presented case study this efficiency is based on the upper bound of 70% reported by prior research [16, 21]. It should be noted that in some cases PMs are also recovered from shredder dust and incineration ashes. Therefore, future research is required to determine the overall recycling efficiency of PMs from electronic products including these material feedback loops. 4. Methodology To correctly evaluate the economic and environmental benefits of implementing design for demanufacturing, it is essential to take into account both the costs and generated environmental impacts of the implementation of design changes, as well as the cost and impact reductions attained during repair, refurbishing, remanufacturing, cannibalization and recycling. For this reason, it is essential to first analyze the products’ material composition, as shown in Fig. 1, and the efficiency of commonly adopted recycling processes, as shown in Fig. 3. When this data are available, both the economic and environmental benefits of implementing design for demanufacturing can be determined with the methodology described in this chapter. 4.1. Lifetime distribution Due to the fact that electronic products are discarded for different reasons, such as production errors, accidents, wear or technological obsolescence, the product lifetime of different types or categories of electronic products can significantly vary. Since early returning products will result in a faster return on investment in design for active disassembly and, accordingly, a higher rate on return, it would be wrong to only take into account the average lifetime. Therefore, a distribution which represent the failure probability in function of the product lifetime is used in the presented methodology. Different probability functions, such as normal and lognormal, can be used to represent the probability FðxÞ that a product will be discarded in the year x: However, prior research has demonstrated that the Weibull function produces the best fit of the lifespan distribution for CRT TVs (Walk, 2009). Therefore, in the presented methodology the failure probability is, in case this data are available for the shape parameter a and the
k scale parameter b, calculated using a Weibull distribution (Pola
palova
, 2012; Van Schaik and Reuter, 2004). Furthermore, and Dra it is assumed that the successful repair rate (RðxÞ ) in the first and second year is 90%. With this approach, the probability FðxÞ that a product will be discarded in year x after its manufacture is defined by Equation (1)

( FðxÞ ¼

1

i¼X1 X i¼1

)  a  x a1   FðiÞ 1  RðxÞ *e *

b b

 a x

b

(1)

Collecting quantitative data to determine the shape parameter a and scale parameter b is resource intensive. Therefore, the authors propose Equation (2) to determine the failure probability when limited data are available, which is the case for the payment terminal and medical monitor. Equation (2) is based on the theory of the bathtub curve used in reliability engineering (Kapur and Pecht, 2014). With this equation the probability FðxÞ that a product will return in year x is calculated by adding up the probability of the following three distinct reasons why a product would return for repair, remanufacturing or recycling: Early failures (E) during the first year due to production and installation errors, constant accidental failures (C) due to external parameters and average lifetime (A) due to wear and technological obsolescence following a normal

7

distribution with a standard deviation (s). Furthermore, it should be considered that a share of the discarded products can be successfully repaired in year x (RðxÞ ).

( FðxÞ ¼

1

i¼X1 X

)     ðif x ¼ 1=EÞ þ C FðiÞ 1  RðiÞ

i¼1 ðxAÞ2 1 þ pffiffiffiffiffiffie 2s2 s 2p

 (2)

4.2. Economic analysis In the presented methodology, the Composite Rate of Return (CRR) is used to determine the economic viability of design for demanufacturing. The CRR, also referred to as external rate of return, is the unique rate of return for an investment that assumes that net positive cash flows are reinvested at a predefined reinvestment rate (Blank and Taraquin, 2005). CRR analysis is used instead of Internal Rate of Return (IRR) analysis, because it is not realistic to assume that the released funds are reinvested at the obtained IRR in case that the IRR substantially differs from the Minimum Attractive Rate of Return (MARR). It should be noted that in the presented case study only the economic viability of implementing design for active disassembly is compared with the CRR to doing nothing. To correctly compare different design for demanufacturing options an incremental rate of return analysis for mutually exclusive alternatives is required (Blank and Taraquin, 2005). The CRR is calculated for the 11 case study products and three types of active fasteners taking into account a reinvestment rate of 10%. For these calculation, the required investment costs for implementing the active fasteners in year 0 (in red) and the attained cost reductions by facilitating repair (in green) and the EoL treatment (in blue) multiplied with the probability of failure at year x are determined, as shown in Fig. 4. To determine the economic benefits of implementing design for active disassembly for material recycling, the treatment costs and material revenues of the different EoL treatment options is first quantified. The results of these calculations are then used to rank the EoL treatment options based on their profitability (PROF). Based on this ranking the potential increase in profitability of the EoL treatment as a result of implementing design for active disassembly is calculated. For this economic evaluation the recyclates values, material processing costs, and pre-treatment and end-treatment recycling efficiencies shown in Table 2 are used. To determine the recovered material value the recyclates value is multiplied with the pre-treatment recycling efficiency. It should be noted that this calculation includes several simplifications and, therefore, does not take into account that the separation efficiency depends on the input composition and that the recyclates value depends on the composition of the output fraction and the compatibility for recycling of the materials present in this output fraction. In the presented case studies the current value of mixed color flakes is used for plastics which are commonly recycled (Plastic News, 2013). Since FR plastics are nowadays rarely recycled in a closed loop system on an industrial scale, the value of separated FR plastic components is for the case study assumed to be approximately 60% of the virgin value, which varies for virgin PC-ABS between 2930 and 3010 V/tonne and of HIPS-PPE between 2130 and 2148 V/tonne depending on the sold volume and the required additives, such as FR (PIE, 2012). This is a realistic assumption, since FR plastic are more valuable than non-FR plastics and because it has been demonstrated that, after cleaning and regranulation, FR plastic recyclates can be used to substitute virgin material (PIE,

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

8

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

Fig. 4. Required investment for pressure sensitive fasteners in a Yomani payment terminal and cost reductions during repair and EoL treatment during the product lifetime.

Table 2 Material values of recyclates (þ), material treatment cost (), avoided () and generated (þ) environmental impact of material recycling, recycling efficiencies and processing costs of pre-treatment (material liberation and separation) and recycling efficiency of end-processing (material recycling) (others include all materials that are not considered for recycling). Recyclate value Recycling impact 1. PM smelter 2. In-depth 3. Partial disassembly disassembly (euro/tonne) (Eco-Pt/tonne) refinery (yield %) (yield %) (yield %) ABS No FR 255 350 HIPS No FR 295 350 PC/ABS No FR 1078 625 PMMA 509 630 FR Plastics 1540 625 Steel 211 236 Aluminum 1188 1,510 Batteries 0 131 High grade PWBs PWB dependent Low grade PWBs þ cables 500 9,570 Others 150 175 Pre-treatment cost average (euro/tonne excl. disassembly) Size reduction based treatment impact (Eco-pt/tonne)

0% 0% 0% 0% 0% 0% 0% 100% 100% 100% 0%

100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 47 0.04

2012). Furthermore, the European steel and aluminum values of shredded scrap are used (2015). The treatment cost of the different scenarios is based on the investment costs reported in prior research (Cryan et al., 2010) and recent discussions with industrial partners in Belgium, a throughput of 2 tonnes/h, 1750 production hours/year, a renting cost of 45 euro/m2/month, an energy cost of 0,15 euro/kWh, an air pressure cost of 0,04 euro/m3, a labor cost including overhead of 27,5 euro/hour and a depreciation time of 7 years for all recycling infrastructure. Furthermore, a plastic identification cost of 0.18 V per identification is used. This cost includes labor cost, operational cost such as maintenance, and the depreciation cost of the required analysis equipment. In the presented economic calculations the cost of recycling batteries or CCFLs are not taken into account, since batteries and CCFLs have to be separately recycled in every EoL treatment scenario to comply with EU legislation.

100% 100% 100% 0% 100% 96% 100% 100% 100% 80% 100% 110 0.84





  

90% 90% 0% 96% 0%

0%

4. Smasher 5. Shredder þ 6. Shredder End processing (yield %) (yield %) (yield %) (no CCFL) (yield %) 90% 90% 0% 96% 0% 99% 83% 100% 70% 70% 100% 215 7.22

90% 90% 0% 96% 0% 99% 83% 100% 19% 50% 100% 205 19.52

90% 90% 0% 0% 0% 99% 83% 100% 0% 50% 100% 102 0.83

90% 90% 90% 90% 90% 100% 79% excluded >98% 100% incineration

recycling processes is accounted for, since other environmental impacts related to the pre-processing can be neglected. The incineration with energy recovery of different plastic types is assumed to have the same impact as incinerating an average mix of plastics of consumer electronics. For plastic recycling only the impact of energy consumption equal to 0.6 KWh/kg and the avoided production of virgin plastic are accounted for. An end-processing yield for steel of 100%, for aluminium of 79% and for copper of 100% A copper concentration of low grade PWBs and cables of 25 wt% (Bigum and Kai-Sorens Brogaard, 2009). For materials other than metals, plastics, PWBs or batteries, the impact of treating residue from shredder fraction in a municipal waste incinerator is accounted for.

5. Results 4.3. Environmental evaluation In the presented case studies the environmental performance of the different recycling scenarios is calculated using only single score end-points determined with the Life Cycle Assessment database Ecoinvent 3 and the ReCiPE H/A Europe method, as shown in the third column of Table 2. These single scores are multiplied with both the pre-processing and end-processing yield to determine the actual avoided environmental impact. For these calculations the following assumptions are made:  Since the market for metals is unsaturated, metal recycling is fully accounted for as avoiding virgin metal production.  Except for direct treatment in an integrated PM smelter-refiner, only the environmental impact of energy consumption of the

The results presented in Table 3 show that the direct treatment in a PM smelter-refinery scenario (1) is only characterized by a high profitability for the Yomani terminal, which has a high PM content, a low content of steel and iron, a high disassembly time and a low concentration of plastics that can be recycled in a size reduction based treatment. The partial or in-depth disassembly scenarios (2 & 3) are the most profitable for the medical monitor, laptops and TVs, which are characterized by a relatively low disassembly time in proportion to the recycling revenue of the PWBs. The smasher scenario (4) results in the highest profit for the Yomani terminal, tablets and setup-box, which are characterized by high disassembly times in proportion to the recycling value of the PWBs. The advanced shredder scenario (5) results in the highest profit for monitors, which have PWBs with a low economic value and with a

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

high concentration of plastics which can be separated after sizereduction. The standard shredder based recycling scenario (6) is for none of the case study products the preferred EoL treatment. To graphically represent the CRR in function of the environmental impact of implementing design for demanufacturing, the by Huisman developed eco-efficiency graph (Huisman, 2003) is used, as shown in Fig. 5. However, instead of the profit of the EoL treatment the CRR with a reinvestment rate of 10% for the investment in design for active disassembly is plotted on the Y-axis. The results shown in Fig. 5 demonstrate that the environmental impact and the CRR on investment in design for demanufacturing strongly dependents on both the product properties and type of active fasteners. Nonetheless, the following three clusters can be identified: The first cluster encompasses all products in the right top quadrant, also referred to as a positive eco-efficiency realized, for which the implementation of design for active disassembly is characterized by a high CRR and a high avoided environmental impact. For all products in this first cluster manual disassembly was, before the implementation of design for demanufacturing, not the most profitable EoL treatment due to high product complexity and, accordingly, high disassembly time. After the implementation of active fasteners a disassembly based treatment becomes substantially more profitable, resulting in a high return on investment for the products in this firs cluster. Since significantly more materials can be recovered from these products by the adoption of a disassembly based EoL treatment, also substantial environmental impact can be avoided by the implementation of design for demanufacturing for these products. The second cluster contains all products in the right bottom quadrant for which the implementation of design for active disassembly is characterized by a CRR below the reinvestment rate, which indicates that the avoided disassembly costs and higher material revenues do not compensate for the cost of implementing active fasteners. However, by the adoption of a disassembly based EoL treatment more materials can be recovered from these products, which results in avoided environmental impact. Therefore, the avoided environmental impact in proportion to the CRR (a) should be determined, as shown in Fig. 5, to evaluate the extent to which financial incentives result in sufficient environmental improvement. This evaluation can, for example, be used by governmental organizations to prioritize the implementation of Ecodesign

9

requirements. The third cluster contains all products close to the y-axis for which the implementation of design for active disassembly does not significantly influences the environmental performances, since manual disassembly was for these products the most profitable EoL treatment before the implementation of active fasteners. The CRR of implementing design for active disassembly for these products strongly depends on the disassembly cost that can be avoided, the product lifetime and the cost of implementing active fasteners. Therefore, the added value of the implementation of active fasteners in the Econova TV, which has already been designed for disassembly, is significantly lower compared with an average LCD TV. However, the CRR of the likely higher production costs of the Econova TV, for example for the simplification of the product structure, is not evaluated in this case study. For the products in this cluster, the implementation of design for active disassembly results in a neglect able environmental impact, since the impact generated during the active disassembly process and for the implementation of the active fasteners is not compensated by a higher material recovery during the EoL treatment, as shown in Table 3. The presented results also demonstrate that the cost of manual disassembly is in general rather small. For example, for LCD TVs and laptops, for which manual disassembly is the preferred EoL treatment, the labor cost is 3,38 and 2,96 V respectively. For this reason and due to the relatively long lifetime of the case study products, the elastomer based fasteners, which have the lowest implementation cost, are characterized with the highest CRR for all case study products. However, it should be noted that these active fasteners might not be suited for products for which a high security level is needed, such as payment terminals, which should not be easily opened by users to prevent fraud. Furthermore, the CRR for pressure and temperature sensitive fasteners, which are assumed to have a similar investment cost, is nearly the same. This small difference also shows that the energy intensity of the active disassembly process, which is substantially higher for a temperature triggered active disassembly process, has only a limited effect on the economic viability. For the performed environmental analyses it should be noted that the main impact categories which are avoided by recycling electronic products are: fossil fuel depletion, climate change, particular matter formation and human toxicity. The impact of

Table 3 Results of the application of the presented methodology for the different case study products (the scenarios with the highest profit are marked in blue, negative values are losses or avoided environmental impacts and positive values are profits or generated environmental impacts).

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

10

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

Fig. 5. Eco-efficiency graph presenting the Composite Rate of Return (CRR) for the investment in design for demanufacturing in function of the environmental impact of implementing design for demanufacturing and the increased recycling efficiency by adopting a disassembly based EoL treatment.

fossil depletion and climate change is reduced by: 1) increasing the recycling of plastics, which avoids the impact of plastic production and incineration, and 2) by lowering the need for virgin metals and the related energy intensive mining and refining processes. The impact of human toxicity is mainly reduced by increasing the recovery of PMs and copper for which significant amounts of acids are used during the refining process. These analyses also indicate that the recycling of PWBs and the herein contained PMs and copper contributes most to avoiding environmental impact. 6. Discussion It should be stressed that for the evaluation of the overall profitability of investing in design for demanufacturing that also product (re-)design and other indirect costs related to the implementation of design for demanufacturing should be taken into account. This is important to determine the extent to which (re-) design and indirect costs can be covered by facilitating the demanufacturing, which depends on the total number of similar products that are sold. Accordingly, when the CRR is similar, the overall economic benefits of investing in design for demanufacturing will be higher for the Yomani terminal of which more than 100.000 products are produced annually than for the medical monitors of which less than 5.000 are manufactured yearly. In addition, it should be considered that the decision of OEMs or customers to repair, refurbish, remanufacture or cannibalize an electronic product depends on many factors, including the ease at which a product can be disassembled. However, the relation between the product design and the likelihood that the lifetime of a product will be extended is difficult to assess and out of the scope of the presented research. Therefore, the economic and environmental benefits of design for demanufacturing for the purpose of repair, refurbishing or remanufacturing are not taken into account in the presented case study. When interpreting the presented results it should also be considered that recycling process parameters, product parameters and other boundary conditions, often depend on the considered region and can significantly fluctuate over time. For example,

Peeters et al. (2015b) have demonstrated that the housing plastic composition of EOL electronic displays will significantly evolve over time and that this will result in a decrease in recycling efficiency of size-reduction based processes for housings of electronic displays from 43% in 2005 to only 28% in 2025 (Peeters et al., 2015b) (Peeters et al., 2015b) (Peeters et al., 2015b). Furthermore, the following factors were identified to significantly influences the value of design for demanufacturing: 1.) The repair rate, since the implementation of design for demanufacturing will result in a faster and multiple returns on investment for products which are frequently repaired. 2.) The product lifetime, as the CRR on investments in design for demanufacturing is negatively related with the average product lifetime, since a longer product lifetime results in a later return on investment. 3.) The labor cost, since the economic benefits of design for demanufacturing are mainly obtained by reducing the labor intensity of repair, remanufacturing, cannibalization and recycling processes, while increasing the material recovery efficiency at EoL. 4.) The collection rate at EoL, since investments in design for demanufacturing only make sense from an economic perspective when the products will be demanufactured by the company or society that invested in the improved product design, and 5.) The product knowledge at EoL, since disassembly operations are more cost efficient with good product knowledge, which results in lower potential economic benefits of implementing design for active disassembly. For example, for the Yomani payment terminal, the medical monitor, the V3 setup-box and the B-Box modem, good product knowledge, which avoids the time required to localize and identify the type of fasteners, is evaluated to result in a decrease in disassembly time of up to 68%. This increase in disassembly efficiency is evaluated to results in a substantial decrease of the CRR on investments in active fasteners, but for the Yomani payment terminal and the medical monitor the availability of product knowledge will not influence the CRR, since even with the availability of product knowledge a disassembly based EoL treatment would not be the preferred EoL treatment. Therefore, it can be concluded that design for demanufacturing is also economically viable for complex products containing high value PWBs, which are sold in a PSS.

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

11

7. Conclusions and further work

Acknowledgements

A methodology is presented to enable OEMs and governments to correctly quantify the economic and environmental benefits of implementing design for demanufacturing, such as design for active disassembly. The presented methodology is applied to evaluate the economic viability of implementing active fasteners in eleven electronic products which are both sold in sales oriented business models and in PSSs, with distinct material compositions, product structures and lifetime distributions. Results of the presented case study demonstrate that the preferred EoL treatments, as well as the economic and environmental benefits of implementing design for active disassembly, are strongly product dependent. The CRR on investment in active fasteners is highest for products with the following properties:

The authors acknowledge the Agency for Innovation by Science and Technology in Flanders (IWT) and the Flemish Environmental Technology Platform (MIP) for funding this research (project numbers 100502 and 111568), and thank the companies TP Vision, Barco, Galloo, Umicore, Van Gansewinkel, Worldline and SIMS recycling for their support and cooperation.

 High collection rate of EoL products  High product complexity and little product knowledge at EoL, resulting in a high disassembly time  Low cost for the implementation of design for demanufacturing  High concentration of PMs  High concentration of plastics which cannot be separated in a size-reduction based EoL treatment, such as FR plastics.  High failing rate and/or short average lifetime Based on the performed analysis, the application of pressure and temperature sensitive fasteners is expected to be economically viable for products placed on the market in a PSS and which will be separately collected with a high collection rate for the purpose of repair, refurbishing, remanufacturing, cannibalization or recycling. Furthermore, it is demonstrated that investments in impulse sensitive fasteners are characterized with the highest CRR, which is mainly due to the low fastener cost. The implementation of impulse sensitive fasteners is considered to be suited for products sold in a traditional sales oriented business and used in a PSS. In addition, the implementation of these fasteners is demonstrated to result for several product categories in a substantial reduction in environmental impact. However, OEMs which sell their products in a traditional sales oriented business model often do not benefit of the implementation of these fasteners, since EoL products are commonly collected by joint collection schemes, which mostly charge a fixed contribution fee per product. Therefore, future research should focus on the development of systems to provide adequate stimuli for OEMs to implement design for demanufacturing. To further assist OEMs to improve their product design and recyclers with improving their processes, future research should focus on better quantifying the effect of the evolution in material composition and product structure on the recycling efficiency of both existing and new recycling technologies. For example, the effect of the adoption of manual disassembly and plastic sorting based on optical identification techniques should be further studied. Furthermore, it is demonstrated that both profitability and resource efficiency can be increased by differentiating the type of EoL treatment depending on product characteristics. For this reason, both guidelines and techniques should be developed to support recyclers to optimize their processes by enabling a better differentiation of the EoL treatment. Finally, it is also demonstrated that design for demanufacturing for products with limited material value, but of which many will reach their EoL in the coming years, such as tablets, is not economically viable. Therefore, the development of new separation, sorting or refining technologies will be required to increase the material recovery from such products in an economically viable manner.

References Bain, P., Manfre, G., 2006. Method and Apparatus for Bonding and Debonding Adhesive Interface Surfaces, US. Balkenende, R., Ochionorelli, V., van Meensel, W., Felix, J., Sjolin, S., Aerts, M., 2014. Greenelec: product design linked to recycling. In: Going Green e CARE Innovation, Vienna, Austria. Beuren, F.H., Gomes Ferreira, M.G., Cauchick Miguel, P.A., 2013. Product-service systems: a literature review on integrated products and services. J. Clean. Prod. 47, 222e231. Bigum, M., Kai-Sorens Brogaard, L., 2009. LCA Modelling of Metal Recovery from WEEE DTU Environment. Technical University of Denmark, Denmark. Blank, L., Taraquin, A., 2005. Engineering Economy, seventh ed. Boothroyd, 2002. Product Design for Manufacture and Assembly, second ed. (Revised and Expanded). Chancerel, P., Meskers, C.E.M., Hageluken, C., Rotter, S., 2008. E-Scrap Metals Too Precious to Ignore, pp. 42e45. Recycling international November. Chancerel, P., Meskers, C.E.M., Hageluken, C., Rotter, V.S., 2009. Assessment of precious metal flows during preprocessing of waste electrical and electronic equipment. J. Ind. Ecol. 13, 791e810. Cheung, W.M., Marsh, R., Griffin, P.W., Newnes, L.B., Mileham, A.R., Lanham, J.D., 2015. Towards cleaner production: a roadmap for predicting product end-of-life costs at early design concept. J. Clean. Prod. 87, 431e441. Chiodo, J.D., Jones, N., Billett, E.H., Harrison, D.J., 2002. Shape memory alloy actuators for active disassembly using 'smart' materials of consumer electronic products. Mater Des. 23, 471e478. Cryan, J., Freegard, K., Morrish, L., Myles, N., 2010. Demonstration of Flat Panel Display Recycling Technologies (WRAP Project.). DirectFix, 2012a. Google Nexus 7 Tablet Repari Directions. DirectFix, 2012b. How to IPad 2 Screen Digitizer Replacement. Dubois, M., Peeters, J.R., 2015. Design for environment incentives from extended producer responsibility. In: OECD (Ed.), OECD Guidance Manual on Resource Productivity and Waste. Duflou, J.R., Willems, B., Dewulf, W., 2006. Towards self-disassembling products design solutions for economically feasible large-scale disassembly. Innov. Life Cycle Eng. Sustain. Dev. 87e110. Duflou, J.R., Seliger, G., Kara, S., Umeda, Y., Ometto, A., Willems, B., 2008. Efficiency and feasibility of product disassembly: a case-based study. CIRP Ann. Manuf. Technol. 57, 583e600. EU Commision, 2011a. Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, a Resource-efficient Europe e Flagship Initiative under the Europe 2020 Strategy. EU Commision, 2011b. A Resource-efficient Europe e Flagship Initiative under the Europe 2020 Strategy. Ghosh, B., Ghosh, M.K., Parhi, P., Mukherjee, P.S., Mishra, B.K., 2015. Waste printed circuit boards recycling: an extensive assessment of current status. J. Clean. Prod. 94, 5e19. Go, T.F., Wahab, D.A., Rahman, M.N.A., Ramli, R., Azhari, C.H., 2011. Disassemblability of end-of-life vehicle: a critical review of evaluation methods. J. Clean. Prod. 19, 1536e1546. Harjula, T., 1996. Design for disassembly and the environment. CIRP Ann. 45, 109. Huisman, J., 2003. The QWERTY/EE Concept, Quantifying Recyclability and Ecoefficiency for End-of-life Treatment of Consumer Electronic Products, Design for Sustainability Program. Delft University of Technology, Delft, Netherlands. Kapur, K.C., Pecht, M., 2014. Reliability Engineering. John Wiley & Sons. Kawaguchi, Y., 2004. Synthesis and properties of thermoplastic expandable microspheres: the relation between crosslinking density and expandable property. J. Appl. Polym. Sci. 93, 505. Kott, N., 2003. Fastener System. Ma, J., Okudan Kremer, G.E., 2015. A fuzzy logic-based approach to determine product component end-of-life option from the views of sustainability and designer's perception. J. Clean. Prod. 108 (Part A), 289e300. Meskers, C., Hagelüken, C., Salhofer, S., 2009. Impact of pre-processing routes on precious metal recovery from PCs. In: European Metallurgical Conference (EMC) Innsbruck, Austria, pp. 527e540. Nick Jones, D.H., Hussein, Habib, Billett, Eric, Chiodo, Joseph, 2003. Towards selfdisassembling vehicles. J. Sustain. Prod. Des. 3, 59e74. Peeters, J.R., Vanegas, P., Duflou, J.R., Mizunoc, T., Fukushigec, S., Umeda, Y., 2013. Effects of boundary conditions on the end-of-life treatment of LCD TVs. CIRP Ann. Manuf. Technol. 62, 35e38. Peeters, J.R., Vanegas, P., Tange, L., Van Houwelingen, J., Duflou, J.R., 2014. Closed

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043

12

J.R. Peeters et al. / Journal of Cleaner Production xxx (2016) 1e12

loop recycling of plastics containing flame retardants. Resour. Conserv. Recycl. 84, 35e43. Peeters, J., Van den Bossche, W., Devoldere, T., Vanegas, P., Dewulf, W., Duflou, J., 2015a. Pressure-sensitive fasteners for active disassembly. Int. J. Adv. Manuf. Technol. 1e11. Peeters, J.R., Vanegas, P., Kellens, K., Dewulf, W., Duflou, J.R., 2015b. Forecasting Plastic Waste Compositions: a Case Study for Plastic Waste of Electronic Display Housings (Accepted for publication in the Journal of Waste Management). Peeters, J.R., Vanegas, P., Van den Bossche, W., Devoldere, T., Dewulf, W., Duflou, J.R., 2015c. Elastomer-based fastener development to facilitate rapid disassembly for consumer products. J. Clean. Prod. 94, 177e186. Pezzini, M., 2012. An Emerging Middle Class. OECD Yearbook. PIE, 2012. Plastic Information Europe. Plastic News, 2013. Resin Pricing - Commodity Thermoplastics.
k, M., Dra
palova
, L., 2012. Estimation of end of life mobile phones generation: Pola the case study of the Czech Republic. Waste Manag. 32, 1583e1591. Pugliesi-Conti, J., Girardiere, C., 1990. Snap-fit Device for Hoining Two Parts Together. Ravi, V., 2012. Evaluating overall quality of recycling of e-waste from end-of-life computers. J. Clean. Prod. 20, 145e151. € Reim, W., Parida, V., Ortqvist, D., 2015. ProducteService Systems (PSS) business models and tactics e a systematic literature review. J. Clean. Prod. 97, 61e75. Sabaghi, M., Mascle, C., Baptiste, P., 2016. Evaluation of products at design phase for an efficient disassembly at end-of-life. J. Clean. Prod. 116, 177e186. Resin pricing e Commodity Thermoplastics. 2015. Scrap monster. Telenko, C., O'Rourke, J., Webber, M.E., Seepersad, C.C., 2015. A compilation of design for environment guidelines. J. Mech. Des. 289e301. Umeda, Y., Takata, S., Kimura, F., Tomiyama, T., Sutherland, J.W., Kara, S., Herrmann, C., Duflou, J.R., 2012. Toward integrated product and process life cycle planningdAn environmental perspective. CIRP Ann. Manuf. Technol. 61, 681e702. Van den Bossche, W., Peeters, J.R., Devolder, T., Duflou, J., Dewulf, W., 2014. Proof of concept of an elastomer based fastener enabling rapid disassembly. In: Lien, T.K.

(Ed.), To be Published in 21st CIRP Conference on Life Cycle Engineering. Elsevier, Trondheim, Norway. Van Ostaeyen, J., Van Horenbeek, A., Pintelon, L., Duflou, J.R., 2013. A refined typology of producteservice systems based on functional hierarchy modeling. J. Clean. Prod. 51, 261e276. Van Schaik, A., Reuter, M., 2004. The time-varying factors influencing the recycling rate of products. Resour. Conserv. Recycl. 40, 301e328. Walk, W., 2009. Forecasting quantities of disused household CRT appliances e a regional case study approach and its application to Baden-Württemberg. Waste Manag. 29, 945e951.
, C.P., 2013. Enhancing e-waste estimates: Wang, F., Huisman, J., Stevels, A., Balde improving data quality by multivariate inputeoutput analysis. Waste Manag. 33, 2397e2407. Willems, B., 2007a. Development of Improved Fastening Techniques in Support of Design for Disassembly Strategies. Centre for Industrial Management. Katholieke Universiteit Leuven, Leuven. Willems, B., 2007b. Development of Improved Fastening Techniques in Support of Design for Disassembly Strategies. Doctoral Dissertation. Faculteit Ingenieurswetenschappen, KULeuven, Leuven, ISBN 978-90-5682-822-6. Willems, B., Dewulf, W., Duflou, J., 2005. Design for active disassembly (DfAD) e an outline for future research. In: 2005 Ieee International Symposium on Electronics & the Environment, Conference Record, 129e134 262. Willems, B., Dewulf, W., Duflou, J.R., 2007a. Active snap-fit development using topology optimization. Int. J. Prod. Res. 45, 4163e4187. Willems, B., Dewulf, W., Duflou, J.R., 2007b. Pressure-triggered active fasteners: design results using topology optimization. In: Proceedings of the 2007 Ieee International Symposium on Electronics & the Environment, Conference Record, 184e189 262. Xiuli, Q., Williams, J.A.S., Grant, E.R., 2006. Viable plastics recycling from end-of-life electronics. Electron. Packag. Manuf. IEEE Trans. 29, 25e31. Ziout, A., Azab, A., Atwan, M., 2014. A holistic approach for decision on selection of end-of-life products recovery options. J. Clean. Prod. 65, 497e516.

Please cite this article in press as: Peeters, J.R., et al., Economic and environmental evaluation of design for active disassembly, Journal of Cleaner Production (2016), http://dx.doi.org/10.1016/j.jclepro.2016.10.043