Ewaste Opportunities

E-waste:

an opportunity by Matthew J. Realff*, Michele Raymond§, and Jane C. Ammons†

The success of the electronics industry over the last decade in developing a mass consumer market for computers, cell phones, and other personal electronic equipment has been phenomenal. Society must now finds ways of safely and economically recovering the materials that are embedded in these products. This will require significant investment by governments, industry, and individuals in technology and education to reshape societal attitudes to waste disposal. This multidimensional and multiscale problem will be a pivotal challenge as we close material cycles and move away from linear material use.

*School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta GA, 30332-0100, USA E-mail: [email protected] †School of Industrial & Systems Engineering, Georgia Institute of Technology, Atlanta GA, 30332-0205, USA §Raymond Communications, Inc. 5111 Berwyn Road #115, College Park, MD 20740, USA

40

January 2004

The wide scale reuse and recycling of products to avoid the disposal of concentrated materials is a strategy that both intensifies materials’ use and reduces disposal. This may help in achieving more sustainable patterns of production and consumption for a growing world population that expects ever higher standards of living. Recycling mass produced consumer products such as white goods, refrigerators, and washing machines, is not new, and sophisticated infrastructures for scrap metal have been in place for decades. However, electronic products are a major new category, which have experienced rapid growth over the last decade, particularly in the area of personal computers. It is estimated that, by 2005, one computer will become obsolete for every new one put on the market. Between 1997 and 2004, 315 million computers will become obsolete. This will result in the discard of 550 x 106 kg of Pb, 900 000 kg of Cd, 180 000 kg of Hg, and 0.5 x 106 kg of Cr VI. This will also yield additional waste in the form of 1800 x 106 kg of plastic and at least 159 x 106 kg of brominated flame-retardants from monitors. The disposal of consumer electronics accounts for 40% of Pb in landfills. Additionally, 22% of the yearly world consumption of Hg is used in electronics1. The successful capture and reuse of these streams of materials will require a combination of government initiatives at local, regional, national, and supranational levels, as well as public willingness and innovations in materials, products, and recycling technologies. Above all, this is a systems problem, where

ISSN:1369 7021 © Elsevier Ltd 2004

INSIGHT FEATURE

innovations at different scales and levels must be tied together to achieve the maximum impact.

economically efficient system designs that combine effective recycling technology with socially acceptable collection systems.

Legislative initiatives An important part of the electronics recycling problem is the legislative framework that has or is being established in a particular region. There are 11 countries that currently have ‘mandatory’ electronics recovery laws on the books. They include Belgium, Denmark, Italy, Netherlands, Norway, Sweden, Switzerland, Portugal, Japan, Taiwan, and South Korea. There are extensive voluntary programs in a number of other countries, such as Germany, and draft takeback bills in several more, including China. The European Union (EU) enacted two directives in January 2003. The first, referred to as WEEE, requires industry to ensure recycling of any electronic product with a battery or a cord. A second, Restriction on Hazardous Substances (RoHS), phases out Hg, Cd, Pb, and Cr VI in all electronic items by July 2006, with a number of exemptions. In Asia, Taiwan and Japan have fee systems in place for takeback of computers, large appliances, and air conditioners. Japan’s private collection system includes TVs, while Taiwan’s includes printers. South Korea enacted new takeback laws for electronics in 2003, covering major electronic items, phasing in small products such as cell phones and cameras in 2005. In the Americas, activity has been concentrated in the north. In Canada, provinces with authority to require takeback of electronics cover about 95% of the population, most of which are expected to demand recycling plans from industry by 2005. In the US, the government initiated the National Electronic Product Stewardship Initiative, a series of talks in 2001 to set up a national recovery system for electronics. While no agreement had been finalized as of October 2003, there were 52 electronics waste (e-waste) bills introduced in 26 states in 2003. California enacted a fee on cathode ray tubes (CRTs) in October 2003, as well as a restriction on heavy metals that mirrors the EU RoHS requirements for CRTs over 4”. From this catalog of activity, it would appear that a worldwide consensus is emerging to regulate the disposal of electronic products and that legislation will be enacted over the next few years. However, the specific tack being taken by regions varies and may lead to significant overheads for global manufacturers trying to comply with different regulations. In all cases, it will be important to establish

Recycling system design The system design level is concerned with strategic decisions that will connect discarded products to final use as recycled products, component parts, materials, or energy. The scope of the system design is to connect the product retirement with a suitable point in the forward production system where its reuse, or that of its components or materials, is most economical and environmentally sound (Fig. 1). It can be considered the equivalent of the forward supply chain, which many companies optimize on a daily basis. We have termed this the reverse production system2. Forward supply chains begin with primary extraction of raw materials, such as mining, drilling, or agriculture. In contrast to mining and drilling, the reservoirs of raw materials for recycling are centers of human population with lower resource density, while in contrast to industrial farming, there is greater diversity in the resource. Unlike physical reservoirs, with behavior defined by geological parameters, reservoirs of retired products are governed by socio-economic factors that we are barely beginning to understand. At the most rudimentary level, the potential supply of products is equal to those sold, but even this data is highly fragmented and incomplete. Companies hold such data closely for marketing purposes, while the rise and fall of manufacturers leads to ‘orphan’ products whose numbers are difficult to quantify and whose material composition may no longer be known. The uncertainty in the total number and profile of the sales of these products is complicated by the profile of their use, obsolescence, and failure. The final filter

Increase in manufactured value

Material manufacturing

component manufacturing

Final assembly Point of sale

Raw material refining Chemical recycling

Material compounding

Collection & sorting Demanufacturing

Decrease in manufactured value

Fig. 1 General forward-reverse production systems scheme. Solid arrows indicate the possible flows of material, while the dotted arrows correspond to one option for a recycling system design.

January 2004

41

INSIGHT FEATURE

is the behavior of the consumer with regard to disposal: how quickly and diligently will the product be recycled? These factors make the prediction of recycling volumes extremely difficult; methods are in their infancy3-7 and data on recycling programs in the US is sparse8-12. What has emerged in the US, however, is that participation rates are relatively low, between 1% and 10% of those who have something to recycle. The spectra of the age and condition of the products are very broad, with products 20 years old or older in the case of TVs. Furthermore, the age of retired computers appears to be increasing. This could be a result of the recent economic downturn, causing individuals and companies to hold on to their existing assets for longer13, or that the pace of functional obsolescence is slowing. For electronic products, there are a number of problems that have to be faced at the strategic level, none of which is more challenging than collection, because of the issues of volume estimation outlined above. The initial collection of discarded products has been approached in many ways and is likely to have very different configurations depending on the locale (e.g. US versus Europe). In the US, for example, several companies allow you to mail products back to them, particularly if they still have residual value, such as current generation monitors. This has a very high cost and is unlikely to be effective for bulky items that are obsolete or broken. Some states have collection centers at which products can be dropped off, and nonprofit organizations have sprung up to take back products that can be reused in other channels, such as zero cost computers for disadvantaged groups. A mechanism that is growing in popularity is the ‘special event’, organized at a public venue such as an electronics retailer or university campus, to which the general public is invited to come and drop off their e-waste, often with a fee associated for each car or item. This can work well in the US where suburban ‘big box’ stores with large car parks are the dominant retail outlet, but is likely to be much more difficult to implement in Europe. Europe, however, has some relatively mature systems for the collection of e-waste. For example, the SWICO system in Switzerland and the Netherlands Association for Disposal of Metalectro Products (NVMP) or ICT for IT equipment programs. The NVMP program established regional centers that collect 80% of e-waste, with the remainder coming through retailers. It is estimated that 77% of TVs and 64% of other small brown goods are recovered.

42

January 2004

The inherent problem with collection mechanisms is that they can have very high variability in volume and, hence, in the cost per item collected. For example, a special event may be ineffective because of poor publicity, location, or weather. A fixed location may lead to a very reliable stream of material, but has high fixed overhead costs. Nonprofit organizations are not obligated to take e-waste and may deter recycling if too many items are refused because they cannot be reused. The price charged to the public may also deter participation and lower the volume collected. In Europe, legislative initiatives have placed the burden on manufacturers to finance and develop schemes to recover such products. In the US, the wider distribution of legislative power to the state level is leading to a more diverse set of solutions, weakening the power to negotiate with companies. What is clear, at the current state of development, is that no one collection mechanism dominates the others and continued data gathering on volumes and costs is warranted. The crucial aspect of the collection mechanism is that it can dominate the cost and scale of the overall reverse supply chain and the types of solutions that can be considered. For example, a scheme that can deliver high and consistent volumes of material will encourage systems with significant capital investment. However, at some point, the additional cost of either intensifying the collection system, by raising the participation rate of consumers through advertising expenditure for example, or raising its geographic coverage and hence transportation costs, will outweigh the economies of scale or saturate the market. In other product systems, such as carpet, this has proved to be a particularly difficult trade-off to manage. Effective technologies exist to recover high value monomers from polyamide-6 depolymerization to caprolactam, but developing the necessary collection systems and co-product networks to use other carpeting types has proved very difficult. In electronics, the problems are compounded by the unpredictability of the stream quality in terms of its reusability and, for some materials, the fact that the manufacturing systems to which they are coupled have moved offshore. This leads to the ironic situation that exporting e-waste may actually be the right course of action because the materials recovery operations might benefit from close proximity to the end markets. However, the Basel Convention, adopted by the EU and other OECD (Organization for Economic Cooperation and Development) members, has focused on limiting the movement of

INSIGHT FEATURE

hazardous waste and treating it as near as possible to its point of generation14. Its goal has been to prevent damage to human and environmental health through unsafe handling of waste in regions where standards are lower or enforcement is lacking. This precautionary approach is entirely appropriate until mature systems and monitoring of environmental and health standards for global recovery operations are available and implemented. The previous discussion has emphasized the human systems aspects of electronics recycling systems. Effective technologies for dealing with e-waste are emerging, which appear to be combinations of disassembly and bulk recycling through reducing the product into pieces of a single or few materials that can be separated using physical and chemical properties. Electronic products, like cars, are a combination of some valuable subcomponents and assemblies – such as the central microprocessor or hard drive – and those that have value only as materials, such as the printed circuit board or housing. The recovery options that are pursued must balance the costs of testing and disassembly, which tend to be labor intensive, with the incremental value of the components over their material value. The technological life cycle of a product has a profound interaction with this decision. For example, CRTs are a mature product for personal computers and have experienced rapid declines in prices in order to compete with liquid crystal displays (LCDs) that are penetrating the market. The resale value of recovered CRTs is, therefore, very low in their original markets. This discourages the testing and refurbishment of CRTs and increases the need to recover the leaded glass, metals, and plastics. However, as the market for CRTs declines, leaded glass itself will become obsolete and we will be faced with a disposal problem at a different level. This problem is more complex for the computer itself, as whole systems, or subsystems, can be reused and the obsolescence rate for chips and fixed drives has slowed, but prices for new components have continued to fall. A further complication is that testing, disassembly, and bulk recycling technologies scale in cost very differently with throughput. Thus, for small scale operations that have been adopted to deal with the historical flows of electronic products, disassembly has been possible, but as the volumes, variability, and age of systems increase, there will be a need to shift toward higher throughput, less manually intensive operations. Prototypical examples of separation processes that might be considered for bulk electronics recycling are shown in

33% Shredding/ grinding

Plastic Glass Other Metal Wood

Metals removal 49%

12% 1% 5%

Density-based plastics separation

2% 3% 16%

Non density-based plastics separation 59% 20%

HIPS ABS PPO PP or PE Other

Post-processing

Fig. 2 Schematic of the recycling process and typical compositions of electronics waste.

Fig. 2. For any separation process, the two important issues are to liberate the constituent materials, so that they are no longer physically or chemically bonded to one another, and to find the physical or chemical properties that most differentiate these materials. The liberation of metals from plastics is normally carried out by intensive shredding and battering of the system to break the physical connectors, such as screws and other fasteners. This also reduces the size of the pieces. Metals can be separated by using their ferro- or paramagnetic properties, before sending to smelters to recover the metal fractions by liquid density. Sometimes this can be done without any separation, just size reduction for densification, with the combustible fraction (plastics) recovered as energy. The recovery of energy from plastics is potentially complicated by the presence of brominated fireretardants in the compounds and polyvinyl chloride plastics. If metal separation is followed by plastics separation, these materials can be recovered on the basis of differences in density, hydrophobicity, spectra, solubility, or surface polarity, as indicated in Table 1.

Case study: a US regional system As part of our research, we have been working closely with the State Legislature of the State of Georgia to understand Table 1 Different mechanical separations options for plastics.

Property

Plastics separation method

Density

Dense medium separation (sink-float tanks/drums, dense medium cyclones) Froth flotation (mechanical tanks, flotation columns) Electrostatic separation (free-fall, drum-type) Spectroscopic separation (near infrared, mass, X-ray) Selective dissolution by solvents

Hydrophobicity Polarity Spectra Solubility

January 2004

43

INSIGHT FEATURE

Fig. 5 Scenarios for the design of recycling infrastructure.

Fig. 3 Possible recycling infrastructure choices for the State of Georgia.

the issues in designing an effective infrastructure for recycling e-waste, specifically, TV and computer waste, which makes up the bulk of the weight of electronic products requiring recycling15. The approach starts by developing an outline of the existing and possible recycling outlets for materials in the state and adds possible collection points based on a division into economic development regions. The set of options is shown in Fig. 3. This is combined with an overall pattern of the flow of material based on different disposal channels, essentially commercial and government sources, which tend to be large scale and TV-poor, or

Residential source -TV, CPU, monitor

Municipal collection sites

residential, which is TV-rich but with older equipment, as shown in Fig. 4. As this article points out, there is significant uncertainty associated with the flow of goods and the quality of this flow. In addition, there may be economic development opportunities associated with attracting a recycling operation, such as CRT glass recycling, to the state. To capture some of this uncertainty, we have developed sixteen scenarios that represent the high and low extremes of four parameters. These are: percentage participation by the population, the percentage of usable TVs and computers with monitors, and whether or not the CRT glass recycling can be done out of state. The scenarios are given in Fig. 5. Optimal networks are designed for each scenario, consisting of a subset of processors and collection sites, and the flows of products between them. The problem is resolved with the objective of finding a robust system that would do well in each different scenario, where the quality of a scenario is measured by the deviation of the performance of a robust system from the optimal value. The robust network is shown

Commercial processing sites Residential end customers

Nonprofit recycling sites

Commercial source -CPU, monitor Commercial end customers

Recycler for large commercial sources Commercial source -outside Georgia

Main stream Auxiliary stream

Fig. 4 Abstract flow of products from end consumers to recyclers.

44

January 2004

INSIGHT FEATURE

4.5

Total profit (millions of $)

4

Equ Robust 16

3.5 3 2.5 2 1.5 1 0.5 0

1

3

5

7

9

11

13

15

Scenario Fig. 7 Total annual profit (in dollars) of recycling systems for the State of Georgia, for more details of assumptions15.

Fig. 6 Robust solution for recycling infrastructure in the State of Georgia.

in Fig. 6 and the objective function values for each individual network and the robust network are shown in Fig. 7. The cost structure for the system includes a $5.38 charge per item entering the system. The results are encouraging, showing that the robust network can do quite well compared to the optimal value in all of the scenarios.

Conclusion E-waste represents a challenging recycling problem for several reasons. First, the material complexity of the product – a combination of valuable metals with hazardous ones, such

as Pb and Hg, and low value plastics – makes its diversion from landfills an important consideration and one that will continue to drive the development of environmentally sound recycling processes. Second, it is a widely distributed and diverse basket of consumer products with highly variable rates of obsolescence and failure. This means that it is hard to predict whether particular collection program types will be cost effective in a given region and how much volume will be generated. The next few years will see considerable growth in the volume of electronic products being retired and an entire reverse supply chain will have to be developed around them. This will lead to challenges and opportunities for diverse disciplines and companies. The question is: who will be the Dell™ of recycling? MT

REFERENCES 1. Just Say No To E-Waste: Background Document on Hazards and Waste from Computers, Silicon Valley Toxics Coalition, 1999 www.svtc.org/cleancc/pubs/sayno.htm 2. Ammons, J. C., et al., Decision Models for Reverse Production System Design, In: Handbook of Environmentally Conscious Manufacturing, Madu, C. N. (ed.), Kluwer, Boston, 2000 3. Leigh, N .G., and Realff, M. J., J. Environ. Plan. Management (2003) 462, 147 4. Kleijn, R., et al., Ecological Economics (2000) 32, 241 5. Marx-Gomes, J., and Rautenstrauch, C., Predicting the Return of Scrapped Products through Simulation – a Case Study, In: Proceedings of the Second International Workshop on Reuse, Eindhoven, 1999, 71 6. Tucker, P., et al., J. Environ. Management (1998) 53, 31 7. Tucker, P., J. Environ. Systems (1997-1998) 26, 365 8. Hainault, T., et al., Recycling Used Electronics: Report on Minnesota’s Demonstration Project, Minnesota Office of Environmental Assistance, 2001, www.moea.state.mn.us/plugin/ElectronicsReport.pdf

9. Managing Electronic Equipment, Minnesota Pollution Control Agency, 1999 www.pca.state.mn.us/publications/w-hw4-15.pdf 10. Domestic Policy Initiatives and Voluntary Programs, National Recycling Coalition-Electronics Recycling Initiative, www.nrc-recycle.org/resources/electronics/reports.htm 11. Speirs, D., and Tucker, P., J. Environ. Management (2001) 62, 201 12. Analysis of Five Community Consumer/Residential Collections End-of-Life Electronic and Electrical Equipment, EPA-901-R-98-003, US Environmental Protection Agency, 1999 13. Old Computers Don’t Fade Away Anymore, Wall Street Journal (May 1, 2003) 14. Secretariat of the Basel Convention, UNEP, Basel Convention 2003 www.basel.int 15. Ammons, J. C., and Realff, M. J., Regional Electronics Recycling System Design Planning and Reverse Production System Model Development, Final Project Report to Pollution Prevention Assistance Division, State of Georgia, Atlanta, 2003

January 2004

45