A universal code for environmental management of products

A universal code for environmental management of products

Resources, Conservation and Recycling 53 (2009) 400–408 Contents lists available at ScienceDirect Resources, Conservation and Recycling journal home...

407KB Sizes 3 Downloads 98 Views

Resources, Conservation and Recycling 53 (2009) 400–408

Contents lists available at ScienceDirect

Resources, Conservation and Recycling journal homepage: www.elsevier.com/locate/resconrec

A universal code for environmental management of products Valerie M. Thomas ∗ School of Industrial and Systems Engineering, Georgia Institute of Technology, Atlanta, GA, USA

a r t i c l e

i n f o

Article history: Received 7 May 2008 Received in revised form 6 March 2009 Accepted 11 March 2009 Available online 16 April 2009 Keywords: Barcode UPC RFID Recycle Electronics Batteries Small appliances

a b s t r a c t The UPC code has provided a foundation for efficient product supply chains, the proliferation of products, and the development of superstores. Extension of the universal product code could make reuse and recycling of consumer products more efficient, and provide a foundation for recycling more types of products. Three types of applications are discussed. Use of product codes in existing recycling and refurbishing enterprises (1) is a baseline application that may be cost-effective. Use of product codes for automated sorting of recyclables (2) is technically challenging, may be cost-effective as a follow-on application, and would provide potential for increased recycling of small appliances, small electronics, batteries, and other small goods. A universal product code also provides a basis for recycling innovations, such as recycling rebates and online market applications (3) that could increase reuse and recycling. Voluntary private sector implementation is feasible, with coordination between recyclers, manufacturers, retailers, government, and non-governmental organizations, although recycling regulation and legislation may be necessary for effective development of recycling programs. © 2009 Elsevier B.V. All rights reserved.

1. Introduction A core challenge for closing material cycles by recycling, refurbishing or reusing products is the shear complexity and diversity of products. A recycling program can be established for any given product, but extending recycling to even a significant fraction of all consumer products is a challenge beyond the reach of most of today’s recycling systems. Some countries have successfully implemented programs to recycle selected consumer products at higher rates: Switzerland now collects more than half of consumer batteries for recycling (Binder et al., 2008). However, as illustrated in Table 1 for the U.S., in most places recycling rates for consumer products are low. It is increasingly recognized that environmental management of products requires a lifecycle approach. Lifecycle environmental impacts are now sometimes considered in the design and manufacture of products (Frosch and Gallopoulos, 1989). Beyond design and manufacture considerations, the supply chain and overall retailing operations also need to be designed to support environmental management of products throughout their lifecycle. That is, products need to be designed to be recyclable, and supply chain technology needs to be deployed to make it easy to recycle the product. In the 1970s, the US grocery industry introduced a technology – the Universal Product Code (UPC) – that allowed for the massive

∗ Tel.: +1 404 384 7254; fax: +1 404 894 0390. E-mail address: [email protected]. 0921-3449/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.resconrec.2009.03.004

expansion of the number of products sold in grocery stores. This innovation was soon taken up by big-box retailers including K-mart and Wal-Mart, who required their suppliers to use the UPC. Within a few years this innovation was adopted into retailing worldwide (Dunlop and Rivkin, 1997). The Universal Product Code, by making product marking, product check-out, and product re-ordering quicker and cheaper, provided a basis for what is called “product proliferation”—a significant increase in the number and variety of products. The number of items in a typical supermarket rose from 9000 in 1974 to 45,000 in 2006 (Dunlop and Rivkin, 1997; Food Marketing Institute, 2007; Holmes, 2001). The changes brought by the UPC bar code have been compared “to the changes of the Second Industrial Revolution, in which the railroad and the telegraph served as key technologies that transformed the industrial landscape. The UPC serves a role similar to that of Morse Code: as an economy-wide standard for communication, it enables innovative forms of coordination within and between organizations” (Dunlop and Rivkin, 1997). The experience of the grocery industry with the UPC – a cost of implementation in the range of 0.7–2.4% of sales, and benefits in the range of 3–6% of sales – suggests that direct financial benefits were significant, especially given the small profit margins in the grocery industry (Haberman, 2001; Thomas, 2007). Yet, compared with claims often made for transformative technologies, savings of a few percent are modest. Initial lower-bound estimates of costs and benefits were not much above break-even, and additional benefits were speculative. Nevertheless, these modest but real savings for grocers supported an innovation that allowed manufacturers and grocers together to increase the number of products supplied to

V.M. Thomas / Resources, Conservation and Recycling 53 (2009) 400–408

401

Fig. 1. Linear UPC supply chain with UPC code on the packaging. UPC product codes on packaging facilitate supply chain management to the point of sale. When the consumer removes the packaging, the product code is thereby removed. Today, retail products are usually landfilled at end-of-life.

and managed by grocery stores—and eventually by the entire retail sector. The use of the UPC ends, with a few exceptions, at the point of sale (Fig. 1). UPC codes are almost always on the package, not on the product. The consumer purchases the product, takes it home, takes it out of the package, and throws away the packaging and the UPC code with it. And, perhaps consequently, today’s end-oflife enterprises – dismantlers, recyclers, material recovery facilities, refurbishers, etc. – are what might be expected from a pre-UPC industry, characterized by heavy industrial processing, as in the recycling of paper, steel, aluminum, and plastics, and by laborintensive piece work, as in computer refurbishing and second-hand sales of furniture and household goods. It is conceivable that the UPC could have significantly greater direct financial benefits for existing recyclers than it has had for groceries and retailing. More plausible, however, is that direct savings for recyclers might, optimistically, be comparable to the scale of direct savings for grocers. These savings might be significant enough for recyclers to start making use of UPC codes. Beyond reducing costs for existing recycling operations, a lifecycle UPC approach might provide a basis for recycling other types of products, such as tools, toys, batteries, a range of electronics products, small appliances, and household furnishings. Mirroring the product Table 1 Recycling rates for selected durable goods in the U.S. Product disposed Computers Consumer batteries Primary Secondary Athletic shoes Small appliances Furniture and furnishings

Number consumed per person per yr 0.2 13.4 0.6 2.2

Recycling rate (percent)

Mass (kg) per person per yr

2–6

2

2 8

0.3 0.1

0.2 2 0

proliferation and the development of superstores that followed the introduction of the UPC (5), a UPC approach to lifecycle management might conceivably engender the recycling and reuse of more types of products, and the development of super-recyclers that can cheaply manage the refurbishment or recycling of many types of products. The use of information technology for product lifecycle management has been addressed by a number of previous studies, with primary focus either on radio-frequency identification applications (RFID) or on use of smart chips or other electronic data-logging approaches (Thomas, 2003a,b; Saar and Thomas, 2003; Klausner et al., 1998). The emphasis here is on a more technology-neutral approach to the issue, with application to either bar codes or RFID tags. 2. Putting the code on the product, not just on the package The simplest lifecycle code is the existing UPC-type code, with the code on the product itself, rather than just on the packaging. There are several versions of the UPC; it has evolved over time and is continuing to evolve. The original UPC is a 12-digit code called UPCA. Fig. 2 shows a standard UPC-A code. It is a number, in this case 0-16000-66060-1, that is also encoded as a bar code. The first digit identifies the type of code: 0, 6 and 7 are for manufactured products and 5 is for coupons. The next five digits identify the manufacturer; following five digits identify the product; the last digit is a check digit.

1 4 30

Data for computers from US EPA (2008a,b), for small appliances and furnishings from US EPA (2003). Derivation for consumer batteries and athletic shoes is in Supplemental material.

Fig. 2. A UPC code.

402

V.M. Thomas / Resources, Conservation and Recycling 53 (2009) 400–408

There is also UPC-E, which is a shortened code for small packages. The European Article Numbering (EAN) system is a 13-digit superset of the UPC and contains a country code; the EAN-13 is now the worldwide standard. There is also EAN-8 for smaller products. There are also two-digit and five-digit supplemental codes for periodicals and books. The power of the UPC is its flexibility; the code itself simply identifies the product. This allows the same code to be used for many purposes, and for the details and even entire applications to change and develop over time. In retail stores, the information related to the product – such as the price – is stored in a database and accessed when the UPC code is read into the system at the check-out counter. When prices change, the database is updated; the code does not have to be changed. In recycling centers, information related to the product – such as how to recycle it, or simply what it is – would similarly be stored in a database and accessed when the product comes in for recycling. An alternative to UPC-type bar codes are radio-frequency identification (RFID) tags. As of 2009 these are beginning to be used in the retail sector on cartons and pallets of products; Wal-Mart and the US Department of Defense are requiring its use in some applications. There are also a few retailers introducing item-leveling RFID tagging of products; in the future RFID tags could at least partially replace bar codes as the standard retail product code. RFID tags can be read from a distance of up to 1 m, and do not require a direct line of sight. They are more expensive than printed bar codes, with costs of 10¢ or more per tag (RFID Journal, 2009). The standard code for RFID is the EPC. The EPC is longer than the UPC or EAN code, and contains letters as well as numbers, so that the code can uniquely identify each item. An RFID tag might be easier to read than a bar code in some endof-life applications. Identifying products in curbside waste bins, identifying products coming in to or out of garbage trucks or recycling trucks, identifying products in large recycling facilities and material recovery facilities are all situations in which reading a barcode on a product could be challenging: the bar code would need to be facing the barcode reader directly. RFID tags might be easier to read in these situations because an RFID tag can be

read through other products. Also, RFID tags might be useful when information on the specific item is needed. For instance, if a computer has been upgraded with new internal parts, RFID tags on those internal parts could be read when the computer enters the recycling or refurbishing facility, without needing to open up the computer. Some products could easily accommodate a bar code on the bottom, back, or base: computers, televisions, printers, small electronics (telephone answering machines, cordless phones, radios), small appliances (toasters, hair dryers, coffee makers), fluorescent light bulbs, etc. But there are other products for which application of a bar code might be difficult or impossible: aluminum foil, small cell phones, sponges, wine goblets, jewelry, clothing, handbags, cutlery, incandescent light bulbs, etc. Some products cannot incorporate a bar code because of the nature of the product (aluminum foil), some because it would be unattractive (wine goblets), others because the product is too small (jewelry). Typical UPC bar codes are as small as 1 cm × 3 cm; UPC-E codes are as small as 1.7 cm × 1.7 cm. RFID tags are typically a bit larger, with an area of more than 10 cm2 . As RFID technology develops, potential applications can be expected to grow. A decision to implement product codes for environmental management does not require a permanent decision about the specific code or technology. The EPC is being developed to be compatible with the UPC, EAN, and other product codes. This allows RFID to be phased in with UPC barcode systems. Early applications of product codes for environmental lifecycle management could use the UPCtype bar code. For some product categories, a UPC may be all that is ever needed to implement an effective recycling or reuse system. For other products, particularly expensive, longer-lived products that might be upgraded over time or for which details about how the item has been used are useful, a different technology, such as RFID or embedded data logging, might be used. Three types of environmental applications will be considered: UPC to assist existing recycling and refurbishing operations; UPC for sorting hazardous or recyclable items at a materials recovery facility; and UPC codes to assist consumers in recycling and reuse. These applications indicated in Fig. 3.

Fig. 3. Closed loop UPC supply chain, with UPC code on the product. The three types of applications discussed below are indicated schematically. Type 1: repair, refurbish, and recycle, with products managed individually. Type 2: materials recovery facility or other automated, conveyor-belt sorting. Type 3: product-code-supported applications such as online market developments, reverse vending, and recycling coupons.

V.M. Thomas / Resources, Conservation and Recycling 53 (2009) 400–408

2.1. Type 1: efficiency for existing refurbishers and recyclers In established recycling programs worldwide, product codes have the potential to provide information which could reduce costs, increase efficiency, and improve the potential for more effective refurbishment and reuse. For example, in many electronics recycling and refurbishing facilities, each item is handled by hand, often several times by several people. The value of old computers, or the recycling fee charged, is relatively high for end-of-life computers, typically US$ 15 or more (Sarkis, 2003). This high value makes it possible for recyclers and refurbishers to spend at least several minutes of total labor time with each item. For example, at the University of Massachusetts’ electronics demanufacturing program, the rate of throughput is 0.8 kg/min (Sarkis, 2003; Pepi, 1998). For a typical PC weighing 20 kg, this implies roughly 25 min per computer. This type of high value, “slow” recycling might be enhanced by information linked to product codes. The product codes on products could identify the make and model, and link to a database which provides information on how to dismantle the product, or to information on the hard drive, RAM, and circuit boards, the type of plastic in the case, and hazardous materials content or other product composition information. This type of system has been described for the recycling of cell phones by Saar et al. (2004), using the GSM bar code underneath the cell phone battery as the product code. RFID tags could also be used to provide assurance that products were not being exported to developing countries in violation of US law or the Basel Convention (US GAO, 2008). The cost of implementing such a system would include the cost of putting the UPC code on the product, the cost of hardware including bar code reader systems, or RFID readers, and the cost of software and database development. 2.1.1. Cost of putting the code on the product The direct cost of printing a bar code on a product is, for all intents and purposes, zero. The cost of the ink is negligible, and most small appliances and products already have printed labels, serial numbers, and codes of several types. Alternatively, if the product code were implemented as an RFID tag, the cost would be perhaps 10¢ per tag (RFID Journal, 2009). A complex electronic product might have as many as four RFID tags: one main tag, one on the hard drive, one on the display, and one tag on a part added later as a upgrade or repair. The cost per item recycled depends on the fraction of items participating in the code-based recycling system. If participation is only 10%, then a 10¢ RFID tag on each item would translate to a cost of US$ 1 per item recycled. 2.1.2. Cost of hardware at the recycling facility A high-quality hand-held bar-code reader costs about US$ 300. A computer to store the data costs roughly US$ 2000. The number of bar code readers would approximately scale with the number of items to be scanned. Assuming one bar code reader is needed for each 100 items scanned per day, averaging one item every five minutes over an 8 h day, and assuming the facility operates 250 days per year, then one bar code reader is needed for every 25,000 items recycled per year. Taking the total hardware cost to be US$ 2500 and amortizing over a 7 years at 5% interest, the annual hardware cost is US$ 475, equivalent to a hardware cost of 2.5¢ per item recycled. RFID readers cost about US$ 1000, so an RFID-based system would have a hardware cost closer to 3¢ per item recycled (RFID Journal, 2009). 2.1.3. Cost of software Database development and maintenance might be done by or for each individual recycling company, with capability built up

403

slowly, or might be centralized by the electronics industry, small appliance industry, UPC specialists, or other organization. A basic database could simply contain the make and model corresponding to the UPC code, and the UPC code would be used for inventory and record-keeping. A more developed database could contain recycling instructions for high-value products, and could share information with manufacturers or recycling databases, could provide rebates to consumes, and so on. With computer recycling rates still relatively low (Table 1), database costs need to be kept low if the system is to be self-financing. For example, a minimum software and database cost of US$ 100K for a U.S.-based system, if allocated to all of the roughly 6 million computers recycled annually, would come to 2¢ per computer recycled. If allocated to all of the computers sold, the cost per computer would be less. Eq. (1) shows how the cost per recycled item (C) depends on the cost of coding each item (P), the fraction of coded items recycled (f), the cost of the hardware per item recycled (H), the cost of the software (S), and the number of items recycled (N). C=

P S +H+ f fN

(1)

Eq. (1) shows that the hardware cost is independent of the participation rate, because it is purchased only by participating facilities. As discussed above this hardware cost would be 2–3¢ per recycled item for computer refurbishers. The cost of coding depends on the participation rate, although this cost could be effectively zero if optical bar codes were used. The cost of the software is taken to be a fixed cost for the entire product class, so the cost per recycled item depends on both the participation rate, f, and the number of items recycled, N. Overall, this approach would cost at least 5¢ per recycled computer using a bar code approach (0¢ + 3¢ + 2¢). For such a system to break even, the economic benefits of the UPC code for electronics recycling would need to be at least 5¢ per computer. An RFID-based approach using one RFID tag per computer and with a 23% recycling rate (Table 1) would add an additional 43¢ per computer recycled (43¢ + 3¢ + 2¢). 2.1.4. Benefits The labor costs at a computer recycler have been reported to be about US$ 2.40 per computer (Pepi, 1998), and often considerably more (W. Cade, P.C. Rebuilders, 2009, personal communication). Grocers were able to achieve labor cost savings of 3–6% with introduction of the UPC. If similar savings could be achieved at computer recyclers, this would be 7–14¢ per recycled computer. Optimistically, then, if computer recyclers and refurbishers could get the same labor savings as grocers, the system could at least break even. The public interest in a UPC lifecycle code is not in saving money for refurbishers and recyclers, but in reducing environmental impacts through increasing reuse and recycling. Is it plausible that use of product codes by electronics recyclers would result in more electronics being reused, refurbished, or recycled? This type of application is analogous to the initial application of UPC codes in grocery stores: it can increase efficiency and perhaps save enough money to more than cover the costs. To achieve “proliferation” of product recycling, or even just to achieve increased recycling rates in existing programs, new applications need to be built on this baseline application, possibly through automated sorting (Type 2), or through a range of other code-mediated applications (Type 3). 2.2. Type 2: product code for sorting products Product codes might be used to sort products at a range of points along the product end-of-life supply chain: in households, at curbside, at recyclers, or for pre-processing before incineration, landfilling or smelting. For this assessment, facilities that sort curbside recyclables and municipal waste (called Material Recov-

404

V.M. Thomas / Resources, Conservation and Recycling 53 (2009) 400–408

ery Facilities or MRFs) will be considered. These facilities typically sort office paper, newspaper, cardboard, metal, steel, glass, and plastic bottles. Many of today’s material recovery facilities already have magnets to sort steel, eddy current separators to sort aluminum, and infrared detectors to sort plastics. Items that might be recoverable using a product code include small electronics, batteries, products containing batteries or electronics, a wide variety of plastics-containing products, small fluorescent light bulbs, athletic shoes, specially tagged valuables, etc. UPC bar codes on products could work for this application if items were separated and moved more or less one-by-one – singulated – down a conveyer belt. RFID might work better, because RFID tags can be detected without a direct line of sight, and RFID detectors could detect batteries or other important components inside products, and could be used on products for which a visible bar code might not be popular (as for athletic shoes, toys, or clothing). The cost of a product code reader system might be comparable to the cost of optical sorters that are increasingly being used at recycling facilities. Optical sorters identify different types of plastic resins and paper from their infra-red signature, and use a puff of air to separate out identified items. Many of the components and much of the design of a product code sorter could be similar to an optical sorter. For a MRF serving a population of one million people and handling 35 tonnes per hour, 15 h per day, 250 days per year, an optical sorting system costs US$ 1 million, including shoots, hoppers, and supports, shipping, installation, and engineering. The annual cost would be US$ 190,000 per year when amortized over seven years at a 5% interest rate (Entec, 2006). This is equivalent to approximately 20¢ per year for each person served by the MRF. A key issue, in addition to identification per se, is the sorting mechanism. Air puff sorters currently work well for items with a mass of up to about 10 g. But most durable products weigh more. A mechanical sorting system, such as used by UPS and the US Postal Service for their barcode-based systems, could be used for products with mass of up to several kilograms. Code-assisted hand sorting is also an option. For the purpose of this evaluation, batteries will be considered as the target product. As shown in Table 1, each person in the US and Canada is estimated to dispose of 14 dry cell batteries per year. Thus a facility serving 1 million people could receive up to 14 million batteries per year, if consumers were instructed to put batteries in their recycling bins. Some batteries are contained in products and disposed with the product, so both the battery itself and the battery-containing product would need to be coded for recycling. If the system could successfully retrieve 50% of the batteries (taking into account that not all batteries would enter the recycling system, and that the retrieval mechanism would not be 100% effective), then a facility receiving 14 million batteries per year could remove 7 million batteries from the recycling waste stream. With a hardware and software cost totaling US$ 190,000 per year, the cost would be slighly less than 3¢ per recovered battery for 7 million batteries. Per person, as stated above, this would come to about 20¢ per year. This system would also retrieve battery-containing products. If the sorting system were based on optical bar codes, the cost of the product code could be taken as zero. But if RFID were used and if the cost of each tag were 10¢, this would add twice that, or 20¢ per battery, for a 50% effective system. Also, if sorting systems were established in, for example, only 10% of geographic regions, the cost per recovered battery would rise to US$ 2. Eq. (2) expresses the cost per recycled item (C) as a function of the cost of coding the product (P), the cost per unit time and per capita of the hardware (H) and software (S) at the MRF, the per capita disposal rate of the product type (n), the participation fraction (f), and the sorting efficacy (e).

C=

P H+S + ne fe

(2)

The costs of coding the product would be high if the participation rate were low, suggesting that zero cost bar codes have an advantage over RFID. The hardware and software costs per item fall as the number of items increase. 2.2.1. Benefits Is recovering a battery for recycling worth 3¢? There is some financial value in retrieving the metal from batteries: 2008 values for zinc, cadmium and nickel are about US$ 1.90/kg, US$ 7/kg, and US$ 21/kg, respectively (USGS, 2009). Since batteries include both the chemically active metals as well as battery casing and inert materials, the valuable metal content is only a fraction of the total battery mass. For example, cadmium is roughly 15% of the total mass of a nickel cadmium battery, and nickel is roughly 20% of the mass (Thomas, 2003a,b). So, for example, a 20 g nickel–cadmium battery would contain about ¢ worth of nickel and ¢ worth of cadmium. Heavier nickel-containing batteries are more valuable; ordinary alkaline zinc–manganese batteries are less valuable. Taking into account all of the material in the batteries, the average material content of a disposed consumer battery is about 9¢; details of the calculation are shown in Supplemental material. This 9¢ value per battery is an average over low value alkaline batteries and high-value rechargable batteries; a system which only retrieved rechargable batteries would have a higher average battery value, and also a higher retrieval cost per battery. Also, the 9¢ per battery material value does not include the cost of recycling itself; the net value of an average battery would be less than 9¢. The public interest in increasing the recycling of batteries is based on environmental considerations rather than economic considerations. An environmental assessment of the costs and benefits of battery recycling would include environmental impacts of the entire lifecycle of the batteries, their material content, and the relative merits of recycling versus incineration or landfilling. For nickel–cadmium batteries, recycling has been calculated to provide an overall reduction in energy use and in emissions (Rydh and Karlström, 2002). It is also possible to evaluate the amounts of key materials that would be diverted from landfills, and to compare the cost of current battery recycling programs to the cost of a product-code based recycling program. The data on battery consumption indicate that about 3600 tonnes of cadmium are consumed in the U.S. in consumer batteries and another 400 tonnes per year in Canada (see Supplemental material). With a recycling rate of 8%, the amount of this cadmium that is recycled is about 320 tonnes, and about 3700 tonnes is landfilled or incinerated. Nickel cadmium batteries are now the main use of cadmium. The product-code-based sorting system described above would increase the recycling rate from 8% to an estimated 50%. If implemented at one facility serving 1 million people, such a system would divert 6 tonnes of cadmium from landfill or incineration. If extended throughout the United States and Canada, such a system would divert 2000 tonnes of cadmium from landfills, resulting in about a 45% decrease of cadmium sent to landfills or incinerators, and a factor of 6 increase in the cadmium recycling rate. In the Mercury-Containing and Rechargable Battery Management Act of 1998, the U.S. Congress found that “it is in the public interest to . . . provide for the efficient and cost-effective collection and recycling or proper disposal of used nickel–cadmium batteries . . .” (PL 104-142, 1998). In the U.S. and Canada, the collection and recycling of rechargeable batteries is managed by the Rechargable Battery Recycling Corporation, and funded by license fees paid by rechargeable battery and product manufacturers. As of 2004,

V.M. Thomas / Resources, Conservation and Recycling 53 (2009) 400–408

405

Table 2 Costs, benefits and feasibility for environmental applications of product codes. Type

Description

Cost

Benefit

Technical feasibility

1

Inventory and recycling management for recycling and refurbishing Automated sorting at MRFs or elsewhere for batteries, small electronics, and other products

>5¢ per recycled computer

∼7–14¢ per recycled computer

Easy

∼3¢ per recovered battery

Demonstration needed

Reuse and recycling innovation: online markets, reverse vending, efficiencies for charity second-hand sales, coupons and rebates for recycling

0 upfront cost

∼9¢ of material value per battery; factor of 6 increase in rechargeable battery recycling; tripling of US cadmium recycling Multiple post-consumer product reuse and recycling options

2

3

fees for C-cell batteries were 3.7¢ for NiCd batteries fees for other NiCd batteries scale with the energy content (RBRC, 2007). There is also a fee on non-NiCd batteries of 0.25–2¢ per battery. The lifecycle environmental impact of non-rechargeable batteries may be larger than that of rechargables (Lankey and McMichael, 2000). Despite the establishment of recycling programs, Table 1 indicates that only a small fraction of the consumer nickel–cadmium batteries are diverted from the municipal waste stream. With a fee of 1¢ to 4¢ per rechargeable battery and only an 8% recycling rate, the cost of the current battery recycling program comes to about 30¢ per collected battery, again, not including the direct cost of recycling per se. This indicates that a product-code based system could have significantly lower costs than the current system, and achieve a significantly higher recycling rate. Overall, then, it is possible that with a cost of about 3¢ per item or 20¢ per person per year, a metal recycling benefit of a few cents per item and an environmental benefit of a few cents per item, the cost–benefit balance might be sufficient to justify this type of system. As another point of comparison, Switzerland finances its battery recycling program with an Advance Recovery Fee on the sale of all consumer batteries. The annual average cost to consumers is 1.45 francs per year, or US$ 1.25 per person per year (Batrec, 2009). Interviews with Swiss waste management experts indicated that an RFID-enabled waste sorting system could significantly increase the battery recycling rate in Switzerland and reduce the number of batteries being incinerated with municipal solid waste (Binder et al., 2008). In the European Union, the sales price of electronic products includes a fee to support the Waste Electronics and Electrical Equipment directive (WEEE) recycling program. The fee ranges from 0.5 D (60¢) for small electronic products to as much as 12 D (US$ 15) for large LCD televisions (Creative, 2009; Dell, 2009). The cost of a universal product code system would be a small fraction of the overall WEEE program cost; the WEEE program could potentially provide a context for testing and developing applications of universal product codes for environmental management. 2.3. Type 3: stimulating innovation with product codes What would happen if most products had a UPC on them? Recyclers could use the codes to make their operations more efficient, as described above. But a host of other applications could piggy-back on the same code. UPC codes on products might stimulate innovation analogous to the introduction of UPC codes on packaging. 2.3.1. Recycling coupons and rebates Organizations could provide a rebate to the consumer for bringing the item back to a store or a recycling center. The consumer could drop the item through a slot into a recycling bin, a UPC reader would read the code as the item goes through the slot, and the bin would print out a coupon for the consumer. Coupon values

Easy

would depend on the value of the item and on the specific recycling plan. Product codes could be printed on products as actual coupon codes, which start with the number 5. Alternatively, the regular UPC code could be used, allowing retailers, individual companies, or recycling centers to choose whether and how much of a rebate to provide. These “recycling rewards” might be provided by the manufacturer, but also by any organization that wanted to establish a recycling reward program. The product code could be used to provide rebates for bringing items back to the store, or they could be linked to a wide range of other potential offers or benefits. 2.3.2. Product repairs Product codes could be used by consumers to order spare parts and repairs, to access user manuals, and to find out about new products or services. Consumers would not need to own a barcode scanner; they could simply type the printed barcode number into the appropriate web site to receive product information. 2.3.3. Increased re-use UPC codes on products also could make it easier for consumers to sell or donate items for reuse. A number of Internet sites already allow customers to use product codes to easily list items for sale. Charities such as Goodwill, the Salvation Army, and church rummage sales could upload UPC codes to better advertise their inventories. The open-ended potential of the UPC might cause manufacturers to resist putting a UPC code on their products. Do all manufacturers really want their products to be recycled or reused? Does a UPC code on the product imply some responsibility for recycling, or to establish a program to recycle the product? Might sales of new products fall if reuse takes off? Might a manufacturer’s competitors be able to take better advantage of a lifecycle-oriented supply chain? All of these questions can be raised in the prospect of establishing a lifecycle-based supply chain. Greater reuse can be expected to reduce sales of new products, although on a less than one-to-one basis. For example, the increases in used book sales on Internet markets has been estimated to have decreased sales of new books by about 0.7 books for every book sold used (Thomas, submitted for publication). The impact of secondhand sales on the profits of manufacturers and retailers can be positive or negative, depending on the durability of the product, the cost of recycling, and how manufacturers and retailers respond to the market changes (Agrawal et al., submitted for publication). Table 2 summarizes the costs, benefits, and feasibility of the types of applications discussed above. 3. Potential environmental drawbacks of RFID If a large fraction of consumer products, including food packaging, someday had RFID tags, the number of tags produced and disposed might grow to one tag per person per day, or more. In the

406

V.M. Thomas / Resources, Conservation and Recycling 53 (2009) 400–408

U.S., this would come to 100 billion tags per year. Global tag consumption could approach two trillion tags per year. Most of these tags would end up in municipal solid waste, or in the recycling streams for paper, cardboard, metals, glass and plastics. RFID tags typically contain antennas made of copper, aluminum, or silver compounds, as well as a silicon integrated circuit, and adhesives, plastics, and paper (DOD, 2004). Specific concerns have been raised about contamination of recycling of glass, steel, aluminum, paper, and plastic (Krauchi et al., 2005). These are discussed in turn below. 3.1. Glass recycling Silicon chips in RFID can present a problem for RFID tags attached to glass containers, because silicon melts at a different rate than glass. If the silicon stays with the glass through the recycling chain and into the glass furnace, this can result in silicon ‘balls’ in new packaging. These are potential weak spots, especially of concern for pressurized containers. This suggests that RFID tags used on glass products need to be completely removable, or be placed in the caps rather than on the glass itself (British Glass, 2005). 3.2. Steel recycling An RFID tag may contain about 20 mg of copper (Copper Development Association, 2005). If used on a typical steel can weighing 20 g, this would correspond to an effective copper concentration of 0.1%, and with continuing recycling, the copper content in steel would build up (Igarashi et al., 2007). This would reduce the quality of the steel (Steel Recycling Institute, 2006). Thus, RFID tags used on steel products need to be completely removable or made with a non-copper antenna. Aluminum-based RFID tags would not interfere with steel recycling. At the very high temperatures of steel making, any incidental scrap aluminum from such tags would be removed in an exothermic reaction. 3.3. Contamination of aluminum recycling Aluminum can be recycled multiple times with very high levels of purity. No contaminants of any type are allowed in the feed stream; contaminants must be screened out or removed via other means (DOD, 2004). Accordingly, RFID tags on aluminum products should be completely removable or made of benign materials. 3.4. Contamination of paper recycling RFID tags can be expected to enter recycling mills as part of old corrugated containers. Testing by the National Council for Air and Stream Research indicates that copper foil antennae are likely to remain intact through the hydropulping process and be readily captured and removed. Silver ink antennas, however, contain 2–3 ␮m particles of silver. Using tags containing about 16 mg of silver, a pilot study indicated that most of the silver remains with the pulp, although enough silver was found in the effluent water to indicate that levels could approach regulatory limits in some circumstances (Maltby et al., 2005). 3.5. Contamination of plastic recycling High-density polyethylene (HDPE) is used for milk jugs, detergent bottles and other applications; polyethylene terephthalate (PET) is used for soda and water bottles and other applications. Together, HDPE and PET are the two most commonly recycled plastics. RFID tags are typically attached to a PET substrate. RFID tags are expected to be readily separable from HDPE bottles during the recycling process, because RFID tags are denser than HDPE and will

separate out in the first separation step of the HDPE recycling process: HDPE floats while RFID tags will sink. However, separating RFID tags from PET bottles is more of a challenge. PET has a specific gravity larger than one (1.2–1.4 g/cm3 ), so both RFID tags and PET flakes sink in water. In fact, since RFID tags typically are made on a PET substrate, there may not be enough of a mass difference to separate PET from RFID tags using cyclonic separation (Randy Stigall, UPM Raflatac, February 19, 2007, personal communication). Thus, any RFID tags used on PET bottles should be designed to be easily removable, whether by using a non-PET substrate, or through other means. Active, battery-powered RFID transponders will be less widely consumed than passive RFID tags, at a rate of perhaps one per person per year. This would add to the already wide use of batterypowered devices, currently approximately 14 batteries per person per year (Environment Canada, 2007). The low recycling rate for consumer batteries suggests that RFID will add to the disposal of batteries in municipal solid waste. In summary, today’s RFID tags are neither biodegradable nor recyclable. RFID tags may get in the way of recycling of many types of packaging and other materials. Choice of RFID material, and removability of tags, would seem to be readily able to avoid these problems. With carbon-based conducting inks as antennas, and biobased plastics or papers as the substrate, the bulk of the RFID tag could be biodegradable or otherwise environmentally benign. The silicon chip, however, may have environmental impacts that are harder to address, if not at the end-of-life stage then certainly at the production stage; the manufacture of silicon chips requires significant energy and water and has significant pollutant emissions (Williams et al., 2002).

4. Mechanisms for adoption Recycling rates for consumer products are generally very low (Table 1). The main reasons are the relatively low cost of disposal and the lack of incentives or requirements for recycling. It is increasingly recognized that voluntary environmental programs can provide limited environmental benefits, but are not sufficient to accomplish significant gains (Morgenstern and Pizer, 2007). Development of systematic recycling of products is likely to require legislation or regulation to provide a framework for action. Cost-effective comprehensive recycling systems will require not only regulation and legislation, but also the full application of information technology, and the unleashing of innovative potential. Could product codes could do for recycling what they have done for retailing? Could codes on products be the foundation – the cornerstone technology – that would allow recycling of many more types of products? Universal product codes could make it more feasible to manage large numbers of diverse products, and provide a technical basis for product management after the point of sale. The three applications discussed above illustrate different ways that product codes could be used for environmental management. The first type – using the code to assist recycling – might provide an overall savings for recyclers of electronics and other durable products. However, in itself this application is unlikely to provide savings large enough to qualitatively transform the recycling industries. Rather, the public benefit that might result from this approach is the potential for recyclers and refurbishers to handle more and a wider range of products, as happened in the grocery and retail sectors when UPC codes were introduced. The second type – using a product code for automatic identification and sorting of products for recycling – is very different

V.M. Thomas / Resources, Conservation and Recycling 53 (2009) 400–408

from the first case, in that it targets items that are currently not being recycled. The preliminary estimate indicates that the costs might – optimistically – be justified by the economic and environmental benefits. But automated sorting is technically challenging. And before product sorting technology would be developed, there would need to be codes on products. Automated product sorting is an application that might be developed after environmental product codes become standard for other applications. The third type – innovations in repair, reuse, and recycling – also provides a mechanism for recycling items that are not currently being recycled. Upfront costs for this approach would low—arguably zero if optical bar codes are used. Conceivably, coupons for recycling, or other applications, could be popular with consumers and have marketing benefits. The full environmental benefit of a universal product code for lifecycle environmental management is likely to be from innovations that cannot be quantified in advance. Nevertheless, introduction of the environmental code must be based on a clear application, rather than on future innovations. Using product codes for environmental lifecycle management could, therefore, be based on both prospects for increasing efficiency in recycling facilities – case 1 – and increasing recycling through consumer assistance – case 3. The simplest and lowest cost application is to use the environmental product code for inventory of recycled products and management of returns for recycling. Together these applications could form a foundation for recycling and refurbishing more types of products, more efficiently. A number of types of organizations could contribute to the development and application of product codes for environmental management. • The UPC and EAN system is managed by GS1. Input from, and support by GS1 and related organizations would be important to the universal adoption of an environmental product code in a way that is compatible with the existing infrastructure. • Retailers could support the development and promote the adoption of environmental product codes. • Organizations that rate the environmental performance of products and firms could be instrumental in shaping the development of the environmental product code and in promoting its adoption. For example, EPEAT, the Electronic Product Environmental Assessment Tool (EPEAT, 2009) is increasingly recognized for its environmental criteria for electronics products, and e-Stewards (2009) is increasingly recognized for its recycling standards. Product codes could become one of the criteria by which the environmental performance of electronics and other products are evaluated. • Governmental environmental agencies could facilitate development, evaluation, and implementation of a product environmental code. In the United States, the US EPA’s Resource Conservation Challenge has a system of partnerships and projects, with recognition and rewards for successful and far-reaching projects, and its Responsible Recycling Practices provides a context for consideration of product codes for recycling (US EPA, 2008a,b). Implementation of UPC-based product lifecycle management may be the type of innovation that the US EPA could support through such a program. In the European Union, the Waste Electrical and Electronic Equipment Directive and the Battery Directive could provide a context for the development of universal product codes for environmental management. • The UPC code has been developed and managed by the private sector, and the extensions to product lifecycle management could also be developed by the private sector. But because the environmental benefits of product lifecycle management are public benefits, government and non-governmental environ-

407

mental organizations may have an important role in developing and promoting comprehensive recycling applications. Acknowledgement This research was initially supported by grant number R82958501 from the US Environmental Protection Agency’s Science to Achieve Results (STAR) program. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.resconrec.2009.03.004. References Agrawal V, Ferguson M, Thomas V, Toktay LB. Is leasing green? The environmental impact of remarketing and disposal under leasing and selling. Manage Sci; submitted for publication. Batrec. Price for battery recycling in Switzerland. http://www.batrec.ch/enus/ihre fragen/medien.html; 2009 [accessed on March 6, 2009]. Binder CR, Quirici R, Domnitcheva S, Staubli B. Smart labels for waste and resource management: an integrated assessment. J Ind Ecol 2008;12(2):207–28. British Glass. RFID tags present challenge to glass industry. http://www.britglass. org.uk/NewsEvents/BGNewsCurrent/RFIDTagsPresentChallengef.html; 2005 [accessed on March 6, 2009]. Copper Development Association. Copper applications in innovative technology area. http://www.copper.org/innovations/2005/10/copper speeds commerce saves money in rfid antennas.html; 2005 [accessed on March 6, 2009]. Creative. WEEE charges. http://ie.europe.creative.com/shop/weeeinfo.asp; 2009 [accessed on March 6, 2009]. Dell. Dell WEEE. http://ireland.dell.com/content/topics/topic.aspx/emea/topics/ weee/charges?c=ie&l=en&s=dfh&∼lt=popup; 2009 [accessed on March 6, 2009]. DOD. Notes from meeting: RFID tags, recycling, and the environment. White House Conference Center. September 14; 2004. Dunlop JT, Rivkin JW. Introduction. In: Brown SA, editor. Revolution at the checkout counter. Cambridge, MA: Harvard University Press; 1997. Entec. An evaluation of optical sorting of plastic resins at the Region of Peel MRF; 2006, http://www.stewardshipontario.ca/bluebox/pdf/eefund/ reports/85/85 optical sorting report.pdf [accessed on April 3, 2009]. Environment Canada. Canadian consumer battery baseline study. Final report. http://www.ec.gc.ca/nopp/docs/rpt/battery/en/toc.cfm; 2007. EPEAT. http://www.epeat.net/; 2009 [accessed on March 6, 2009]. e-Stewards. http://www.e-stewards.org/; 2009. Food Marketing Institute. Supermarket facts. www.fmi.org/facts figs/superfact.htm; 2007 [accessed on September 2, 2007]. Frosch RA, Gallopoulos N. Strategies for manufacturing. Sci Am 1989;261(3):144–52. Haberman AL, editor. Twenty-five years behind bars. Cambridge, MA: Harvard University Press; 2001. Holmes TJ. Bar codes lead to frequent deliveries and superstores. Federal Reserve Bank of Minneapolis, Research Department Staff Report 261; May 2001. Igarashi Y, Daigo I, Matsuno Y, Adachi Y. Estimation of the change in quality of domestic steel production affected by steel scrap exports. ISIJ Int 2007;47(5):753–7. Klausner M, Grimm WM, Hendrickson C. Reuse of electric motors in consumer products: design and analysis of an electronic data log. J Ind Ecol 1998;2(2):89–102. Krauchi P, Wagner PA, Eugster M, Grossmann G, Hilty L. Impacts of pervasive computing: are RFID tags a threat to waste management processes? IEEE Technol Soc Mag 2005;24(1):45–53. Lankey RL, McMichael FC. Life-cycle methods for comparing primary and rechargeable batteries. Env Sci Technol 2000;34:2299–304. Maltby V, McLaughlin D, Unwin J. Fate of copper and silver from RFID labels during recycling of OCC and potential environmental and product implications. National Council for Air and Stream Research (NCASI); June 2005. Morgenstern RD, Pizer WA. Reality check: the nature and performance of voluntary environmental programs in the United States, Europe, and Japan. Washington, DC: Resources for the Future; 2007. Pepi J. Scrap electronics processing. technical report 7. Chelsea Center for Recycling and Economic Development. http://www. chelseacenter.org/pdfs/TechReport7.pdf; August 1998 [accessed on September 21, 2007]. Public Law (PL), 104–42; 1998. Rechargable Battery Recycling Corporation (RBRC). Standard License Agreement; 2007. RFID Journal. Frequently asked questions. http://www.rfidjournal.com/faq/; 2009. Rydh CJ, Karlström M. Life cycle inventory of recycling portable nickel cadmium batteries. Resour Conserv Recycl 2002;24(4):289–309. Saar S, Thomas V. Toward trash that thinks: product tags for environmental management. J Ind Ecol 2003;6(2):133–46. Saar S, Stutz M, Thomas VM. Toward intelligent recycling: a proposal to link bar codes to recycling information. Resour Conserv Recycl 2004;41(1):15–21.

408

V.M. Thomas / Resources, Conservation and Recycling 53 (2009) 400–408

Sarkis J. Operations of a computer equipment recovery facility. In: Kuehr R, Williams E, editors. Computers and the environment: understanding and managing their impact. Dordrecht: Kluwer Academic Publishers; 2003. Steel Recycling Institute. Radio Frequency Identification (RFID) tags: copper content detrimental to recycling. http://www.recycle-steel.org/rfid.html; 2006 [accessed on March 6, 2009]. Thomas VM. Product self-management: evolution in recycling and reuse. Env Sci Technol 2003a;37(23):5297–302. Thomas VM. Theoretical calculation of product contents: battery and cathode-ray tube examples. Env Sci Technol 2003b;37:2016–9. Thomas V. A universal code for lifecycle management of products. In: Proceedings of the international symposium on electronics and the environment. Orlando; 2007. Thomas VM. The environmental potential of reuse. J Ind Ecol; submitted for publication. US EPA. Municipal solid waste in the united states: 2001 facts and figures. Office of Solid Waste and Emergency Response. EPA530-R-03-011; 2003.

US EPA. Statistics on the management of used and end-of-life electronics. http://www.epa.gov/epawaste/conserve/materials/ecycling/manage.htm and fact sheet: managed of electronics waste in the United States. http://www.epa.gov/epawaste/conserve/materials/ecycling/docs/fact7-08.pdf; 2008a. US EPA. Responsible recycling practices. http://www.epa.gov/epawaste/conserve/ materials/ecycling/r2practices.htm; 2008b. US GAO. Electronic waste. EPA needs to better control harmful U.S. exports through stronger enforcement and more comprehensive regulation. GAO-081044; August 2008. USGS. Mineral commodity summaries. http://minerals.usgs.gov/minerals/pubs/mcs/; 2009 [accessed on April 3, 2009]. Williams ED, Ayres RU, Heller M. The 1.7 kg microchip: energy and material use in the production of semiconductor devices. Env Sci Technol 2002;36(24): 5504–10.