Recycled Plastics in Infrastructure: Current Practices, Understanding, and Opportunities (2023)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

219 Appendix E Overview of Recycled Plastics Supply and Demand: Identifying the Critical Market Bottlenecks for Closing the Loop Basuhi Ravi, Karan Bhuwalka, Elizabeth Moore, and Randolph Kirchain Massachusetts Institute of Technology STATE OF PLASTICS WASTE GENERATION In 2019, global plastic production totaled 368 million metric tons (MMT), of which 60 MMT (McKinsey 2019) were used in the United States. Only a handful of plastic types make up the majority of this vast sum: polyethylene (PE), polyethylene terephthalate (PET), polypropylene (PP), polyvinyl chlo- ride (PVC), polyurethanes (PURs), and polystyrene (PS) together constitute 88 percent of all thermoplastic1 production (Geyer et al. 2017). Plastics are versatile materials—even within a particular plastic type, their chemistries and associated properties vary considerably across applications and sectors. While this complicates waste handling at end of life, it also means that there are many opportunities for plastics waste to be recycled. The packaging sector is the largest consumer (45 percent) of plastics worldwide and in the United States—and almost all of it is single use, des- tined to be waste within a year. This short-lived application utilizes large quantities of PE (bags and films, milk containers, tubs), PET (bottles for wa- ter, soft drinks, cleaners), some PP (food packaging, caps, sweets wrappers), and PS (foodservice packaging). Plastics are also used in the automotive 1 Excluding textile fibers, rubber, and thermosets. Plastics are polymers, materials that are made up of repeating units of monomers. Many synthetic fibers, rubber, and thermosets are also polymers, but the term plastic is less discerning or consistent.

220 RECYCLED PLASTICS IN INFRASTRUCTURE sector (many automobile parts use PP), building and construction (PVC is used for cables, window profiles, roofing, flooring, high-density poly- ethylene [HDPE] for pipes, PUR for insulation), consumer products (toys, mattresses), electronics, and more. Plastics applications are innumerably large, and they are growing faster than our waste management capabilities. The United States is the largest producer of plastics waste in the world (both by total and per capita estimates). In 2018, the country generated 37 MMT of plastics waste, of which less than 8 percent was recycled, 15 percent was incinerated with energy recovery, and the rest was landfilled (USEPA 2020a) (with up to 0.4 MMT mismanaged and potentially enter- ing oceans [Law et al. 2020]). Plastics waste from households, termed as post-consumer plastics waste, accounting for 54 percent of all plastics waste generated (see Figure E-1), is often more diverse and mixed with other waste materials generated at home such as metal cans, paper, cartons, and organic waste. As plastics are lightweight, they typically constitute only 5 to 12 percent of household waste by weight (The Recycling Partnership 2020), despite occupying a larger volume. Post-commercial waste genera- tion sources include retail stores, hospitals, large office spaces, restaurants, and trucking and constitute 35 percent of all plastics waste generated (see Figure E-1). Some commercial entities have less mixed and more homo- genous plastics waste streams such as retail back-of-store tertiary packag- ing (mostly low-density polyethylene [LDPE]) and warehouse crates (often HDPE), while others have mixed packaging (restaurant containers and films) or contaminated waste (agricultural mulch films). Plastics waste from automotive shredder residues and construction and demolition activities constitute 7 percent (2,800 kt) and 4 percent (1,200 kt) of plastics waste generated, respectively (Heller et al. 2020; Townsend et al. 2019). Identify- ing the type and composition of plastics waste for the various generators is key to understanding end-of-life opportunities for reuse and recycling. Col- lection streams will vary significantly between residential, commercial, and other sources: residential collection takes place through municipal curbside collection or drop-off programs, while commercial collection often takes the form of individual contracts between establishments and recyclers. The State of Curbside Recycling Report (The Recycling Partnership 2020) estimates 7.5 million tonnes (out of more than 20 million tonnes of residentially generated post-consumer plastics waste) is practically re- cyclable (i.e., it can be recycled efficiently within the current recycling sys- tem in the United States without major technological advances). While all thermoplastics are theoretically recyclable, practical recyclability is judged based on acceptability in the local recycling context and depends on several factors including but not limited to resin type, product type, product vol- ume, ease of sorting, and maturity of secondary recycling processes. Table E-1 further breaks down recyclable residential waste by resin type.

APPENDIX E 221 Commercial waste generation in the United States is estimated for different business groups and plastic resin types by Meyer et al. (2020). The source data for this estimate is a one-of-a-kind commercial waste characterization study by California that reports plastics waste generation (and recycling) from commercial sources per employee for different business groups. The breakdown of approximately 12 million tonnes of plastic generated from commercial sources (not considering recyclability criteria) in the United States is reported by resin/product type in Table E-2. FIGURE E-1 U.S. plastics waste generation by source type. NOTE: kt = 1,000 metric tons; MSW = municipal solid waste. SOURCES: Automotives waste estimate from Heller er al. 2020. Construction and demolition waste estimates from Townsend et al. (2019). Commercial MSW estimate from Meyer et al. (2020). U.S. Environmental Protection Agency (USEPA) waste generation information includes residential and commercial MSW, and this is used to estimate residential MSW. TABLE E-1 Practically Recyclable Residential Waste Generated in the United States by Resin/Product Type Material Recyclable Residential Waste Generated (kt) PET Bottles 2,618 Non-Bottle PET 553 HDPE Natural Bottles & Jars 542 HDPE Pigmented Bottles & Jars 831 Other Plastic Packaging (~#3-#7) 1,764 Bulky Rigid Plastics 1,227 Total 7,535 SOURCE: Data from The Recycling Partnership 2020.

222 RECYCLED PLASTICS IN INFRASTRUCTURE The rest of the report is organized as follows: The second section discusses the current system for collection, sorting, and recycling of recy- clable waste plastics and discusses various strategies to improve quantity and quality of the supply of recycled plastics. The third section delves into the consumption of recycled plastics in the United States and characterizes applications and actors based on their motivation, willingness, and con- straints to substitute for recycled plastics. The fourth section inspects the supply–demand gaps that exist in the recycled content market today for the more recycled products: PET bottles, rigid HDPE, and PE films. Factors introduced in the second and third sections frame this brief market review. The final section concludes by reporting on policy actions and legislative levers to mitigate supply–demand imbalances. THE SUPPLY OF WASTE-DERIVED PLASTIC FEEDSTOCKS Collection and Sources The collection of recyclable plastics waste is the first step in securing re- cycled plastic supply, and all estimates of the current system are rooted in the U.S. Environmental Protection Agency’s (USEPA’s) Advancing Sustain- able Materials Management annual study (USEPA 2020a). Between 2000 and 2018, annual plastics waste generation has increased from 23 to 33 TABLE E-2 Commercial Plastics Waste Generation Plastic Type Commercial Generation (kt) Durable Plastic Items—Number 2 and 5 Bulky Rigids 411 Durable Plastic Items—Other 1,488 Film Products 67 HDPE Plastic Containers 787 Miscellaneous Plastic Containers 626 Non-bag Commercial and Industrial Packaging Film 1,041 Other Film—Other 3,388 PETE Plastic Containers 1,070 Plastic Grocery and Other Merchandise Bags 318 Plastic Trash Bags 3,157 Total 12,353 NOTES: HDPE = high-density polyethylene; kt = 1,000 metric tons; PETE = polyethylene terephthalate. Commercial plastic waste generation as estimated by the authors from data in California’s commercial recycling survey (CalRecycle 2015) using an approach similar to Meyer et al. (2020).

APPENDIX E 223 million tonnes, while the fraction collected for recycling has gone up from 5.3 to 8.5 percent (2,800 kt in 2018). While USEPA uses a hybrid top-down and bottom-up approach using plastics waste production, use, and lifetime statistics alongside field surveys and reports breakdown by plastic resin/ product type, uncertainties in these data are profound. Figure E-2 shows generation and collection for recycling by plastic resin type. Almost a third of the collected plastic is of an undetermined type (mostly from durable goods such as appliances and furniture), and it is further unclear whether it is collected for recycling or simply represents other channels for disposal. The largest known category of plastic that is recycled is PET (from bever- age containers), followed by HDPE (from milk bottles and other consumer product bottles like detergent) and LDPE (from commercial wraps and films). Other plastic types in municipal solid waste (MSW) are often not collected for recycling, due to distributed uses among diverse use applica- tions that hinder economies of scale or because of difficult and expensive processing requirements or both. Figure E-3 overlays recycling and combustion fractions (collectively called diversion from landfills) on U.S. plastics waste generation statistics seen earlier in Figure E-1. Residential recycling is driven by curbside pro- grams in the United States, where consumers can conveniently set out their waste plastics alongside other recyclables in the recycling bin and opt for frequent collection services. Drop-off collection services, where customers transport their waste to a central drop-off location, supplement curbside ser- vices (often for large items) or can be the only mode of collection in sparsely populated regions. Access to collection modes is a critical determinant of recycling rates in municipalities across the nation. The State of Curbside FIGURE E-2 Recycled fractions for various plastic types compared to generation volumes. NOTE: L(L)DPE = LDPE and LLDPE. SOURCE: USEPA 2020a, Advancing Sustainable Materials Management: 2018 Tables.

224 RECYCLED PLASTICS IN INFRASTRUCTURE Recycling Report (The Recycling Partnership 2020) found that more than 94 percent of households in the United States have access to some form of collection for recycling (curbside, drop-off, or both). Sixty percent access curbside services, 14 percent can choose to but do not subscribe to curbside options, while 21 percent have access to drop-off services only. What Are We Collecting? Not all plastic product types are equally collected for recycling: some are more practically recyclable than others. Curbside capture rates (fraction of waste generated by a household that is correctly put into the recycling bin) for plastic that are accepted by recycling programs vary from 22 to 53 percent (The Recycling Partnership 2020). The largest capture rates are for natural HDPE bottles (milk jugs), pigmented HDPE bottles (detergent bottles, house- hold cleaning agents containers, etc.), and PET bottles (beverage containers), respectively. Mixed plastic packaging and bulky rigid plastics have the lowest capture rates. Looking at total collection volume, PET bottles have the largest amount of collection (830 kt) followed by wrapping films (420 kt; mostly from commercial sources), pigmented/colored HDPE bottles (263 kt), and FIGURE E-3 U.S. plastics waste generation and diversion by source. DATA SOURCES: Automotive waste based on Heller et al. 2020. Construction and demolition waste based on Townsend et al. 2019. Commercial MSW generation from Meyer et al. 2020. Total MSW generation, recycling, and combustion from USEPA 2018. Assumed that commercial and residential streams have the same recycling rate: Analysis done by the authors of this report based on commercial waste data for California finds a collection rate of approximately 9 percent for commercial MSW in the United States, which is similar to the collection rate for residential MSW.

APPENDIX E 225 natural HDPE bottles (200 kt). Within the largest collected product, PET bottles, a third of the total collection comes from deposit return systems in 10 states, while the rest comes from curbside sources. While deposit return systems (DRSs; commonly known as “bottle bills”) are only applicable to PET bottles in the United States, they are very effective in increasing PET bottle collection rates (Basuhi et al. n.d.) and are discussed in detail in a later subsection as one of the proven strategies to improve collection rates. What Is the Cost of Collecting Waste Plastics? The cost of collection is a critical factor in assessing the economic viability of recycling activities, and estimates of costs must be separately made for each mode of collection: deposit, curbside, and drop-off. Costs of curbside collection vary based on diversion rates (how much was collected), the frequency of collec- tion, and whether recyclables were single stream or dual stream. The costs range from an average of US$90/ton to US$280/ton (see Table E-3) with lower costs for less frequent collection with high diversion rates. As recyclables are typically lighter in weight than trash, collecting recyclables typically is more costly than collecting trash: a study by New York City in 2016 found that it cost US$191/ ton to collect recyclables compared to US$99/ton to collect trash (IBO 2017). For beverage containers in the 10 “bottle-bill” states, handling fees ap- proximate the cost of collecting bottles for return centers (see Table E-4). Handling fees are typically charged by the state government to the bottle man- ufacturer and then paid to the return centers (e.g., grocery stores, redemption centers, etc.) as compensation for the costs of running the deposit program in their facility. They are based on the number of bottles collected, and the fees TABLE E-3 Average Costs of Curbside Collection by Collection Type Collection Type Cost per Household per Year (US$) Cost per Ton Waste (US$) Single Stream, Weekly Collection, High Diversion Rate (~40%) 54 139 Single Stream, Every Other Week Collection, Low Diversion Rate (~40%) 52 278 Single Stream, Every Other Week Collection, High Diversion Rate (~25%) 33 89 Dual Stream, Weekly Collection, High Diversion Rate (~40%) 58 141 Dual Stream, Every Other Week Collection, High Diversion Rate (~40%) 45 103 SOURCE: Data from USEPA 2016.

226 RECYCLED PLASTICS IN INFRASTRUCTURE range from 1.5 to 4 cents/bottle, which translates to a cost between approxi- mately US$400/ton and US$1,200/ton. While deposit return systems have higher collection costs than curbside programs (typically less than US$300/ ton), they do not require additional sorting and yield a source-separated low- contamination, higher-quality, higher-value material. What Is the Cost of Sorting Waste Plastics? In the United States, more than 80 percent of household curbside recycling is single-stream commingled collection (Koerth 2019). A critical step in the end-of-life plastic recycling supply chain is the sorting of post-consumer plastics waste from commingled recyclables, which takes place in materials recovery facilities (MRFs). A combination of automated sorting equipment for cardboard, paper, glass, metals, and plastics and manual quality control is employed to produce material-specific bales of recyclables of various grades. The sale of MRF-separated recyclables is usually not enough to cover the cost of sorting and separation. Therefore, MRFs typically have contractual agreements with several municipalities (households) and commercial entities and charge “processing” fees (on a mass or household basis) for their service. Processing fees are estimated based on incoming volume and composition, operating costs, and market value of material bales and vary widely across the United States depending on the variation in these factors. Aggregated average MRF processing fees for various USEPA Regions are shown in Figure E-4. MRFs must adapt to the long-term changing waste composition and market conditions to remain viable, and many MRFs evolve along with TABLE E-4 Handling Fees in U.S. States That Have Deposit Return Systems State Handling Fee (US$/bottle) IA 0.01 CA 0.016 CA 0.024 CT 0.02 MA 0.0225 VT 0.032 MN 0.033 HI 0.03 NY 0.035 ME 0.04 MA 0.0325 SOURCE: Container Recycling Institute 2022.

APPENDIX E 227 their recyclable inputs by adding extra sorting equipment, adding extra conveyor lines, and upgrading to more automated workflows to improve sorting efficiencies and lower costs (such as labor or overhead). However, with the growth in use, cases of light plastic materials, for many MRFs, the evolving tonne is progressively decreasing in value evaluated per tonne of output (RRS 2018). In the short term, volatile market conditions can greatly exacerbate this concern: when bale prices fall, revenue is further eroded. Rising transportation or fuel costs and contamination can also make matters worse. Figure E-5 shows a map of the United States with all curbside program closures in the period between 2018 and 2023 along FIGURE E-4 MRF processing fees (in US$/tonne) per region. SOURCES: Data from The Recycling Partnership 2020. Map adapted from USEPA. FIGURE E-5 Locations and reasons for curbside programs shutting down across the country between 2018 and 2023. SOURCES: Data from Waste Dive Team 2019.

228 RECYCLED PLASTICS IN INFRASTRUCTURE with reasons for closure. The often-cited reason is cost; typically, MRFs make 3- to 5-year contracts with municipalities, but the fluctuating market landscape has made the case for more flexible contracts and renegotiations to keep MRFs, which perform a vital service for the municipalities, afloat. Recycling markets faced economic pressures when China’s import restric- tions were put into effect in 2018, as they lost a ready market for low-value recyclables (such as mixed paper and mixed plastics waste) that otherwise would have to be landfilled, incurring tipping fees. Why Plastic Recycling Is Challenging: Contamination and Quality Contamination is a critical challenge to plastics waste recovery and subse- quent recycling. Contamination can be due to product function, product design, or collection mode; contamination from product function includes wear and tear, residues from small molecules from food, cosmetics, personal and household care formulations, and so forth and is inevitable to some ex- tent. Several washing steps can be employed to remove these contaminants, but washing (water-intensive) must be followed by drying (energy-intensive) because moisture can cause property degradation during mechanical re- extrusion. Contamination due to product design is intrinsic in nature and may require simple sorting techniques (density separation can separate PET bottles from PP caps because PET flakes are heavier than water, while PP is lighter) or complicated multistep processes (multilayer films laminated together require careful solvent extraction for each layer and subsequent energy-intensive separation of the polymer dissolved in the solvent). The latter is expensive in terms of energy and cost and renders much of the fraction of laminated films practically unrecyclable. Simplifying packaging design, eliminating overpackaging; using recycling-friendly inks, sleeves, and adhesives; and developing easy-to-disassemble composite packaging are some strategies to lower the possibility of intrinsic contamination at end of life. However, this requires iterative coordination of upstream product and packaging design with downstream waste management activities; in the United States, the Association of Plastic Recyclers (APR) has undertaken this task to produce their APR design for recycling guide (APR 2022a). Contamination due to collection mode is what is conventionally referred to as contamination, where unwanted materials are added to the recyclables bin (before sorting, due to consumer error, lack of knowledge, or careless- ness) or find their way into recycled bales (after MRF sorting due to imper- fect sorting technology). The immense diversity of materials and products and differing standards across municipalities means that it is difficult for consumers to easily know what is recyclable and what is not. A survey of 212 cities in California, Oregon, and Washington by The Recycling Partner- ship (Tanimoto 2020) found that inbound contamination to MRFs is usually

APPENDIX E 229 10 to 20 percent and can be as high at 46 percent in some cases. Inbound contamination increases the amount of residual waste that a MRF must pay to sort and dispose and affects cost along with recycling efficiency. For plas- tics, most automated sorters observe product attributes such as one or more of color, shape, size, and structural and chemical composition. The product attributes are measured using cameras (visual imaging) and spectroscopic techniques (hyperspectral imaging), and a combination of image processing and chemometrics is used to identify target polymer product type and sort it to the correct stream. With increased material diversity, sorting algorithms can be confused; for instance, polylactic acid (PLA) polymer is a biobased alternative to PET for some applications, but the incorrect sorting of PLA to the PET waste stream can have detrimental effects in downstream recycling (Gere and Czigany 2020). Similarly, a full-cover PE sleeve may register a PET bottle as PE and sort it into the residual stream. Putting It All Together: What Does the Plastic Recycling Value Chain Look Like? Post-consumer and post-industrial plastics waste follow different paths to- ward reuse in new products. As described in the previous section, contami- nation is a determinant in the complex recycling process of post-consumer plastics waste, as well as potential product quality issues. Depending on the end-market, quality deficiencies coming from post-consumer recycled plastics are overcome by introducing virgin polymers. On the other hand, post-industrial plastics waste typically possesses low levels of contamina- tion and, therefore, adequate quality, making its entry into the value chain direct and without the need of virgin polymer additives to improve quality. The flowchart in Figure E-6 illustrates the steps and outcomes in the plastic recycling value chain. FIGURE E-6 Plastic recycling value chain.

230 RECYCLED PLASTICS IN INFRASTRUCTURE What Are Some Viable Strategies for Improving Collection Quantities? Deposit Return Systems While most DRSs originated to manage glass bottles and metal cans, usu- ally from alcoholic beverages, they have generalized the system to include PET bottles used for soft drinks as their volume in the waste stream has grown exponentially in the past 40 years. A DRS incentivizes a customer to return their empty containers to a participating retail location or deposit center by adding redeemable deposit fees to the purchase of the product in deposit-eligible containers. The observed impact of DRS in PET collection is significant: 50 percent of U.S. post-consumer PET comes from DRSs, even though it is present in only 10 states (consisting of 34 percent of the na- tion’s population) (Smith et al. 2022). In fact, PET from deposit schemes is 25 percent of all plastic collected and recycled in the country, even though PET bottles comprise roughly 1 to 2 percent of all plastics waste generated (USEPA 2020b). States with bottle bills recycle about three times more per capita on average than states without. However, states vary widely in their design of the deposit system, and several parameters affect the outcomes of the system. One important parameter is the deposit fee, which varies depending on the state (5 to 15 cents) and sometimes, depending on type of product (i.e., 15 cents for liquor containers in Maine and Vermont) or size of container (larger bottles have a 10-cent deposit). Higher deposit fees typically correlate with a higher redemption rate (percentage of eligible containers returned): Michigan with a 10-cent deposit has a 90 percent plus redemption rate and in Oregon, when the deposit fee was increased from 5 to 10 cents in 2017, studies found that the recycling rate had increased from 64 percent (measured in 2016) to 86 percent (in 2019) (Profita 2019). More than 40 countries besides the United States have deposit programs, and many in the European Union have deposit fees set above 10 cents (ad- justed for price parity) and boast high redemption rates (80 to 95 percent) (Reloop 2020). With abundant empirical evidence, several states (including Connecticut and Massachusetts) have considered increasing their deposit fees, and the Break Free from Plastic Pollution Act2 has suggested imple- mentation of a national deposit return scheme with a 10-cent deposit to boost collection of PET bottles. To be successful collection mechanisms, DRSs should focus on improv- ing coverage, connectivity, and convenience. Coverage denotes the fraction of all containers that are eligible for return; this number can be as low as 50 percent in several states (Container Recycling Institute 2022) where 2 Break Free from Plastic Pollution Act of 2021. H.R. 2238, 117th Congress (2021-2022). https://www.congress.gov/bill/117th-congress/house-bill/2238/text.

APPENDIX E 231 only carbonated drinks are included (some allow bottled water as well). As deposit fees can be viewed as a tax on consumption (despite redeem- ability), expansion to many food products can be controversial, and other products may have greater food residues that yield contaminated feedstock for recycling. Connectivity and convenience are linked: while it is always easier to throw a bottle into the recycling bin, a well-designed network of deposit centers or a participating retail location that optimally serves the population density of the region can encourage return behavior along with a reasonable deposit fee incentive. When consumers do not return the bottle (by throwing it in curbside recycling, a trash bin, or littering), “unredeemed” fees accumulate, and states differ in the handling of these unredeemed deposits. Iowa allows pro- ducers to keep it and operate the deposit system, whereas in Connecticut, Massachusetts, and Maine, the state uses it for general funds; New York gives producers a fraction to cover system costs while the state uses the rest; Michigan gives the retailers a fixed percentage to cover handling costs; and yet others, like California and Vermont, earmark it for specific purposes such as a beverage container recycling fund or clean water programs (Re- loop 2020). California is also unique in having a centralized system that operates and administers the deposit system through CalRecycle, while in most other states, the beverage industry and sometimes collection logistics partners operate the system in a decentralized manner. The operational and financial characteristics of a deposit system play an important role in determining practical recycling rates, and careful assessment of these factors is needed to devise optimal nationwide strategies for improved collection outcomes. Extending deposit systems to containers beyond PET bottles is slowly gaining momentum (e.g., some provinces in Canada have one for HDPE milk bottles [Morawski 2020]), but without a steady market, the proliferation of broader container schemes will struggle to be cost effective. Pay-As-You-Throw Policies Another strategy used to increase collection rates is a pay-as-you-throw (PAYT) program. Instead of residential customers paying the traditional tax or flat fee for waste disposal, customers pay a variable rate based on the amount of waste disposed of. This variable rate pricing is like the rates for other utilities (e.g., electricity and water). The minimum participation rate in the Netherlands was found to increase when curbside recycling was replaced with drop-off centers and when a PAYT program was introduced (Thoden van Velzen et al. 2019). In the United States, more than 7,000 cities and towns (e.g., Austin, Portland, Seattle) have developed PAYT programs to reduce waste genera- tion and improve recycling rates. In Massachusetts, towns with the PAYT

232 RECYCLED PLASTICS IN INFRASTRUCTURE program disposed of 1,239 pounds of trash per household compared to an average of 1,765 pounds per household in towns without the program. USEPA reported that communities with PAYT programs reduce the amount of waste disposed of by 14 to 27 percent on average. Residents reduced their waste generation through increasing their recycling rates, reusing or donating goods, and/or composting. PAYT programs also improve sorting rates and overall recycling rates, particularly in municipalities with low sorting and recycling rates of plastic packaging (Saure 2018). The main obstacle in implementing PAYT programs is setting the PAYT fees. The fee must be high enough to influence behavioral change. However, a fee that is too high may result in illegal dumping and burning of waste. In the United States, municipalities must also consider legal and jurisdictional ordinances before implementing a PAYT program to set a variable rate, set weight limits on disposal containers, and create bans on illegal diversion. What Are Some Viable Strategies for Improving Feedstock Quality? Initiatives for Waste Source Separation Besides increasing collection quantity by incentivizing return, deposit systems provide source-separated PET bottles that are considered higher quality than their curbside-collected counterparts. There is growing awareness that single-stream recycling, which is convenient to custom- ers, has also increased contamination due to lack of source separation (Damgacioglu et al. 2020; Tonjes et al. 2018). Some cities are removing single-stream programs (Waste Dive Team 2019) because there has been reduced quality in all recycling materials, including mixed paper and glass, reverting to dual-stream recycling or creating recyclables drop-off programs (Rosengren et al. 2019). In a dual-stream system, consumers sort their paper, plastic, and glass into separate bins before it is picked up. This system reduces contamination and improves the quality of col- lected materials. In Germany, where recycling rates are high, consumers follow a waste hierarchy and sort their own recyclables into multiple bins (Keramitsoglou and Tsagarakis 2019). Advanced Waste Sorting In the absence of source separation of waste, advanced waste-sorting prac- tices are needed to recover greater fractions from the mixed recyclable waste. We consume a wide variety of products made of various types of materials. While advanced reprocessing technologies are being developed rapidly, sorting of product fractions remains a critical bottleneck in obtain- ing quality feedstock for reprocessing. 

APPENDIX E 233 Optical sorting works remarkably well for some large-volume plastic products such as PET bottles (e.g., carbonated soft drinks, water bottles), as the system is optimized (by the equipment manufacturer) for select products. However, the technology is unable to keep pace with changing, ever-so-diverse waste streams, and flexible solutions that can be customized to the needs of the facility are needed. Many technology providers are har- nessing computer vision and other data-driven methods based on observed product attributes to complement real-time sorting decisions. While most methods remain proprietary, a preliminary model called CircularNet pub- lished by TensorFlow (Sanjeev et al. 2022) highlights the possibilities and challenges of the approach. To scale such methods, gathering and collation of useful training data are paramount. We need robust data collection and sharing policies that involve all relevant stakeholders. Furthermore, insights from models can be further used to inform upstream design for recycling about problematic shapes, sizes, colors, and additives and accelerate the improvement in the overall recyclability of plastic packaging. Parallel to this development, digital tracers that can be used to identify plastic products are also being investigated in pilot projects (Gasde et al. 2021; Schmidt et al. 2021). Feedstock Recycling Some plastic products such as multilayer films are difficult to recycle con- ventionally, as the different components are incompatible under thermome- chanical processing. Even perfectly sorted and separated plastics undergo quality degradation (Schyns and Shaver 2021) in repeated mechanical re- cycling experiments as thermal and shear forces during reextrusion cleave or extend chains, impact molecular distributions, and consequently affect processing and mechanical properties. For such multimaterial products that cannot be recycled conventionally or monomaterial products that have un- dergone molecular degradation, chemical and feedstock recycling processes provide new pathways to circularity by breaking down the long-chain polymer molecules into monomers (depolymerization), oligomers and other derived small molecules (chemical recycling), or fuel-range hydrocarbons (pyrolysis, gasification). The classification of feedstock recycling processes varies widely; some categorize pyrolysis under chemical recycling because fuel feedstocks can be converted to chemicals via additional cracking and synthesis steps. However, if the end products are fuels that are not intended for chemicals or plastics production, it is unclear how to make the distinction. The envi- ronmental benefits for pyrolysis are considerably lower than for mechanical recycling and may even be worse than the status quo depending on yields, and experimental conditions such as temperatures and pressures; however,

234 RECYCLED PLASTICS IN INFRASTRUCTURE development of low-temperature processes aided by catalysis is a promising pathway for PE and PP wastes. Regulation that conflates plastics-to-fuels and plastics-to-plastics pathways may disincentivize materials circularity. States vary in whether they consider advanced recycling a waste manage- ment activity or a manufacturing activity and whether producing fuels from plastics can be classified as recycling. Currently, Ohio, Texas, and Iowa allow advanced recyclers to produce fuels, while states like Kentucky and Arkansas prohibit energy recovery or fuels production from being grouped with advanced recycling (Hogue 2022), and yet others lack any specificity on the matter. Such variations mean that recycling targets and the roadmaps to achieving them will look significantly different in different states. Many advanced recycling technology providers exist in the United States (Closed Loop Partners survey and summarize 60 of them [Closed Loop Partners n.d.]) but most are at pilot stages or operate small-scale facilities. PET, a condensation polymer, undergoes relatively facile depoly- merization (Barnard et al. 2021) (back to monomers) with relatively mild reagents: water (hydrolysis), methanol (methanolysis), glycol (glycolysis), or enzymes (enzymatic recycling). These monomers can be polymerized back to PET in the same way as virgin PET production, and the process has been demonstrated at scale. Other than PET depolymerization and polyolefins-to-fuels production, the PS-to-styrene (monomer) pathway has been developed and deployed by Agilyx with a capacity of 10 tons of PS per day (Pyzyk 2020). Other monomer recovery activities are usually laboratory scale and face several scale-up challenges, including low yields, high post-processing and separation costs, and lack of large-volume point sources. It is unclear what degree of mixed waste can be economically converted back into feedstock. Closed Loop Partners survey costs for these technologies and the range of capital expenditures (CAPEX), operating expenditures (OPEX), as well as feedstock cost drivers are summarized in Table E-5 below. TABLE E-5 Cost Estimates for Feedstock Recycling Technologies Technology CAPEX (US$/ tonne) OPEX (US$/tonne) Revenue (US$/tonne) Feedstock Cost, Comments Pyrolysis 2,000-2,700 250 560 Assumes very low feedstock costs of less than US$20/ tonne PET Depolymerization 1,300-2,300 1,200 1,400 Feedstock costs account for a fourth of the operating costs

APPENDIX E 235 The economic viability of these processes depends critically on feed- stock cost, and many assume the current costs of obtaining feedstock when it is low quality and/or has undeveloped recycling systems but have not considered the cost of developing out and scaling up reverse logistics and collection end points for less-desired feedstock they claim to recover and recycle; as such this may cause increased competition for existing limited supply of recycled plastics. CONSUMPTION OF RECYCLED PLASTICS Secondary End Uses for Recycled Plastic Recycled plastics are used in a host of applications to supplement or dis- place virgin plastics: much of this demand has historically been driven by the low price of recycled plastics. Recycled PET as polyester in textiles or carpets or PE in municipal pipes and decking were some of the largest end users in the United States in 2018 (see Figure E-7). While the recycled plas- tic usually originates from packaging applications (PET and HDPE bottles, films, other rigids), the end uses were often consumer or institutional products that could tolerate a greater degree of mixing and contamination. This has led to concerns about “downcycling”—loss of economic value as well as technical properties. Many have ascribed lower environmental benefits to “downcycling” as compromised quality in these applications does not allow further recycling and the recycled product is often destined to be landfilled. The concept of product circularity espouses keeping the material in the original application (in this case, packaging), maintaining FIGURE E-7 Distribution of end-use markets for (a) PET, (b) HDPE, and (c) film packaging. NOTE: NAPCOR reports that in 2021 more PET went into bottle end uses than fiber end uses (NAPCOR 2022). SOURCES: NAPCOR 2018 (PET); APR and ACC 2019 (HDPE); and More Recy- cling 2018.

236 RECYCLED PLASTICS IN INFRASTRUCTURE quality, and recycling several times over. A more careful approach (Geyer et al. 2016) shows that “downcycling” is not inherently environmentally inferior and product circularity must be less energy- and emissions-intensive than virgin production to be sustainable. From a market perspective, demand from users driven by low price has been unable to motivate higher recovery rates at higher costs to the recy- cling system. The loss of economic value and lack of demand pull provide little incentive for materials recovery facilities or municipalities to improve plastics recovery or quality when high-quality applications such as food- grade bottles were a minority. With product circularity, this trend has buck- led: growing sustainability expectations from packaging or policy-driven pressures from recycled content mandates (such as one in California or the European Union) and anticipated eco-modulated packaging Extended Producer Responsibility (EPR) laws (in several U.S. states [Sustainable Packaging Coalition 2021]) have created demand for high-value recycled plastics. In 2020, for the first time, more recycled PET was incorporated in bottle end uses than in fiber and sheeting end uses (NAPCOR 2022). With limited supply, this has raised prices for PET, HDPE, and films and priced out some of the original low-quality, highly price-elastic end users. Data for higher-quality end uses is also more widely available than for low-quality end uses. Many niche and nascent uses are less well documented and we discuss them briefly in the market for mixed plastics. Exports of waste plastic have plummeted over the past decade, driven in large part by a greater demand for high-quality plastics by domestic ac- tors and China’s 2017 National Sword Policy that placed restrictions on import of mixed and contaminated waste (see Figure E-8). In 2021, the United States exported a third of the scrap plastic it did in 2017. From FIGURE E-8 Percentage of plastics waste collected for recycling domestically and exported outside the United States and Canada. SOURCES: Data from NAPCOR (2018) and Stina (2021).

APPENDIX E 237 2017 to 2018, PET exports dropped from 16.4 to 7.7 percent of the to- tal PET bottles collected (more than half of the 7.7 percent was going to Canada). Similarly, from 2016 to 2018, exports of recovered film bales halved. With increasing international legislation on the trade of plastics waste, it is likely that exports will be a smaller fraction of end-use markets for recycled plastic. MARKET MATCHING OF SUPPLY AND DEMAND This section discusses the balances and imbalances associated with recycled plastics supply and demand. The way supply interacts with demand and how both respond to price is key to designing effective policy. The section presents learnings from the market interactions of the three most recycled plastics: PET, HDPE, and LDPE. Quality is an important factor that in- forms end-use opportunity for demand as well as technological improve- ment for supply; this bridging element is discussed in the context of polymer chemistry, and novel advances in chemical recycling are highlighted. The Importance of Quality in Demand–Supply Matching The previous section provides a rough idea of which demand end uses have a stringent need for high-quality recycled feedstock; however, quantitatively defining quality with adequate generality across plastics or even a single type of plastic is challenging. From a demand perspective, quality must be defined with respect to an application and is contingent on the proper- ties critical for the processing as well as safe and robust use of the plastic product/packaging that incorporates recycled content. From a supply per- spective, loss of quality can be linked to accumulation of contaminants during collection and degradation of polymer chains under high-stress and high-temperature conditions of reextrusion in mechanical recycling. Coming out of the MRF or deposit center, recycled material grades use the percentage and type of contamination in plastic bales to characterize sup- ply (e.g., model bale specifications from the Association of Plastic Recyclers [APR 2022b] list 24 different specifications for Nos. 1-7 plastics). While this harmonizes expectations between sellers and buyers for baled recycled ma- terial, further material quality checks are performed by end users to ensure their processing and application-specific properties are being met. Often, recycled material falls short of virgin properties that were tailor-made and optimized by larger chemical manufacturers for that specific application. In many cases, recycled feedstock is blended and diluted with virgin plastics or the product design is varied to accommodate the property mismatch. Food- grade use of plastics adds regulatory requirements to quality constraints to ensure food safety of packaging: the U.S. Food and Drug Administration

238 RECYCLED PLASTICS IN INFRASTRUCTURE (FDA) issues No Objection Letters (FDA 2022) to recycling companies after evaluating recycling process, feedstock source, decontamination steps, and intended use case conditions (temperature, pH, contact time) of the food product. Due to this rigorous assessment, in most cases, food-grade is the highest grade of recycled plastic available and demand for this feedstock from both food and nonfood packaging industries has risen dramatically (Schneider 2021a). Higher demand with limited supply has increased prices for food-grade HDPE and PET, and this higher value has incentivized many recyclers to improve the quality of their feedstock and recycled products. In a perfectly circular scenario, all feedstocks will be high quality and all products will be infinitely recyclable into themselves. Far from perfect circularity, we have both lower-quality sources of supply (either due to mixed collection or product-level contamination) and some demand actors who can tolerate this lower quality of recycled material to various extents with some degree of “downcycling.” Adequately characterizing this spec- trum of quality considerations for each plastic type can allow us to match demand and supply more efficiently, drive up overall recycling rates, divert more plastics from landfills, and promote broader economy-wide materials circularity which may come at the cost of individual product circularity. HDPE provides a good example of cascading demand. The demand for high-quality plastic is usually driven by consumer pressure and existing (California 25 percent recycled content mandates [CalRecycle 2022a]) or anticipated (Break Free from Plastics Pollution Act of 2021) regulation and tends to be more price-inelastic than noncircular low-quality applications. However, on the supply side, high-quality plas- tics are scarce and lower-quality secondary feedstock is abundant. While converting the lower-quality feedstock to higher-quality recycled plastics should be prioritized, processes can be cost-prohibitive and energy-intensive (e.g., multilayer film separation, mixed rigids, etc.), and research suggests that mechanical recycling degrades properties after several cycles, making it unfit for high-quality applications. Whether low-quality plastics waste is a result of imperfect collection or sorting practices, or multicycling, the current strategy has been to wait for better technologies (advanced sorting, chemical recycling, etc.) to catch up. However, these technologies may be decades from maturation and high efficiencies for many of the plastic resins other than PET. In the meanwhile, developing demand and robust markets for the lower-quality feedstock is important to incentivize their collection and diversion from landfills and the natural environment. The Market for Recycled PET PET bottles have the most mature recycling system today and, consequently, the highest recycling rates (23 percent) among all plastic products in the

APPENDIX E 239 United States. In other words, recycled PET (R-PET) has well-defined supply pathways. However, the demand for R-PET exceeds the currently available supply. Demand from beverage manufacturers who have made recycled con- tent commitments, combined with legislative mandates, have increased circu- lar demand for high-quality R-PET significantly. Polyester, which constitutes more than 50 percent of textile material globally, refers to fibers made from PET; textile manufacturers have also made their own sustainability commit- ments, but textile recycling is still nascent, and they rely on bottle-derived R-PET. Combining these factors, the expected demand for recycled PET is more than three times larger than the current supply (see Figure E-9). Increased competition for limited supply has led to large price premiums (from −25 percent in 2018 to +27 percent in 2020; see Figure E-10a) for food-grade R-PET (Leardini 2020); beverage producers are concerned that sustainable clothing brands can afford large margins and out-price them in the market (Schneider 2021b). Competition within the beverage industry has so far improved supply only marginally as larger actors are integrating recycling facilities within their value chain (Closed Loop Partners 2022). The total recycled content has not necessarily gone up; the recycled materials get redistributed among companies that are willing and able to pay more. From 2020 to 2021, Coca Cola was able to double its recycled content (from 10 to 20 percent), while other companies, like Niagara Bottling, reduced recycled content (CalRecycle 2022b) (see Figure E-10c). California law requires 15 FIGURE E-9 Comparing demand growth from 2020 to 2030 with recycled supply and disposal quantities. NOTE: Bottle and fiber demand are estimated using recycled content commitments publicly made by large companies (RRS 2020). California has mandated 50 percent recycled content in bottles by 2030 (CalRecycle 2022a). SOURCES: Data from CalRecycle 2022a and RRS 2020.

240 FI G U R E E -1 0 (a ) Pr ic es f or f oo d- gr ad e R -P E T a nd v ir gi n PE T in t he U ni te d St at es ( W es t C oa st ) fr om 2 01 8 to 2 02 0. ( b) C ol le ct io n ra te o f po st -c on su m er P E T b ot tl es f ro m 2 01 0 to 2 02 0. ( c) T he p er ce nt ag e re cy cl ed c on te nt u se d in P E T b ot tl es in C al if or ni a by t he fiv e la rg es t be ve ra ge p ro du ce rs in 2 02 0 an d 20 21 . N O T E S: D at a re po rt ed b y be ve ra ge m an uf ac tu re rs t o C al R ec yc le . B ill 7 93 w as p as se d in S ep te m be r 20 20 , r eq ui ri ng a m in im um o f 15 p er ce nt r ec yc le d co nt en t fo r bo tt le s by 2 02 2 (C al if or ni a A B 7 93 , 2 02 0) . T he s iz e of t he c ir cl e re pr es en ts t he a bs ol ut e qu an ti ty o f re cy cl ed P E T u se d by t he c om pa ny . T he d ot te d bl ac k lin e re pr es en ts t he c ha ng e in t he r ec yc le d co nt en t of a ll bo tt le s in C al if or ni a. V -P E T = V ir gi n PE T. SO U R C E S: ( a) D at a fr om I nd ep en de nt C om m od it y In te lli ge nc e Se rv ic es . In e ar ly 2 01 9, t he p ri ce o f re cy cl ed P E T b ec am e gr ea te r th an t ha t fo r vi rg in P E T o w in g to h ig h de m an d fo r re cy cl ed c on te nt . (b ) T he c ol le ct io n ra te h as b ee n st at ic e ve n th ou gh d em an d ha s be en in cr ea si ng ( N A PC O R 2 01 8, 2 02 0) . ( c) C al R ec yl e 20 22 b.

APPENDIX E 241 percent recycled content in PET bottles by 2022 (50 percent by 2030) with a penalty of 20 cents/pound. Without unlocking supply, such a policy may, therefore, inadvertently end up penalizing smaller companies who cannot af- ford to pay price premiums. Historically, a quality divide in waste PET supply made only source-separated deposit collected bottles well suited for FDA- approved food-grade resin. However, to ease the demand pressure, more curbside PET bales are being processed further up to food-grade R-PET, but the transformation is slow and has not kept up with rapidly rising demand. Tracing back the recycled PET value chain, the “suppliers” are consum- ers who often have no direct incentive to respond to increased prices in the R-PET market. As such, the price feedback, existing in many economic mar- kets, is broken. Research finds that in places where consumers have a direct incentive to recycle (bottle bill states), collection rates are four times higher (Edwards and Grushack 2021). Increasing the deposit value in Oregon from 5 to 10 cents increased the collection rate from 65 to 90 percent. When the market does not automatically adjust to high prices, concerted supply-side policies are needed to address the supply–demand gap. The Market for Recycled HDPE Recycled HDPE (R-HDPE) markets follow a cascading structure, as shown in Figure E-11. Natural HDPE bales (from milk jugs collected via curbside recycling) are often used by consumer goods brands for use in pigmented nonfood HDPE bottles (e.g., detergent bottles). Circular recycling opportu- nities back to food-grade milk bottles are limited due to cross-contamina- tion concerns. At end of life, pigmented HDPE bottles are often collected and reprocessed together, leading to dyes mixing into a black or brown color that is undesirable to consumer-facing brands but still applicable to end uses like municipal pipes where cosmetic concerns are minimal. This difference in usability and desirability of pigmented and natural HDPE bales has important implications for the recycled HDPE market. On the supply side, more pigmented HDPE is generated and collected than natural HDPE (although the recycling rate of natural HDPE is higher; see Figure E-11). Consequently, because consumer goods are a large, di- verse market, and have pledged voluntary commitments to increase recycled content in their products (Pyzyk 2021), the demand for recycled natural HDPE bales (to be used in both natural and pigmented HDPE bottles) is much stronger than the demand for pigmented HDPE bales (to be used in pipes). With two times more pigmented HDPE generated than natural HDPE (see Figure E-11), even if all milk jugs were collected, only 50 per- cent of pigmented bottles could be made from recycled material. A short- age of supply for natural HDPE bales leads to a large increase in prices for natural HDPE bales. In 2020, the price of natural HDPE bales was four

242 RECYCLED PLASTICS IN INFRASTRUCTURE times higher than that for pigmented HDPE bales (see Figure E-11). More recently, the prices of pigmented HDPE bales have also risen, but the sup- ply–demand gap remains wide. The HDPE market highlights the importance of quality and design in the economic value of recycled materials. Some containers now use pig- mented R-HDPE in interior layers sandwiched between virgin resin layers to meet their recycled content commitments. Targeted efforts to reduce the supply–demand gap include incorporating more pigmented material, color- based sorting of recycled HDPE, and design for recycling measures such as use of removable pigments or unpigmented material. The Market for Recycled Films Commercial films have more well-developed but opaque recycling value chains, and they have a much higher collection rate (21 percent) com- pared to residential film (4 percent) (More Recycling 2018). Commercial and industrial (C&I) collection is often enabled by backhauling capacity utilized to transport clean waste films from large retail stores, warehouses that maintain supplier relationships to large reclaimers. Commercial col- lection from smaller businesses and residential post-consumer collection FIGURE E-11 Left: Prices for natural and pigmented HDPE bales in the United States from 2009 to 2020. Right: Quantity of HDPE generation segmented by end use. NOTES: (Right) Milk jug represents natural HDPE consumption and detergent bottle represents pigmented HDPE consumption for bottles. Recycled milk jugs are used in pigmented bottles, while recycled pigmented HDPE bottles are used for pipes and other end-use applications. SOURCES: (Left inset) Generation and Collection quantities for HDPE in the United States (2018) (USEPA 2020b). Data from “Recycling Markets” website as reported on resource-recycling.com (RecyclingMarkets.net 2021). Prices have been increasing significantly, especially for natural HDPE, due to high demand.

APPENDIX E 243 are currently cost ineffective: a study in Ontario estimated the gross cost of collection and sorting via three municipal curbside and drop-off programs to be as high as 2511 CAD/tonne, with the sorted bale fetching only 29 CAD/tonne (Stewardship Ontario 2017). Initiatives such as Materials Re- covery for the Future are piloting new film recovery workflows in MRFs, but currently films are considered contamination in most MRFs across the United States due to operational challenges and lack of value for curbside film grades (Figure E-12 shows the spread of recycled and disposed films). This lack of value points to the nature of demand for PE recycled film res- ins: films lack circular recycling opportunities back to films due to the high processing performance required for blown-film molding, which is easily and consistently achieved by custom-made virgin resins. Producers of food-grade packaging (rigid and flexible), which are will- ing to pay more to meet sustainability requirements, do not typically source recycled content from films, and currently there is no film reclaimer in the United States with FDA No Objection Letter (NOL) to sell to food-contact or food-adjacent uses. This leaves noncircular end-use producers that are interested in films as a low-cost alternative, do not have any strong incen- tives to pay higher prices to meet the costs of developing the film recycling infrastructure, and would easily switch to virgin resins if cheaper. Plastic lumber, wood–plastic composites consume 47 percent of flexible film bales in the United States (More Recycling 2018), and they have a unique ad- vantage in that sales growth is driven by enhanced product performance features compared to treated wood lumber. An interesting case of film recycling is the use of recycled content (sourced from clean C&I films) in trash bags, a market that was developed by California’s requirement for manufacturers of regulated trash bags to include at least 10 percent recycled material (CalRecycle 2022c). This indicates that legislative levers and public FIGURE E-12 Quantity of films generated and recycled from various end-use applications. NOTES: C&I represents commercial and industrial end uses such as shrink wraps for tertiary packaging. C&D represents construction and demolition waste. SOURCES: Data scaled from California Commercial Study (CalRecycle 2015); USEPA waste characterization (USEPA 2020b), and National Film and Bag Report (More Recycling 2018).

244 RECYCLED PLASTICS IN INFRASTRUCTURE procurement activities that target noncircular end uses can aid development of film recycling infrastructure in the United States. The Market for Mixed Plastics The market for mixed plastics has been shrinking since export of mixed plastics waste to China was restricted but was recently partially revived by chemical recycling (Staub 2020). It is unclear whether they are using all of Nos. 1-7 or 3-7 mixed plastics or, like some plastic reclaimers (EFS Plastics), simply looking to separate out PE and PP ending up in mixed plastics as a cheaper alternative to well-sorted PE in high demand. There is increasing interest in the use of recycled plastics in newer ap- plications such as infrastructure and manufacturing. There have been 412 studies on using recycled plastics as an asphalt binder in pavements since 2000 (Ma et al. 2022), mostly focused on PET, HDPE and LDPE, and PP. A pavement trial in the Regional Municipality of Durham, Ontario, found that plastics added to asphalt binder increase the strength of the pavement yet reduce the strain tolerance. The recommendations based on these find- ings were to change the base asphalt to a softer grade (Ma et al. 2022) to allow for increased volumes of plastics. Recycled plastics have also been proposed as a substitute for aggregate in concrete, since plastics can improve concrete’s durability (Ponmalar and Revathi 2022), increase the bending resistance, and provide better abrasion and wear resistance (Baciu et al. 2022). However, the differences in physi- cal properties of recycled plastics must be considered when mixing plas- tics in asphalt and concrete, including density, melting point, degradation temperature and viscosity, and fillers that may be present. Other potential challenges include the unknown consequences and environmental impacts of long-term plastic use in infrastructure, such as fuming, leaching, and microplastic emissions (Ren et al. 2021). In additive manufacturing, the use of recycled plastics in 3D printing filament (Zhou 2022) has been proposed as a potential end use of post- consumer plastics. A collection-recycling-manufacturing model has been proposed to improve recycling rates and reuse plastics for 3D-printed parts and components (Wu et al. 2022). For example, Domingues et al. (2017) experimented with 3D printing of larger components and parts from tire and plastics waste. However, the recycling process can have an impact on material properties including changes in viscosity, molecular weight, and breaking strength due to exposure to high temperatures and extrusion of the material (Mikula et al. 2021). In both the asphalt and concrete infrastructure application as well as in additive manufacturing, the source of the plastics waste and impacts on the material properties during the recycling process must be considered.

APPENDIX E 245 Research into the sources of recycled plastic supply and effects of the recy- cling process must be studied further to understand the potential impact on material properties and the resulting end products. Long-term environmen- tal impacts of recycled plastics in infrastructure should also be considered in order to understand possible effects of microplastic emissions. SUGGESTIONS FOR POLICY This section discusses some recommendations we derive from our analysis and highlights key data gaps that are important to understand the recycling system. For R-PET, like many other recycled materials, there is low price elas- ticity of supply in recycling markets; that is, an increase in prices does not lead to a significant increase in supply. The “suppliers” of recycled material are usually consumers deciding to place the waste in a recycling bin and are largely unaffected by the market price of R-PET bales. The price feedback is broken. As a result, we find that for most high-quality recycled content de- mand, there is a shortage of supply. Policies, such as deposit return systems, that directly create economic incentives for consumers to “supply” more material are effective in improving collection rates. Deposit return systems are further effective in reducing litter and do not show a major impact in beverage consumption. Concerns about the cost of developing robust systems of recycled sup- ply have delayed decision making. In the United States, the cost burden of collection, sorting, and handling of packaging waste has been largely borne by the municipal (via contracts with MRFs, waste collection actors, etc.) or state (deposit handling fees) budgets (and indirectly, all taxpayers). In the current system, end-of-life concerns of the increasing amount of packag- ing waste remains external to the actors who produce this waste. EPR is a policy approach that aims to internalize this packaging waste externality by holding producers, importers, and distributors accountable for the plastic (broadly, packaging) waste their products generate. Currently, 4 states in the United States have legislated EPR frameworks—Oregon, Colorado, Maine, and California—and 12 others have introduced legislation for pack- aging EPR in 2022 (Sustainable Packaging Coalition 2021). Historically, recycled plastics markets have prices that are strongly correlated with virgin plastic prices (RRS 2021; Stromberg 2004) (which are strongly determined by oil prices [Issifu et al. 2021]). Inexpensive and custom-made virgin plastics also pose a threat to recycled supply; ensuring and maintaining robust demand for recycled materials is also necessary. Policies such as California’s recycled content mandates increase demand and raise prices. Currently, lower-quality (or contaminated) plastics wastes have very limited end uses and low value. Increasing the potential end-use

246 RECYCLED PLASTICS IN INFRASTRUCTURE markets for these materials through legislative requirements for noncircular uses or public procurement can incentivize processing of these materials. EPR can provide legislative instruments to link supply-side actions and demand-side safeguards through eco-modulation. California’s EPR includes an eco-modulated fee structure that rewards higher post-consumer recycled content and punishes lack of recycled content, thus promoting demand. Similarly, metrics for recyclability also influence fees. While a flat EPR fee structure can be adjusted to pay for waste management and recycling activi- ties at end of life, it does not provide any incentive to use recycled materials or change the packaging design to enable ease of recycling. Eco-modulation provides legislative levers that can potentially influence decision making about upstream packaging design that resides with producers. However, there is limited evidence of eco-modulation of packaging waste in practice (Joltreau 2022) and there is a need to quantitatively understand the link be- tween EPR fee structure, demand potential, and design toward recyclability. While EPR is primarily a policy tool to place the financial and/or opera- tional responsibility on the management of waste their products generate, it is likely that the cost of doing so will be passed on to the consumer. It is unclear whether this may negatively and unjustly impact consumers; pack- aging is found to be usually no more than 2 percent of the total costs of the product and, depending on consumer price elasticity, usually less than 100 percent of the increase in costs is passed on by the producer. A study by Columbia University (Bose 2022) focuses on grocery spending and finds that if EPR compliance costs lead to doubling of packaging costs to the producer, a consumer is likely to observe that their spending increased by less than 0.7 percent (on average US$4 per month per household). Operationalizing an EPR scheme often requires strong take-back struc- tures or collection schemes that can provide the right incentives to everyone (including consumers) in the value chain. Extending deposit return struc- tures to other plastic products and packaging requires careful assessment of the cost of incentivization, implementation, and monitoring compared to the value proposition of the take-back structure itself. Understanding the interplay of supply and demand, institutions, and policy in a rapidly evolving recycling landscape requires market-based modeling approaches that can simulate the behavior and interactions of actors across the supply chain. The data that can inform these approaches are often missing, and there lies great uncertainty in both economic and environmental assess- ments of plastics waste management choices today. Information Gaps and Summary In this report, we have attempted to summarize information about plastics recycling markets scattered across academic and gray literature. In our

APPENDIX E 247 literature review and analysis, we find a general need for models that ad- dress the interplay of policy and technology levers to identify cost-efficient pathways to improving recycling outcomes in the United States; to develop such models more granular and contextual data are needed. Below, we highlight some of the data gaps we think are critical to developing a more complete understanding of recycled plastics markets. Municipal Waste Collection While there is national-level information about waste collection and some state-level analyses, organized municipal-level data on waste collection re- main rare. Some state environmental departments (e.g., California) release county- or municipal-level waste generation and collection estimates, but comprehensive data across the United States is nonexistent. Data about participation rates, municipal- or county-level policies such as collection frequency, funding models, MRF contracts, accepted materials, and so forth are lacking. Such information can be instrumental in quantifying the impact of policies in improving collection and recycled materials supply. Under- standing the effectiveness of these policies will go a long way in designing policies and determining strategies to overcome the main bottleneck in the recycling system: stagnant collection. While new infrastructures and standards are likely needed to measure and collect some of the data described above, other kinds are traceable but are currently not reported openly. For example, DRSs involve a mon- etary exchange between municipalities, manufacturing, and redemption centers, and budgets and accounts can be queried to track bottles collected at redemption centers. In turn, this can help better design the network of deposit centers. For sorting, data on the costs of individual MRFs may be proprietary, but processing fees, paid by municipalities to MRFs to cover their cost, serve as a proxy to analyze changing trends in packaging waste recyclability, MRF viability, and system effectiveness. For successful EPR implementation, collation and tracking of information on collection and sorting of plastics will be imperative. Commercial Waste Collection A large amount of waste generation and diversion takes place from com- mercial sources, but very little information exists about the volumes of these flows. California conducted a waste audit of commercial facilities in 2014 (and has mandatory commercial recycling rules) but remains the only state to release this information. A better understanding of waste types and quantities from commercial sources as well as the economics of waste col- lection in business-to-business contracts can improve overall recycling rates

248 RECYCLED PLASTICS IN INFRASTRUCTURE by targeting point sources with promising economies of scale and existing logistics networks. End-Use Demand To comment on the economic viability of the system, it is crucial to know how much the different end-use markets might be willing to pay for recy- cled materials in different scenarios. Very little is known about the demand for various grades of plastics by end users. We use price data with recycled plastic consumption trends to tease some insights into demand. However, better information is needed to predict future demand and understand the price elasticity of demand. While an individual company may not be will- ing to report how much they are willing to pay for plastics, information on the demand of different industries on aggregate should be collected and reported. REFERENCES Association of Plastic Recyclers (APR). 2022a. APR Design® Guide. https://plasticsrecycling. org/apr-design-guide –––. 2022b. Model Bale Specifications. https://plasticsrecycling.org/model-bale-specifications Association of Plastic Recyclers (APR) and American Chemistry Council (ACC). 2019. 2018 United States National Post Consumer Bottle Recycling Report. https://plasticsrecycling. org/images/library/2018-postconsumer-bottle-recycling-report.pdf Baciu, A. M., Kiss, I., Desnica, E., and Sárosi, J. 2022. Reinforcing concrete with recycled plastic wastes. Journal of Physics: Conference Series 2212:012031. https://iopscience.iop. org/article/10.1088/1742-6596/2212/1/012031/pdf Barnard, E., Arias, J. J. R., and Thielemans, W. 2021. Chemolytic depolymerisation of PET: A review. Green Chemistry 23:3765-3789. Basuhi, R., Bhuwalka, K., Roth, R., and Olivetti, E. n.d. Evaluating the cost of high recycling rates under uncertain demand conditions: A case study of PET bottles in the United States. In preparation. Bose, S. 2022. Economic impacts to consumers from extended producer responsibility (EPR) regulation in the consumer packaged goods sector. [PowerPoint presentation] September 28. https://doi.org/10.7916/n2af-vv87 California AB 793. 2020. Recycling: Plastic Beverage Containers: Minimum Recycled Content. https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=201920200AB793 CalRecycle. 2015. 2014 Generator-Based Characterization of Commercial Sector Disposal and Diversion in California. https://www2.calrecycle.ca.gov/WasteCharacterization/ PubExtracts/2014/GenSummary.pdf –––. 2022a. Plastic Minimum Content Standards (AB 793). https://calrecycle.ca.gov/ bevcontainer/bevdistman/plasticcontent –––. 2022b. Plastic Recycled Content Reporting. https://calrecycle.ca.gov/bevcontainer/ bevdistman/plasticrpt –––. 2022c. Recycled-Content Trash Bag Program. https://calrecycle.ca.gov/buyrecycled/ trashbags

APPENDIX E 249 Closed Loop Partners. 2022. PepsiCo Beverages North America Invests $35 Million to Help Close Gap in Recycling Access Through Investment in Closed Loop Local Recycling Fund. Closed Loop Partners. https://www.closedlooppartners.com/pepsico-beverages- north-america-invests-35-million-to-help-close-gap-in-recycling-access-through-invest- ment-in-closed-loop-local-recycling-fund –––. n.d. Assessing Molecular Recycling Technologies in the U.S. and Canada. https://www. closedlooppartners.com/wp-content/uploads/2022/04/AR-report-V25_final.pdf Container Recycling Institute. 2022. Redemption Rates and Other Features of 10 U.S. State Deposit Programs. https://www.bottlebill.org/images/Allstates/10_States_Summary_and_ Notes_July_2022.pdf Damgacioglu, H., Hornilla, M., Bafail, O., and Celik, N. 2020. Recovering value from single stream material recovery facilities—An outbound contamination analysis in Florida. Waste Management 102:804-814. Domingues, J., Marques, T., Mateus, A., Carreira, P., and Malça, C. 2017. An additive manu- facturing solution to produce big green parts from tires and recycled plastics. Procedia Manufacturing 12:242-248. Edwards, S., and Grushack, S. 2021. The 50 states of recycling; a state-by-state assessment of containers and packaging recycling rates. Eunomia. https://www.eunomia.co.uk/reports- tools/the-50-states-of-recycling-a-state-by-state-assessment-of-containers-and-packaging- recycling-rates Gasde, J., Woidasky, J., Moesslein, J., and Lang-Koetz, C. 2021. Plastics recycling with tracer- based-sorting: Challenges of a potential radical technology. Sustainability 13:258. Gere, D., and Czigany, T. 2020. Future trends of plastic bottle recycling: Compatibilization of PET and PLA. Polymer Testing 81:106160. Geyer, R., Kuczenski, B., Zink, T., and Henderson, A. 2016. Common misconceptions about recycling. Journal of Industrial Ecology 20:1010-1017. Geyer, R., Jambeck, J., and Law, K. L. 2017. Production, use, and fate of all plastics ever made. Science Advances 3(7). https://www.science.org/doi/full/10.1126/sciadv.1700782 Heller, M. C., Mazor, M. H., and Keoleian, G. A. 2020. Plastics in the US: Toward a material flow characterization of production, markets and end of life. Environmental Research Letters 15:094034. Hogue, C. 2022. Chemical recycling of plastic gets a boost in 18 U.S. states—but environmentalists question whether it really is recycling. Chemical & Engineering News. https://cen.acs.org/ environment/recycling/plastic-recycling-chemical-advanced-fuel-pyrolysis-state-laws/100/ i17 Independent Budget Office, NYC (IBO). 2017. Per Ton Cost Differential Between Recycling and Refuse Increases. https://a860-gpp.nyc.gov/concern/nyc_government_publications/ vq27zn97b?locale=en Issifu, I., Deffor, E. W., and Sumaila, U. R. 2021. How COVID-19 could change the economics of the plastic recycling sector. Recycling 6:64. https://www.mdpi.com/2313-4321/6/4/64 Joltreau, E. 2022. Extended producer responsibility, packaging waste reduction and eco- design. Environmental and Resource Economics 83(3):527-578. https://doi.org/10.1007/ s10640-022-00696-9 Keramitsoglou, K. M., and Tsagarakis, K. P. 2018. Public participation in designing the recy- cling bins to encourage recycling. Sustainability 10:1240. Koerth, M. 2019. The era of easy recycling may be coming to an end. FiveThirtyEight. https:// fivethirtyeight.com/features/the-era-of-easy-recycling-may-be-coming-to-an-end Law, K. L., Starr, N., Siegler, T. R., Jambeck, J. R., Mallos, N. J., and Leonard, G. H. 2020. The United States’ contribution of plastic waste to land and ocean. Sciences Advances 6(44). https://www.science.org/doi/10.1126/sciadv.abd0288

250 RECYCLED PLASTICS IN INFRASTRUCTURE Leardini, P. 2020. INSIGHT: California mandate seeks to develop market for recycled plas- tic. ICIS Explore. https://www.icis.com/explore/resources/news/2020/10/21/10565676/ insight-california-mandate-seeks-to-develop-market-for-recycled-plastic Ma, J., Nawarathna, H. M. C., and Hesp, S. A. M. 2022. On the sustainable use of recycled plastics in flexible asphalt pavements. Journal of Cleaner Production 359:132081. McKinsey. 2019. Accelerating plastic recovery in the United States. https://www.mckinsey. com/industries/chemicals/our-insights/accelerating-plastic-recovery-in-the-united-states Meyer, D. E., Li, M., and Ingwersen, W. W. 2020. Analyzing economy-scale solid waste gen- eration using the United States environmentally-extended input-output model. Resources, Conservation and Recycling 157:104795. Mikula, K., Skrzpczak, D., Izydorczyk, G., Warchol, J., Moustakas, K., Chojnacka, K., and Witek-Krowiak, A. 2021. 3D printing filament as a second life of waste plastics—a re- view. Environmental Science and Pollution Research 28:12321-12333. Morawski, C. 2020. Who Pays What? An Analysis of Beverage Container Recovery and Costs in Canada. CM Consulting. https://www.cmconsultinginc.com/wp-content/ uploads/2016/09/2010.pdf More Recycling. 2018. 2016 National Post-Consumer Plastic Bag & Film Recycling Report. American Chemistry Council. https://plasticsrecycling.org/images/library/2016-film- report.pdf NAPCOR (National Association for PET Container). 2018. Report on postconsumer PET container recycling activity in 2017. www.plasticsrecycling.org –––. 2020. NAPCOR’S 2020 PET Recycling Report Reveals an 800 Million Pound Increase of Recycled Pet for End Market Use Over the Past Decade. https://napcor.com/news/ napcors-2020-pet-recycling-report-reveals-an-800-million-pound-increase-of-recycled- pet-for-end-market-use-over-the-past-decade –––. 2022. NAPCOR’s 2021 PET Recycling Report Shows Largest Amount of Postconsumer PET Ever Collected in U.S. https://napcor.com/news/2021-pet-recycling-report Ponmalar, S., and Revathi, P. 2022. Waste recycled plastic granules substitute for aggregate in concrete—Review. Materials Today: Proceedings 65:1441-1448. Profita, C. 2019. Oregon bottle deposit system hits 90 percent redemption rate. NPR. https://www.npr.org/sections/thesalt/2019/02/04/688656261/oregon-bottle-deposit- system-hits-90-percent-redemption-rate Pyzyk, K. 2020. Plastic processors moving ahead on polystyrene recycling plant in Illi- nois. Waste Dive. https://www.wastedive.com/news/plastic-agilyx-polystyrene-chemical- recycling-illinois/586321 –––. 2021. Brand commitments contributing to higher recycled HDPE prices. Waste Dive. https:// www.wastedive.com/news/recycled-hdpe-natural-milk-brands-sustainability/605418 The Recycling Partnership. 2020. State of Curbside Recycling. https://recyclingpartnership.org/ wp-content/uploads/dlm_uploads/2020/02/2020-State-of-Curbside-Recycling.pdf RecyclingMarkets.net. 2021. Prices Stabilize for PET and HDPE bales. Plastics Recycling Update. https://resource-recycling.com/plastics/2021/09/15/prices-stabilize-for-pet-and-hdpe-bales Reloop. 2020. Global Deposit Book 2020. https://www.reloopplatform.org/wp-content/ uploads/2020/12/2020-Global-Deposit-Book-WEB-version-1DEC2020.pdf Ren, Z., Gui, X., Xu, X., Zhao, L., Qiu, H., and Cao, X. 2021. Microplastics in the soil- groundwater environment: Aging, migration, and co-transport of contaminants—A criti- cal review. Journal of Hazardous Materials 419:126455. Resource Recycling News (RRS). 2018. Data corner: The evolving ton over 25 years. https:// resource-recycling.com/recycling/2018/10/29/data-corner-the-evolving-ton-over-25-years –––. 2020. Data corner: Supply possibilities to meet U.S. RPET demand—Resource recy- cling. Resource Recycling News. https://resource-recycling.com/recycling/2020/10/26/ data-corner-supply-possibilities-to-meet-u-s-rpet-demand

APPENDIX E 251 –––. 2021. Data corner: The importance of decoupling recycled and virgin plastic prices. Resource Recycling News. https://resource-recycling.com/recycling/2021/04/18/data-corner-the- importance-of-decoupling-recycled-and-virgin-plastic-prices Rosengren, C., Witynski, M., Li, R., Crunden, E. A., Boteler, C., and Pyzik, K. 2019. How recycling has changed in all 50 states. Waste Dive. https://www.wastedive.com/news/ what-chinese-import-policies-mean-for-all-50-states/510751 Sanjeev, S., Little, R., Sabir, U. 2022. CircularNet: Reducing Waste with Machine Learning. TensorFlow Hub. https://blog.tensorflow.org/2022/10/circularnet-reducing-waste-with- machine.html Saure, I. O. 2018. The Use of Pay as You Throw Schemes and Central Sorting in Mu- nicipal Solid Waste Management—A Case Study of Potential Measures to Increase Sorting and Recycling Rates of Plastic Packaging in More and Romsdal, Norway. Norwegian University of Science and Technology. https://ntnuopen.ntnu.no/ntnu-xmlui/ handle/11250/2577208 Schmidt, J., Auer, M., Moesslein, J., Wendler, P., Wiethoff, S., Lang-Koetz, C., and Woidasky, J. 2021. Challenges and solutions for plastic packaging in a circular economy. Chemie Ingenieur Technik 93:1751-1762. Schneider, S. 2021a. U.S. Plastics Recycling Struggles to Meet Demand, Highlight- ing Investment Gap. SP Global Commodity Insights. https://www.spglobal.com/ commodityinsights/en/market-insights/blogs/petrochemicals/022621-us-plastic- recycling-pet-bottles-packaging-waste-investment –––. 2021b. Supply-Demand Gap a Key Challenge for Recycled Plastics Sector. S&P Global Commodity Insights. https://www.spglobal.com/commodityinsights/en/market-insights/ blogs/petrochemicals/120921-recycled-plastics-california-pcr-mandate Schyns, Z. O. G., and Shaver, M. P. 2021. Mechanical recycling of packaging plastics: A review. Macromolecular Rapid Communications 42:2000415. Smith, R. L., Takkellapati, S., and Riegerix, R. C. 2022. Recycling of plastics in the United States: Plastic material flows and polyethylene terephthalate (PET) recycling processes. ACS Sustainable Chemical Engineering 10:2084-2096. Staub, C. 2020. Brightmark Seeking 2.4 Billion Pounds of U.S. Mixed Plastics. Plastics Re- cycling Update. https://resource-recycling.com/plastics/2020/08/05/brightmark-seeking- 2-4-billion-pounds-of-us-mixed-plastics Stewardship Ontario. 2017. 2016 Stewardship Ontario Annual Report. https://stewardshipon- tario.ca/wp-content/uploads/2022/06/2016_SO_AnnualReport-3.pdf Stina. 2021. 2020 U.S. Post-Consumer Plastic Recycling Data Dashboard. https://circularity- inaction.com/2020PlasticRecyclingData Stromberg, P. 2004. Market imperfections in recycling markets: Conceptual issues and empiri- cal study of price volatility in plastics. Resources, Conservation and Recycling 41:339- 364. https://www.sciencedirect.com/science/article/abs/pii/S0921344904000424 Sustainable Packaging Coalition. 2021. EPR Guide to All Policies. https://epr.sustainable- packaging.org Tanimoto, A. 2020. The West Coast Contamination Initiative: Results from California, Or- egon, and Washington. The Recycling Partnership. https://recyclingpartnership.org/blog- the-west-coast-contamination-initiative-results-from-california-oregon-and-washington Thoden van Velzen, E. U., Brouwer, M. T., and Feil, A. 2019. Collection behaviour of light- weight packaging waste by individual households and implications for the analysis of collection schemes. Waste Management 89:284-293. Tonjes, D. J., Aphale, O., Clark, L., and Thyberg, K. L. 2018. Conversion from dual stream to single stream recycling results in nuanced effects on revenues and waste stream amounts and composition. Resources, Conservation and Recycling 138:151-159.

252 RECYCLED PLASTICS IN INFRASTRUCTURE Townsend, T. G., Ingwersen, W. W., Niblick, B., Jain, P., and Wally, J. 2019. CDDPath: A method for quantifying the loss and recovery of construction and demolition debris in the United States. Waste Management 84:302-309. U.S. Environmental Protection Agency (USEPA). 2016. Collection Costs | Improving Recy- cling’s Economic Profile | Tools for Local Government Recycling Programs. https:// archive.epa.gov/wastes/conserve/tools/localgov/web/html/collection.html –––. 2018. Plastics: Material-Specific Data. https://www.epa.gov/facts-and-figures-about- materials-waste-and-recycling/plastics-material-specific-data –––. 2020a. Advancing Sustainable Materials Management: 2018 Tables and Figures. https:// www.epa.gov/sites/default/files/2021-01/documents/2018_tables_and_figures_dec_2020_ fnl_508.pdf –––. 2020b. National Overview: Facts and Figures on Materials, Wastes and Recycling. U.S. Environmental Protection Agency. https://www.epa.gov/facts-and-figures-about-materials- waste-and-recycling/national-overview-facts-and-figures-materials U.S. Food and Drug Administration (FDA). 2022. Submissions on Post-Consumer Recycled (PCR) Plastics for Food-Contact Articles. https://www.cfsanappsexternal.fda.gov/scripts/ fdcc/?set=RecycledPlastics Waste Dive Team. 2019. Where Curbside Recycling Programs Have Stopped and Started in the US. WasteDive.com. https://www.wastedive.com/news/curbside-recycling-cancellation- tracker/569250 Wu, H., Mehrabi, H., Karagiannidis, P., and Naveed, N. 2022. Additive manufacturing of recycled plastics: Strategies towards a more sustainable future. Journal of Cleaner Pro- duction 335:130236. Zhou, D. 2022. Choosing the optimal recycled plastic for making 3D printing filament by ELECTRE decision model. Proceedings of the SPIE 12255:186-192. https://ui.adsabs. harvard.edu/abs/2022SPIE12255E..0UZ/abstract