Plastics value chain - Abatement of greenhouse gas emissions

1.   Introduction

Chemicals are the cornerstones of modern life, and plastics are chemicals produced in bulk quantities. The value chain of plastics can be separated into two different parts; virgin plastics production from cradle-to-gate and recycling of plastic waste. The recycling side comprises of the consumption of plastics, waste generation, disposal of waste and recycling of plastic waste. The share of the OECD countries production capacity is estimated to decrease from 40% to 20% when low energy prices place about 60% of new basic chemical production capacity in non-OECD regions by 2030. This development would increase the global carbon dioxide (CO₂) emissions of the chemical sector by 50% ^[1]. The vast majority of feedstock used to make plastics, such as ethylene and propylene, originate from fossil hydrocarbons, and the plastics waste produced is seldom biodegradable accumulating, rather than decomposing, in landfills or in the natural environment. On the recycling side increasing urbanization, especially in the Indian Peninsula, Southeast Asia and Africa, causes waste management problems, resulting in the accumulation of plastics not only on land but also in the oceans. The plastic waste in the oceans increases at the rate of about eight million metric tons every year ^[2]. The environmental losses of global plastic value chain were estimated by Ryberg et al. ^[3] to 9.2 Mt in 2015. This illustrates the magnitude of the plastics recycling problem that awaits a solution. The World Bank ^[4] and the World Wildlife Fund ^[5] have both recently published reports on plastic waste management including future projections up to 2030 and 2050 with growth rates up to 40% in the production of plastics until 2030. Collection of plastic waste prior to landfilling and preventing the leakage of already landfilled plastic waste would considerably reduce the volume of plastics entering the oceans. The environmental impact of plastics was mapped by Ciel ^[6]. Several countries already restrict the use of disposable single-use plastic utensils. In addition, concerns regarding human consumption of microplastics are growing ^[7]. The increasing accumulation of plastics in ecosystems needs innovative solutions since a total ban on plastics is hardly an option to solve the problem. The Global Commitment on plastics has already over 450 signatories, mainly large companies, that are determined to start building a circular economy for plastic with a recent progress report ^[8]. A recent study mapped material flows of seven major plastics in the United States in 2015 stressing the low end-of-life recycling rate ^[9].

As for greenhouse gas (GHG) emissions, the chemical sector, producer of plastics, is the third largest industrial emitter of carbon dioxide after steel and cement production with 5.5% of global emissions ^[10]. In addition, the chemical sector is the largest industrial energy consumer and in all responsible for approximately 7% of global GHG emissions ^[11]. Much research is being directed to reduce the energy, and to improve the efficiency of chemicals production. Energy intensive pyrolysis is proposed as one option to utilize waste plastics as an intermediate for transport fuels ^[12,13]. A recent study compared the environmental impacts of chemical recycling via pyrolysis of mixed plastic waste in comparison with mechanical recycling and energy recovery. The climate impact of pyrolysis and mechanical recycling were comparable and the avoided GHG emissions of chemical recycling were significant ^[14]. The research on chemical recycling of waste plastic is increasing and a variety of technologies are identified ^[15]. Considerable efforts are directed to recycling, reuse and substituting plastics ^[16], recycling of hard plastic waste ^[17] and managing plastics in municipal waste ^[18,19], and to developing plastics from renewable resources including low carbon bioethylene ^[20], partly biobased and recycled polyethylene terephthalate (PET) ^[21], fully bioderived PET ^[22] and polyesters derived from CO₂ ^[23]. Recent research highlights life cycle impacts of recycled polyethylene (PE) both bio and fossil based by Tonini et al. ^[24] and of PET, virgin, recycled and bioderived bottles ^[25] and fossil and biobased PET ^[26]. The results of Blank et al. ^[27] on the future of CO₂ and plastic waste are optimistic on manufacture of multiple products from simple bulk chemicals to pharmaceuticals using biotechnology. Walker and Rothman ^[28] critically reviewed and compared the life cycle assessments (LCA) of fossil and biobased plastics to the European Union Product Environmental Footprint (EU PEF) standards. The results recommend wider use of the PEF standard to avoid the different system boundaries of LCA. A recent study estimated the supply chain energy and GHG emissions for annual plastics consumption in the United States for 18 major polymer types providing a benchmark against which to compare new technologies, renewable plastics and recycling of plastic waste ^[29].

Gaps in global production and materials flow data of plastics exist. The identified research gaps are manifold; the relative advantages and disadvantages of dematerialization, substitution, reuse, material recycling, waste-to-energy and conversion technologies need careful assessment. We need less littering, more recycling, and new routes for recovering and utilizing plastic waste as a raw material. Increasing efficiency in the value chain together with recycling and production of renewable plastics offer win-win opportunities to close the carbon loop and to reduce greenhouse gas emissions. The goal of this study is to assess the GHG saving potential in the global value chain of plastic. This study focuses on the midpoint climate impact of efficiency improvements in the virgin plastic production and improved recycling in the plastics value chain compared to the current GHG emissions identified in this study. The study is restricted to the polymers and their intermediates produced in annual quantities over 20 million tons a year (Mt/y) on global scale. The detailed assessment of different mechanical and chemical recycling technologies is excluded from the scope.

2.   Materials and methods

The aim of this study is to identify the possibilities of combining the abatement of GHG emissions with resource efficiency, recycling or reproducing key bulk plastics from waste streams on global scale. The key bulk plastics referred here are polymers or resins produced in quantities over 20 Mt/y. The analysis includes also the high-volume raw materials for plastics production. In this study the value chain of plastics is separated into two different parts; virgin plastics production from cradle-to-gate and recycling of plastic waste. The recycling side comprises of the consumption of plastics, waste generation, disposal of waste and recycling of plastic waste. The methodology includes collection of data on the production and consumption of plastics, followed by the evaluation of options for scenario development. Finally, a robust life cycle analysis to evaluate the most promising scenarios completes the study.

2.1.   Collection of data on production and consumption of plastics

Attention was especially placed on polymers from fossil resources that accumulate in landfills or litter the environment and leak to seas and lakes. The aim of the survey was to identify the key chemicals/polymers in the value chain of plastics originating from a fossil resource base. The survey includes data collection, analysis and assumptions. Chemicals/polymers produced in quantities over 20 Mt/y were considered. Bio-based or biodegradable plastics currently having a global production capacity of only 2.1 Mt ^[30] are excluded from this study. However, as new options for future, the renewable production routes can be included.

2.1.1.   Data collection

The production statistics of plastics including key feedstock chemicals were compiled from existing public sources. Attention was especially paid on plastics from fossil resources that accumulate in landfills or litter the environment and leak to seas and lakes. The distribution of global plastics production is from ^[30]. The fossil raw material flows, intermediate chemicals and proxies for chemicals consumed in quantities less than 20 Mt/y on the upstream side were identified. Key data were retrieved from several sources: ammonia ^[31,32,33], methanol ^[34,35], urea ^[32], ethylene ^[36] and propylene ^[37]. The actual plastic production, waste generation, recycling and incineration data rely mainly on Geyer et al. ^[38] and WWF ^[5]. Comprehensive data on global geographical distribution of the plastics production was not readily available from public sources and not included in this study.

2.1.2.   Data analysis

The current annual production volume, waste generation, recycling and incineration rate reveals the chemicals beneficial for recycling from waste streams or for regeneration/reuse from waste or renewable resources. For this purpose, the global production in the value chain of plastics in this study is defined as follows:

where i is the chemical/polymer, Prod is the amount produced in Mt/y, Waste is total waste generation (including recycled and incinerated amounts) in Mt/y, Recycle is the amount recycled waste in Mt/y, Incinerate is the amount incinerated waste in Mt/y, and Difference is the amount of plastics remaining in long-term use. The maximal additional recycling potential of plastics was estimated from Eq 1 by deducting the current amounts of recycling and incineration from the generated waste amounts.

2.1.3.   Data assumptions

Key raw materials, chemicals, precursors and plastics produced in the virgin plastics value chain were selected using the approach of Levi and Cullen ^[39] for mapping of chemical flows i.e. based on mass balances, key chemical reactions and existing technological solutions. An average standard deviation between −5 and +5% was assumed for the production and consumption amounts found from public sources. It was assumed that end-of life textiles with synthetic fibres are incinerated or discarded as municipal solid waste. Currently no global statistics is available on the recycling or reuse of synthetic fibres, although this is a growing business. Comprehensive data on geographical distribution of the plastics production is not readily available from public sources. Assumptions on production data of chemicals are given in section 3.1. Assumptions used in the preliminary robust life cycle assessment are given in section 3.3.

2.2.   Evaluation of options for scenario analysis

The production routes of selected chemicals are analyzed with emphasis on the GHG emissions. Different production routes for each chemical are assessed using the volumes of production and waste generation from Eq.1 as a base. The criteria for efficiency improvements in the virgin plastics value chain is reduced GHG emissions. Renewable options for plastics production are not excluded. On the recycling side the target is to identify recycling possibilities reducing the GHG emissions. The evaluation results in scenarios for the robust life cycle analysis.

2.3.   Preliminary life cycle assessment

Life cycle assessment (LCA) is a common standardized tool ^[40,41]. LCA consists of four phases: (1) the goal and scope definition, (2) the life cycle inventory (LCI) analysis, (3) the life cycle impact assessment (LCIA) phase and (4) life cycle interpretation.

2.3.1.   The goal and scope definition

The goal of this LCA is to assess the GHG saving potential in the global value chain of plastic, and to determine the midpoint climate impact of selected three scenarios illustrated in Figure 1. The scope covers the current virgin plastics product system (Scenario A), improved efficiency in the virgin plastics product system (Scenario B) and increased recycling of waste plastics (Scenario C) in global scale. The system boundary and approach is on a cradle-to-gate basis for virgin plastics production scenarios. The system boundary condition for recycling Scenario C excludes the use phase of plastics and is on waste-to-cradle or new product gate basis. The functional unit (FU) used is 1t of chemical/polymer produced or recycled plastic waste in the LCI for emission coefficient estimation, and Mt of chemical/polymer or recycled plastic waste for global LCIA, the system boundaries are shown in Figure 1. Each chemical/polymer/plastic included in the value chain of study is listed with a LCI result reflecting the FU and the global summary GHG emissions are calculated by multiplying the production or recycled amounts with the corresponding emissions coefficients. The LCIA results are explained in section 3.4 LCIA - Results of scenario analysis. The change from single chemical/polymer LCI to a summary global LCIA is explained also in the methods section. GHG emission allocation is used for the production emissions of ethane and propane in petroleum refineries. The impact category is restricted to midpoint climate impact. The methodology for impact assessment was global warming potential (GWP) calculated for the identified GHG emissions of carbon dioxide (CO₂), methane (CH₄) and nitrous oxide (N₂O) utilizing emission coefficients from IPCC AR4 (2007) in own calculations i.e. 1 for CO₂, 25 for CH₄ and 298 for N₂O. If the reference data contained other GHG's they were not used for GWP calculation. CO₂ emissions alone were used to define the GHG emissions when data on CH₄ and N₂O emissions were not available. The GHGs are expressed in carbon dioxide equivalents (CO₂eq). The results obtained are calculated to correspond the FU's defined for emission coefficients and global production or recycled amounts. A summary for the scope of scenarios is illustrated in Table 1. The robust LCA includes several assumptions relating to the production data of chemicals, e.g. selection of precursors for chemicals not included in this study (section 3.1). The assumptions for LCI data are specified in section 3.3. The GHG emissions connected to land use and land use change (LULUC) are not considered in this study. Detailed LCAs for each chemical/polymer included in this study, and on recycling of all plastics waste generated on a global scale are beyond the scope of this study. The reason behind this is the fact that the plastic production, waste streams, collection data and landfill specifics would require a geographic information system (GIS) meta-analysis from every spot on the globe. This kind of data were not available. The identification of the chemicals/polymers included in this study, and their use in the value chain of plastics is a part of our results (section 3.1).

2.3.2.   The LCI analysis and LCIA

The LCI input/output data are collected from public sources or estimated from the available process and technological data. Only GHG emissions are considered. The LCI results are calculated using the specific emission coefficients, e.g. t CO₂eq/t PUR. Analysis of the robustness and uncertainty of LCI data and assumptions is included. After data collection, the data were validated and related to the reference flow of the functional unit of each chemical/polymer included in the study. The results of LCIA are reported as total global GHG emissions for Scenario A and as total GHG emission reductions identified for Scenarios B and C. Detailed chemical reactions and unit processes in the production chain are not reported.

2.3.3.   Life cycle interpretation

The interpretation of LCA results includes evaluation of consistency with the defined goal and scope, identification of the significant issues, evaluation of completeness, sensitivity, limitations of the obtained results and conclusive recommendations. The environmental hot spots are excluded due to the preliminary nature of the LCA and missing geographical data on production locations.

3.   Results and discussion

The results start with the identification of the key bulk chemicals in the production chain of plastics on a global scale. Thereafter follows the analysis of the production routes of selected chemicals aimed to identify the precursors of chemicals not included in the study. The results cover the evaluation of options for scenario development, taking into account the specific consumption patterns of each chemical. Finally, a robust LCA of the selected scenarios completes the results.

3.1.   Identification of chemicals and polymers in the value chain of plastics

The analysis of available production data revealed 12 bulk chemicals/polymers as the most prevalent for further screening with potential for recycling and reuse/regeneration from waste or renewable resources in the value chain of plastics. In order of magnitude, these chemicals/polymers are as follows: ethylene (C₂H₄), propylene (C₃H₆), high-density polyethylene (HDPE) and low-density and linear low-density polyethylene (LD&LLDPE), methanol (CH₃OH), polypropylene (PP), polyester, polyamide and acrylic (PP&A) fibres, polyvinylchloride (PVC), ammonia (NH₃), polyethylene terephthalate (PET), polyurethane (PUR) resin and polystyrene (PS). The global production of plastic resins and fibres including their additives reached over 400 Mt in 2015. On average, non-fibrous plastics contain 93% polymer resin and 7% additives by mass. In addition to the plastic resins and fibres the production of other polymers and additives accounted for 41 Mt. NH₃, C₂H₄, CH₃OH and C₃H₆ are intermediates in the production of both plastic resins and fibres and are commodity chemicals widely utilized for various applications. The global annual production of the key feedstock chemicals and their share in the value chain of plastics is shown in Table 2. The estimates for NH₃ and CH₃OH use in the production chain of plastics are also precursors for the use of all other fossil feedstock chemicals not listed here and, therefore, do not represent their actual consumption figures in the production of plastics. The use of NH₃, C₂H₄, CH₃OH and C₃H₆ as precursors for other chemicals is explained in Table 2. The annual global production of plastics is shown in Table 3. The geographical distribution of C₂H₄ and C₃H₆ production is illustrated in Figure 2.

The estimates for the waste generation, recycling and incineration of chemicals/polymers are based on collected data and several assumptions. The assumption was that all NH₃ is utilized without recycling or incineration. Similarly, it was assumed no direct waste generation, recycling or incineration for C₂H₄, CH₃OH and C₃H₆ in the value chain of plastics. Plastic resin and fibre waste generation accounted for 74.2% of total polymer production in 2015. The recycling and incineration of plastics vary considerably depending on the geographical location. Detailed global data on recycling and incineration for different non-fibrous plastics is scattered and scarce. An 18% global recycling and 24% global incineration rates was assumed. PVC is practically not recycled (only 0.5 Mt in Europe), and no data is available on its incineration. PE, PP and PET are most common in packaging materials and more easily recycled. PS and PUR are seldom recycled and thus the assumption was zero recycling and 24% incineration. PP&A fibre waste practically all landfilled or incinerated, and 11% incineration on the global scale was assumed ^[4]. Production of recycled fibres exists in small quantities, e.g. recovered from plastic waste in the oceans. The volumes of waste generation, recycling and incineration of chemicals included in this study are illustrated in Table 4.

Total primary non-fibrous plastic waste generation accounts for 232 Mt in Table 4, nearly 53% of which consists of packaging waste. The overall conversion efficiency of feedstock, including precursors, to plastics listed in Table 1. is 67%, and 74% if all plastics and their additives from fossil resources are included.

3.2.   Evaluation of options for scenario analysis

The evaluation started with the analysis of the production routes of selected chemicals. Different options, both in the virgin plastics production and recycling of plastic waste, were evaluated, taking into account the specific consumption patterns of each chemical. In particular, the possibilities for increased direct recycling and resource-efficient technologies were highlighted.

3.2.1.   Virgin plastics production

Traditionally, NH₃ is produced from natural gas (CH₄) via Haber-Bosch synthesis and the process is energy-intensive. China produces CH₄ by coal gasification. Similarly, CH₃OH is produced from CH₄ or coal. C₂H₄ is produced in petroleum refinery cracking units using crude oil and natural gas as primary resources. C₃H₆ is produced by the propane dehydrogenation units of petroleum refineries and other on-purpose units. These intermediate chemicals in the value chain of plastics do not generate physical long-life waste (in normal use and excluding accidents) in the environment, only emissions into the atmosphere. Therefore, direct recycling, reuse and waste-to-energy are not realistic options for these chemicals. Changing to renewable resources and utilizing resource-efficient technologies are the best options for achieving better material efficiency and reduction in the use of fossil resources in the value chain of plastics. Development of alternative, substitute materials for plastics is also an option worth considering as a long-term solution. Similarly, chemical recycling of plastic waste, i.e. as a feed stream for petroleum refineries or other chemical treatment facilities, is a realistic option available before 2030.

3.2.2.   Plastic waste recycling options

Plastic resins and fibres, i.e. HDPE and LD&LLDPE, PP, PP&A fibres, PVC, PET, PUR and PS, all generate long-life waste in the environment. Therefore, direct recycling, reuse, substitution with renewable materials, material's use reduction and waste-to-energy solutions are the primary options for material efficiency and reduction in the use of fossil resources. The current waste generation of these plastics equals 274 Mt/y and recycled and incinerated waste amounts to 102 Mt/y (Table 4). Consequently, maximal additional recyclable waste amounts to 172 Mt/y. The near-term potential for improving recycling, starting from the consumption pattern of each non-fibrous plastic was evaluated. The polymer-specific end-use data per market sector was combined with the sector-specific waste generation data and the current polymer-specific incineration and recycling amounts were deducted from the obtained values and detailed in the Figure S1. The additional potential for near-term recycling of non-fibrous plastics amounts to 86 Mt/y, and that of fibres to 8 Mt/y. The maximal and near-term potential additional recyclable waste volumes are illustrated in Figure 3.

The estimates indicate that improving near-term recycling (the values shown in brackets are additional volumes eligible for recycling) is easiest for packaging materials, i.e. LD&LLDPE (22 Mt/y), HDPE (14 Mt/y), PP (17 Mt/y) and PET (12 Mt/y). Additional recycling or incineration of PS (8 Mt/y) is also possible. PVC (7 Mt/y) and PUR (6 Mt/y) would need partial or full chemical recycling. The total global recycling rates identified in this study are lower than e.g. those estimated for the EU by Tallentire and Steubing ^[42]. Better regulation of the global plastic waste export trade would reduce the amount of poorly recyclable, low-quality waste. Increased recycling and reuse of plastics would allow more time to develop replacement technologies for plastics, with alternative, preferably renewable materials that have positive impacts on the environment.

3.3.   Preliminary life cycle assessment

A robust life cycle estimation of the GHG emissions of the potentially best options for reducing virgin material consumption and GHG emissions in the global supply chain of each chemical/plastic produced or plastic waste recycled complete this study. Three different scenarios, base case and two based on technological choices from Figure 1 are included and explained in Table 1 section 2.3.

3.3.1.   Key inventory data and assumptions for Scenarios A and B

The global GHG and CO2 emissions are taken from UNEP ^[43]. The corresponding chemical sector emissions were calculated using the percentages of global emissions given by IEA ^[11]. It was assumed that % not amounts were unchanged between 2013 and 2018 for the chemical industry on the global scale. Key LCI data for scenarios A and B is shown in Table 5. For the selected impact category of midpoint climate change the parameter is GHG emissions expressed in t CO₂eq/t of chemical produced.

3.3.2.   LCI data for improved recycling of plastic waste

The embedded emissions of PET amount to 2.04 t CO₂eq /t, and those of PE and PP are 3.14 t CO₂eq /t. The production emissions of PE and PP amount to 2.38 t CO₂eq /t. Accordingly, PET production emissions are 3.26 kg CO₂eq /kg. The GHG emissions from incineration of PET trays is 2.0 t CO₂eq/t, and 1.3 t CO₂eq/t by avoiding the use of fossil energy. The mechanical recycling of PET trays has GHG emissions of 0.15 t CO₂eq/t and avoided emissions of 2.35 t CO₂eq/t by avoiding the use of virgin material. Similarly, mechanical recycling of mixed plastics emits 0.8 t CO₂eq/t and would have avoided emissions of 0.5 t CO₂eq/t by avoiding the use of virgin material. In general, the recycling losses from incineration would amount to 1.7 t CO₂eq/t of mixed plastics by avoiding the use of fossil energy (incineration and recycling data from Bergsma ^[44]). The recycling of PVC is challenging; it is assumed that the annual quantity of products that can be considered as substitutes for virgin PVC consists of 90% pre-consumer waste and 10% post-consumer waste. The recycling rate of PVC waste in the EU reached 16% in 2014, and the average GHG emissions of recycled PVC total 1.9 t CO₂eq/t, a saving of 0.9 t CO₂eq/t compared to virgin PVC.

3.4.   LCIA - Results of scenario analysis

This section describes the results of the robust LCA for the selected three scenarios, and give an overview on GHG emissions of the current production chain of plastics, estimates the impact of process efficiency improvements and highlights near term GHG benefits of increased recycling of plastic waste.

3.4.1.   Scenario A: Base case

The GHG emissions of the current value chain of plastics on a cradle-to-gate basis were evaluated in this Scenario A Base Case. The key LCI results for Scenario A are in Table 6. The global chemical sector GHG emissions amounted to 3444 Mt CO₂eq/y and the summed GHG emissions of NH₃, CH₃OH, C₂H₄ and propylene, used as key feedstock chemicals in the value chain of plastics total 616 Mt CO₂eq/y, corresponding to 18% of chemical sector GHG emissions. The GHG emissions of plastics production amounted to 1146 Mt CO₂eq/y and contributed 33.3% to the chemical sector GHG emissions. Thus, the whole value chain of plastics is responsible for 51% percent of the current chemical sector emissions. The average GHG emission resulted 4.81 t CO₂eq/t of plastics.

3.4.2.   Scenario B: Improved virgin plastic production efficiency

The European Union (EU) has benchmarked chemical sector performance on GHG emissions (Table 5). The benchmark values relate to 10% of the best industry performers in the EU. These values indicate improvement potential in reducing GHG emissions in the production of chemicals when best available techniques (BAT) are implemented. The first step to improve the GHG balance of NH₃, C₂H₄ and C₃H₆ would be to strive for the EU benchmarked values of GHG emissions on the global scale. For NH₃, this would mean a reduction of 0.382 t CO₂eq/t of ammonia produced. Correspondingly, for C₂H₄ and C₃H₆ the reduction would be 0.716 t CO₂eq/t of C₂H₄ and 0.756 t CO₂eq/t of C₃H₆. These and all other GHG impacts of improved resource efficiency on the global scale are illustrated in Figure 4. CH₃OH production has no EU benchmark. Currently the lowest GHG emissions of CH₃OH production are in Sweden and amount to 0.462 t CO₂eq/t of CH₃OH (Table 5). Consequently, switching the current CH₃OH production to the GHG emission level of CH₃OH in Sweden would reduce GHG emissions by 2.148 t CO₂eq/t of CH₃OH.

The emission reductions for PE, PET and PP were calculated applying the same C₂H₄ and C₃H₆ emission reduction targets for the feedstock-to-product (plastic tray and PE film). Applying the EU benchmark values, it was possible to reduce the GHG emissions of polymerization with 0.25 t CO₂eq/t of PVC. Similarly, the emissions for PS and PUR were reduced by 0.36 t CO₂eq/t. The global emissions reduction potential of PP&A fibres remained unsolved. The identified near-term possibilities for reducing the GHG impact of the chemicals selected for this study amount to 531 Mt CO₂eq/y. This would mean a 15.4% reduction of all global chemical sector GHG emissions. With these reductions the average GHG emission would drop to 3.36 t CO₂eq/t of plastics. However, the probability for all global production facilities to reach the EU benchmarked values is unclear. To conclude, the option for improved efficiency in the virgin plastics production is readily available for implementation if the financial gains and political will lower the threshold to start GHG reductions in the chemical sector.

3.4.3.   Scenario C: Improved recycling

The estimation of GHG emission reductions from improved recycling (Figure 3) would require detailed analysis of the recycling value chain of each polymer, which is beyond the scope of our study. Therefore, only a robust estimate of GHG emission reductions for near-term additional recycling of plastics was performed. Generally, long-life plastics are considered as fossil carbon sinks (embedded CO₂ emissions only), provided that production emissions (cradle-to-gate) are counted as emissions. The incineration of waste plastics also releases the embedded CO₂ emissions. The generation of energy reduces the GHG impact of incinerated plastics. Hence, mechanical and chemical recycling of plastics should be favored. This section includes both LCI data and LCA results for improved recycling of plastic waste.

A robust estimate for the avoided GHG emissions of mixed mechanical recycling of HDPE, LD&LLDPE, PP and PET in the near term (Figure 3) would mean 65 Mt/y of additional recycled plastics with annual avoided GHG emissions of 37.5 Mt CO₂eq. Improved segregation of plastics recycling would considerably increase the amount of avoided GHG emissions: for 12 Mt/y recycled PET the GHG saving alone would be 28 Mt CO₂eq. Assuming the same relation between virgin product emissions and segregated recycled production emissions as for virgin and segregated recycled PET, avoided emissions of 1.716 t CO₂eq/t is possible for recycled HD-PE, LD&LLDPE and PP by avoiding the use of virgin material. This would give avoided GHG emissions of 108 Mt CO₂eq annually on the global scale with an additional recycled volume of 65 Mt/y. The assumption was that it is possible to reach the same PVC recycling rate on the global scale as in the EU. The avoided GHG emissions of additionally recycled PVC would be 6.3 Mt/y. The estimates for the GHG savings from additional recycling of PUR, PS or PP&A fibres remained unsolved. The identified downstream near-term recycling possibilities for segregated plastics would reduce the GHG impact by 142.3 Mt CO₂eq/y. This would mean an additional 4% reduction of all global chemical sector emissions. The downstream GHG impacts of recycling on global scale are illustrated in Figure 5.

3.4.4.   Discussion on green value chain of plastics

We restricted our study on high volume plastics produced from fossil resources. The growing production of renewable plastics and their carbon footprints give mixed results. A LCA study of Tonini et al. ^[24] found the carbon footprints of biobased production of HDPE from sugarcane, polylactic acid (PLA) from organic waste and polybutylene succinate (BPS) exceeding the recycled fossil HDPE value of 1.7–1.8 t CO₂eq/t. PLA had a carbon footprint of 19–20 t CO₂eq/t. Green C₂H₄ production from corn ethanol reduces GHG emissions only if the energy used in the process is from renewable sources. This is because its production is highly energy-intensive, including H₂ production in the NH₃ fertilizer manufacture and the dehydration of ethanol with consequent water separation. Corn photosynthesis only partly offsets these emissions by CO₂ removal from the atmosphere. Assuming natural gas as the energy resource in green C₂H₄ production, the cradle-to-gate GHG emissions amount to 1.347 t CO₂eq/t of C₂H₄ ^[53]. The use of natural gas as the energy resource contributes almost 85% of the GHG emissions. Consequently, green C₂H₄ production utilizing only renewable energy resources would reduce the GHG emissions to 0.202 t CO₂eq/t of C₂H₄. Increasing the production of green C₂H₄, and in the future also green C₃H₆, with fully renewable energy would reduce, accordingly, the current GHG emissions by an astounding 1.216 t CO₂eq/t of C₂H₄ and 1.256 t CO2eq/t of C₃H₆. However, we consider this not to be a near-term option. Green NH₃ technology, using air, water and renewable electricity, is under development and expected to be piloted by 2025 ^[54]. New CH₃OH technologies include production from CO₂ and hydrogen (H₂) ^[34], and from renewable forest resources ^[55,56]. The methanol production from captured CO₂ and renewable H₂ delivers a negative emission of −0.752 t CO₂eq/t of CH₃OH. The highest negative emission amounts to −0.914 t CO₂eq/t of CH₃OH produced from forest biomass ^[46]. Accordingly, captured CO₂ and renewable H₂ technology would reduce current GHG emissions by 3.363 t CO₂eq/t of CH₃OH, and forest biomass technology by 3.525 t CO₂eq/t of CH₃OH. The current production of biobased plastics amounts 2 Mt/y and is no match to the current 400 Mt/y fossil based production of plastic resins, fibres and additives in the value chain of plastics. Increased recycling and reuse of fossil based plastics would allow more time to develop replacement technologies with alternative, preferably renewable materials with positive impacts on the environment.

3.5.   Interpretation of results

The interpretation of LCA results includes an assessment of consistency with the defined goal and scope, identification of significant issues and evaluation of completeness of the results. The robustness and sensitivity of the obtained results are addressed, and the limitations of the method used are discussed. Conclusions (section 4) include recommendations for future work. The environmental hot spots are excluded due to the preliminary nature of the LCA and missing geographical data on production locations.

The results are consistent with the defined goal to assess the GHG saving potential in the global value chain of plastic. The implemented scope of the preliminary LCA is consistent with the set target to assess the midpoint climate impact of the current global production of plastics (Scenario A) and to identify the GHG emission reduction potential by improving the efficiency of the virgin plastics production (Scenario B) and by increasing the recycling of waste plastics (Scenario C). Significant issues identified include the found possibilities to improve the global efficiency of the virgin plastics production and considerable potential to increase the amount of recycled plastic waste. The completeness of the results is adequate for the preliminary and robust level set for the LCA.

3.5.1.   Robustness, sensitivity and limitations of the results

We have based our study on references and several assumptions about them. Therefore, our results should be considered with a degree of caution, especially concerning the additional recycling amounts of plastic waste. The preliminary LCA of GHG emissions and their climate impact use published emission values. We have assumed that the LCI values based on references have a maximum confidence interval of between +5 and −5%. Data quality has geographical limitations on production technology used, precision, completeness and representativeness of sources that increase the uncertainty of results. The global data of different plastics produced or consumed differ in various sources e.g. the data set of Rydberg et al. ^[3] composed from different sources from 2015–2017 differ from the data set of this study. We assume no serious overlapping of GHG emissions in scenarios B and C. Our average emissions for the current plastics production chain are 4.81 t CO₂eq/t compared to a recent average estimate is 5.05 t CO₂eq/t of plastics ^[5] and even 6 t CO₂eq/t of LDPE or PET are reported ^[57]. Considerably lower cradle-to-gate emissions in the value chain of plastics, an average of 2.81 t CO₂eq/t for plastics produced in the United States of America (US), are reported using not LCA but the US-focused supply chain modeling tool, Materials Flows through Industry ^[29]. Efficiency improvement in the virgin plastic production is highly sensitive to the GHG emissions of C₂H₄ and C₃H₆. Mechanical recycling of plastic waste is highly sensitive to the segregation of HDPE, LD&LLDPE, PP and PET. The impact of additional, fully segregated recycling of the plastic waste PE, PP and PET 136 Mt CO₂eq/y drops to 37.5 Mt CO₂eq/y of mixed plastics. Figure 6 illustrates the sensitivity of avoided emissions to the efficiency improvement of C₂H₄ and C₃H₆ production. Figure 7 illustrates the sensitivity of avoided emissions to the segregated recycling of plastics.

The total impact of resource efficiency is the upper limit of resource efficiency measures without bioplastics or renewable ammonia, methanol, ethylene and propylene, and can be reached incrementally. We estimated the global plastic value chain emission to 1762 t CO₂eq/y in 2016 and corresponding emission reduction to 531 Mt CO₂eq/y. This compares well with the 1781 t CO₂eq/y in 2015 and an emission reduction of 535 Mt CO₂eq/y reported by Zheng and Suh ^[58] although the approaches differ. We estimated maximum avoided emissions of 1.716 t CO₂eq/t for recycled HDPE, LD&LLDPE and PP. The results of Cascone et al. ^[59] of avoided emissions of 1.495 t CO₂eq/t for recycled LDPE pellets in Sicily is 13% lower than our maximum estimate. Jeswani et al. ^[14] estimated that chemical recycling of plastic waste generates 2.3 t CO₂ eq/t less than the virgin plastic. However, the value chain started from nafta and virgin plastics production emission was estimated to 1.89 eq/t. The result is based on high quality plastic packaging waste.

4.   Conclusions

This study focuses on the abatement of global GHG emissions of intermediates and polymers in the value chain of plastics produced in annual quantities over 20 Mt. The GHG emissions of the global plastics production are 1762 Mt CO₂eq/y on a cradle-to-gate basis. This makes the sector responsible for 51% percent of the current chemical sector emissions. Focusing on improved efficiency for NH₃, CH₃OH, C₂H₄, C₃H₆, PE, PP, PET, PS and PVC in the value chain of plastics has the potential to reduce global GHG emissions by 531 Mt CO₂eq /y, provided that all of the current global production is upgraded to the level of the EU BAT. These upstream improvements in resource efficiency would mean a 15.4% reduction of all global chemical sector emissions. However, the probability for all global production facilities to reach the EU BAT is unclear. On the recycling side, increasing the recycling rate of non-fibrous plastic resins from the current 18% to 42% would reduce global GHG emissions by 142.3 Mt CO₂eq/y provided that incineration is not increased and that the segregation of recycling is improved. These downstream improvements in recycling would mean an additional 4% reduction of all global chemical sector emissions. Fastest immediate GHG emission reductions come from the improved recycling (either mechanical or chemical) of plastics. Efficiency improvements in the virgin plastics production chain come next and after that the new renewable replacements of current plastics. Better regulation of the global plastic waste export trade would reduce the amount of poorly recyclable, low-quality waste. Increased recycling and reuse of plastics would allow more time to develop replacement technologies for plastics, with alternative, preferably renewable materials with positive impacts on the environment. This study estimated the near-term possibilities to reduce the global GHG impact of the plastics value chain. The implementation would require both financial resources from the industry and political will from governments. Current caps in the global, plant level production data restricted to obtain more than preliminary and robust LCA results. Future research needs include a detailed LCA assessment of the global supply chain of each chemical and polymer, complemented by investment cost estimates of the best options.

Acknowledgements

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. R. Kajaste acknowledges with gratitude grants from Fortum and Häme Foundations.

Conflict of Interest

All authors declare no conflicts of interest in this paper.