C-Thru - Deep Dives
C-THRU
Recycled polymers as a feedstock for United States chemical...
Deep Dives
Deep Dive 3

Recycled polymers as a feedstock for chemical manufacturing supply chains in the United States: A network analysis for polyethylene pyrolysis

Authors:
Zhichao Chen
Yosuke Kimura
David T. Allen

Center for energy and environmental Resources,

University of Texas at Austin, United States
Download pdf

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

If 5 Mt of PE waste were chemically recycled each year, and the resulting ethylene and propylene were put to use by the chemical industry, other sources of these chemicals would be displaced. The knock-on effects of this would span a large number of products and processes. These wider repercussions need to be considered in any evaluation of chemical recycling as an emissions-reduction strategy.

the problem

Most plastic recycling is currently mechanical, meaning waste plastic is ground down and reshaped, with minimal change to the chemical structure of its polymers. This can lead to reduced quality and versatility in the recycled plastic across repeated recycling stages. Chemical recycling offers a route for plastics to be recycled without loss of quality by converting plastic back to its raw materials before making it into plastic once again.

For chemical recycling to be viable, the chemical products it generates need to find markets. Currently, capacity for this is limited, so we need to know:

  • How cheap does chemical recycling need to be for waste plastic to become a viable alternative raw material?
  • How would an increase in chemical recycling affect the rest of the chemical production network?
  • How do these changes impact emissions in the sector?

‍

Methods

To answer these questions, we have developed a network model of the US chemical manufacturing industry. The model tracks flows of material and energy in a complex network of substances(e.g. raw materials, intermediate goods, final products) and processes through which the chemicals are transformed. We introduce recycled polyethylene (PE) to the network as a raw material, with an assumed maximum production capacity of 5Mt per year, and explore how much of the possible 5 Mt of recycled PE can be utilised by the industry at different cost points. This allows us to find the point at which chemical recycling becomes cheap enough to be adopted at any scale by the manufacturing network.

‍

Key Results
  • Much of the network of chemical manufacturing processes is affected by introducing recycled PE into the raw materials pool.
  • Developing the PE recycling process would require capital investment of $2 billion. However, the impact of introducing recycled PE on the chemical network, and the resulting expansion of industrial capacity to use the recycled materials, would cost an additional $11 billion.
  • Identifying existing processes with excess capacity to utilise the recycled materials will be a critical consideration in techno-economic analyses of recycling plastics.
  • Our network model shows that chemical recycling of PE can be fully adopted by the industry when the cost of recycling is lower than $0.8 per kg.
  • PE recycling has significant knock-on effects for other processes and products in the chemical sector network. When the cost becomes low enough, other processes (e.g. methane-to-olefins process or ethane cracking) are displaced by PE recycling. The complexity and interconnectedness in the chemical sector mean that introducing recycled PE sets off a cascade of further changes throughout the manufacturing network.

Keep scrolling to read the full Deep Dive

Explore more Deep Dives

Deep Dive 1
Model-based energy and emission analysis of ammonia production for improved process sustainability

Ammonia production is highly unsustainable. Our model-based analysis reveals that in the US, emissions from the ammonia industry could be reduced by 10% by taking measures to enhance the efficiency of existing plants. To reduce emissions to near-zero by 2050, more fundamental changes to production pathways are necessary. Low-carbon technologies such as water electrolysis and carbon capture and storage will need to be developed and deployed to achieve this near-zero target with minimum cost.

Deep Dive 2
Global mass flow of nitrogen fertilisers and mitigation options

Nitrogen fertilisers account for 5% of global greenhouse gas (GHG) emissions. Our global mass flow model uses a whole-systems approach to explore combinations of mitigation interventions. If all the interventions we consider were applied together, GHG emissions from synthetic fertilisers could be reduced by 84% by 2050. The single most effective intervention is increasing the efficiency of nitrogen fertiliser application to plants during their use phase.

Deep Dive 3
Recycled polymers as a feedstock for chemical manufacturing supply chains in the United States: A network analysis for polyethylene pyrolysis

If 5 Mt of PE waste were chemically recycled each year, and the resulting ethylene and propylene were put to use by the chemical industry, other sources of these chemicals would be displaced. The knock-on effects of this would span a large number of products and processes. These wider repercussions need to be considered in any evaluation of chemical recycling as an emissions-reduction strategy.

Deep Dive 5
The economic context

We need to consider the potential wider economic impacts of climate change mitigation actions taken within the petrochemical sector. Therefore, we have developed an integrated macroeconomic model that combines both material and economic flows. The model is calibrated for the US context so that it represents the real-world stocks and flows of mass, energy, population, and money.

Deep Dive 6
Why is a Network Analysis of the Petrochemical Industry useful?

A network analysis of the petrochemical industry reveals the complexity and interconnectedness of the sector and indicates how changes made in one plant or location would have implications elsewhere. This knowledge helps policymakers to target strategic locations or companies for emissions mitigation.

Deep Dive 4
Material flow analysis of petrochemicals production in the UK

We use a new methodology to map the emissions and energy use of the petrochemical sector in greater detail than existing studies. An effective, targeted emissions mitigation strategy requires both a big-picture view of the sector and an understanding of the granular detail so that broad goals can be translated into specific interventions. Here, we demonstrate the use of semantic web technologies, which allow us to navigate multiple levels of detail and aggregation with the same tool. This has not been possible before now.

Deep Dive 1
Model-based energy and emission analysis of ammonia production for improved process sustainability

Ammonia production is highly unsustainable. Our model-based analysis reveals that in the US, emissions from the ammonia industry could be reduced by 10% by taking measures to enhance the efficiency of existing plants. To reduce emissions to near-zero by 2050, more fundamental changes to production pathways are necessary. Low-carbon technologies such as water electrolysis and carbon capture and storage will need to be developed and deployed to achieve this near-zero target with minimum cost.

Deep Dive 2
Global mass flow of nitrogen fertilisers and mitigation options

Nitrogen fertilisers account for 5% of global greenhouse gas (GHG) emissions. Our global mass flow model uses a whole-systems approach to explore combinations of mitigation interventions. If all the interventions we consider were applied together, GHG emissions from synthetic fertilisers could be reduced by 84% by 2050. The single most effective intervention is increasing the efficiency of nitrogen fertiliser application to plants during their use phase.

Deep Dive 5
The economic context

We need to consider the potential wider economic impacts of climate change mitigation actions taken within the petrochemical sector. Therefore, we have developed an integrated macroeconomic model that combines both material and economic flows. The model is calibrated for the US context so that it represents the real-world stocks and flows of mass, energy, population, and money.

Deep Dive 6
Why is a Network Analysis of the Petrochemical Industry useful?

A network analysis of the petrochemical industry reveals the complexity and interconnectedness of the sector and indicates how changes made in one plant or location would have implications elsewhere. This knowledge helps policymakers to target strategic locations or companies for emissions mitigation.

Read the full Deep Dive

Abstract

For chemical recycling of plastic wastes to be viable, chemical products generated in the recycling need to find markets. A network model of the U.S. chemical manufacturing industry was used to assess at what cost points, and the extent to which, chemical products from thermal pyrolysis of polyethylene might find markets in the current U.S. chemical manufacturing industry. The network modelling determined the cost points at which the simulated industry network utilized the thermal pyrolysis products, and which processes were displaced by the supply of recycled materials. The characteristic feature of the simulations is the large number of processes in the chemical manufacturing network that are impacted by the availability of a relatively small number of products from polyethylene recycling. In the case of polyethylene recycling, the capital cost requirements for expanding capacity to effectively utilize the recycled materials is greater than the capital required for the pyrolysis process. This suggests that identifying scenarios where recycled materials can be utilized in processes that have excess capacity will be a critical consideration in techno-economicanalyses of recycling plastics.

Introduction

Global production of plastics has been growing rapidly since1950, reaching more than 400 million metric tons (Mt) in 2015.1 Plastics have been extensively used in packaging, consumer products, construction, and transportation applications, however, most plastics are single use and recycling is limited. The U.S. Environmental Protection Agency (EPA) reports that 36 Mt of plastic wastes were generated in United States in 2018, and about9% of plastic wastes were recycled, with 76% going to landfills.2

Much of current plastics recycling in the United States involves mechanical reprocessing of the polymers, with relatively limited chemical recycling. For chemical recycling of plastic wastes to be viable, chemical products generated in the recycling need to find markets.  A number of analyses of plastics waste management and recycling have been reported. These analyses have examined topics such as the optimization of waste plastic reverse logistics3 and the economic viability of waste recycling infrastructures.4  Chemical manufacturing in the United States isa highly networked system, however, and few analyses have been performed, that examine the impact of plastics recycling on the entire chemical manufacturing network.  This work will use network models of the chemical manufacturing industry to assess at what cost points, and the extent to which, chemical products from depolymerization of plastic wastes might find markets in the current U.S.chemical manufacturing industry. The pyrolysis of polyethylene (PE) will be used to demonstrate the methodology.

Polyethylene production represents about 25% of the total production of plastics in the US, and therefore represents a logical target for recycling. Multiple processes have been proposed for recycling of polyethylene, including both thermal and catalytic pyrolysis. Pyrolysisis the process of thermally degrading polymer molecules with intense heat and in absence of oxygen.5 Thermal pyrolysis is usually performed at temperatures in excess of 500 °C.5 The three major products from thermal pyrolysis of polyethylene are oil, gas and char. The products of pyrolysis depend on temperature, pressure, residence time, reactor configurations, and other process parameters.6,7 Reddy et al.8 reviewed the influence of process parameters on product distribution for thermal pyrolysis of polyethylene and report that gas products generally contain H2, CH4,C2H4, C2H6, C3H6and C3H8, while liquid products generally consist of alkanes, alkenes and aromatic hydrocarbons, with the exact composition and yields depending on process conditions.9,10 Thermal pyrolysis under other conditions, or catalytic pyrolysis, can lead to different reaction mechanisms and product compositions.11–13 There are also various reactor configurations for plastic pyrolysis, including fixed beds, fluidized beds, and rotary kilns.14 This work will investigate potential use of the products of thermal pyrolysis of polyethylene in chemical manufacturing.

A network model of the US chemical manufacturing industry, developed by DeRosa et al.,15 will be adapted to assess the potential use of these products of chemical recycling. The model tracks material and energy flows in a complex network of feedstocks, processes, intermediates, and final products. Feedstocks are converted into intermediate chemicals via various processes to obtain final end products. The same intermediates and final end products can be produced by different technologies with the same feedstocks. For example, ethylene can be produced from ethane by steam cracking of either ethane or naphtha, with naphtha cracking leading to greater yields of by-products than ethane cracking. The model developed by DeRosa, et al.15 consisted of 873 chemical processes that produce 283 different chemicals to represent the US chemical manufacturing industry. The process data was obtained from the IHS 2012 ProcessEconomic Program Yearbook. In the original model formulation, natural gas, natural gas liquids (NGLs) and crude oil were the primary raw materials, and a linear programming model with supply, demand, and mass balance constraints, was used to minimize the total production cost. In this work, waste polyethylene(PE) will be assumed to be available as raw material, and thermal pyrolysis processes will be assumed to be available to convert these waste plastics into chemicals. The impact of plastics recycling process on the use of processes in the network will be examined as a function of the costs of the waste plastics and the pyrolysis process. Key issues to be addressed by the work are the cost points at which the simulated industry network utilizes the recycling technology, and which processes are displaced by the supply of recycled materials. The results from the network model provide insight into how, and to what extent, chemical recycling of plastics might be adopted by the current chemical manufacturing infrastructure in the United States.

Scroll to keep reading

Method

Polymer recycling processes

The products from thermal pyrolysis of PE can be classified as gas, oil and char.8–10 In this work, it is assumed that thermal pyrolysis of PE yields 80% gas product and 20% oil and char. Product composition from a simplified PE pyrolysis is shown in Table 1. For this work it will be assumed that ethane, ethylene, propane,and propylene from pyrolysis will be used as feedstock chemicals for manufacturing. Char is assumed to be used as a fuel to provide heat for the reactor for thermal pyrolysis. Methane, hydrogen, and oil are assumed to be used for energy recovery and result in energy credits. The details of stoichiometry of the reaction, heat requirement for pyrolysis, and energy recovered from the products are described in the SI (Tables S1-S3).

Base Case Model development

The development of the original network model of the chemical manufacturing industry is described by DeRosa et al.15. Demand of final end products, and the cost of processes are retained in this work. The current model includes851 processes and 885 materials. To characterize the markets for the pyrolysis products of polyethylene (ethane, propane, ethylene and propylene), several additional model modifications were required:

  • A new raw material of waste PE was assumed to be available at zero cost. A new thermal pyrolysis process for PE was added to the network model. The stoichiometry for the added process is described in Table S1in the SI.
  • The supply of waste PE was assumed to be 5Mt/year. If the approximately 40 million tons per year of plastic waste in theUnited States is 25% polyethylene, then 5 million tons per year represents half of US polyethylene waste. The maximum available annual supplies of raw materials to the modified network model are listed in Table 2.
  • Processes for ethylene, polypropylene and olefins productions that are no longer in industrial use were removed. Processes from methane to gasoline and diesel production, which are not in widespread use in the United States, were removed.
  • To account for the different grades of propylene(refinery grade, chemical grade and polymer grade) used in the model, inter-conversion processes for the various grades of propylene were added. PE pyrolysis was assumed to yield refinery grade propylene and the added processes enabled the conversion of refinery grade propylene into chemical and polymer grade propylene. In the base case analyses, the costs of interconversion among propylene grades were assumed to be zero.
  • Cost for three refinery processes forNG, NGLs and crude oil are represented by raw material costs based on market price of year 2012 (NG $0.09/lb16, NGLs $0.22/lb17, and crude oil $0.31/lb18). No cost interconversion among gasoline, kerosene, diesel, light naphtha, and naphtha are allowed, so that crude oil refining satisfies both industrial demand (such as naphtha) and fuel demand (such as diesel and gasoline) without strict constraint across these light hydrocarbons.  This assumption is consistent with refiners adapting the operation of their processes to yield products in ratios demanded by markets.

‍

Natural gas, NGLs and crude oil are among the primary raw materials in the model. They provide important feedstocks including methane, ethane, propane, butane, and naphtha. For natural gas, the supply in the model was assumed to be equal to industrial natural gas consumption in the UnitedStates, which was 7,226 billion cubic feet in 2012.19 Using density of 0.042lb/cubic foot20, the supply was 300 × 109 lb in2012. The distribution of components in natural gas was assumed to be 93.07%methane, 3.21% ethane, and 0.59% propane (higher hydrocarbons and non-hydrocarbons are ignored in the mass balance).21

The supply of NGLs is assumed to be 20% of that of natural gas, as they are extracted as byproducts in the production of natural gas.22 The composition of NGLs is based on 2012 production data from EIA23 and shown in Table S4 in theSI.

For crude oil, the supply in the model is assumed to be the same as the total refinery input level. This includes use of refinery products as feedstock for chemical manufacturing and as fuel and other final end products. Total crude oil refinery input for all U.S. refineries in 2012 was5,489,516,000 barrels,24 with a weighted average APIGravity of 31.0.25 This API Gravity gives a density of 7.267 lb/gal which leads to 1,675 × 109 lb crude oil supply in 2012. The crude yield is approximated for 2012 using EIA refinery yield data.26

The supply of coal is based on industrial use in 2012, which was 85.6 × 109 lb.27 Lignite comprises 7% of U.S.coal production by weight.28 The supply of lignite is assumed to be 6 × 109 lb. In addition to the constrained raw materials supplied in Table 2, the model utilizes a series of raw materials, such as minerals, solvents, initiators, and catalysts, without supply constraints. About 90% of these materials were inorganic materials such as phosphate rock and sulfur dioxide. The use of organic materials without supply constraints is about 7 Mt/yr, which is less than 5% of the use of primary raw materials in the base case.

Demands are specified by final end products. There are 141 final end products in the demand list. Totaldemands across all final end products are 748 × 109 lb. The detaileddemand data are described by DeRosa et al.15.

‍

Model statement

The objective of the model is the minimization of total costfor chemical manufacturing (Eq. 1)

where Cj is the cost of process j in cent/pound of primary product, Xj and is annual production level of process j in pounds/year. The model has the following supply and demand constraints:

where ai,j is the input-output coefficient for material i in process j, Si is the annual supply of material i, Di is the annual demand of material i. There are a total of 851 variables and 401 constraints. The problem was modelled using General Algebraic Modeling System (GAMS). BDMLP solver is used to find optimal values for production levels that meet the total demand of all final products with the constraints of supply and minimizing total production cost.

Process cost determination

Production cost of each process was determined per pound, generally per pound of a primary product (e.g., per pound of ethylene, for steam cracking of ethane to form ethylene). The cost is the sum of operating and variable costs (Eq 5). Operating cost is estimated from IHS 2012 ProcessEconomics Program Yearbook, including maintenance, operation, labour, overhead, taxes and depreciation over a 10-year period. These costs are all directly reported in the IHS 2012 Process Economics Program Yearbook. Variable cost includes raw material cost, byproduct credit (by sale) and utility cost. The price of utilities and all raw materials, byproducts and primary products are used as reported in IHS yearbook, except for the costs of petroleum, natural gas, and NGLs, which were based on commodity prices in 2012.16-18

The costs of the PE pyrolysis process were based on processing per pound of recycled polymer. The total costs were varied from 0 to160 cents per pound of recycled polymer. For the process cost of PE pyrolysis, a byproduct credit is included. The byproduct credit is a reduction in process cost due to energy recovery from hydrogen, methane, and oil products. The details of calculation of byproduct credit can be found in the SI. The byproduct credit is 2 cents per lb of waste PE.

Base case supply availability

Two different cases, with different NGLs availability were simulated. In Case 1, as described above, supply of NGLs alone does not meet the demands of light alkanes (C2-C4) in the model. In Case2, supply of natural gas and NGLs is increased so that the availability of these feedstocks could be used to satisfy all light alkane demands. The maximum supply of natural gas (483 Mt/yr) and NGLs is 3.5 times larger than that inCase 1. The supply of NGLs (97 Mt/yr) is still assumed to be 20% of natural gas. Other inputs are kept the same as in Case 1.  

‍

Scroll to keep reading

Table 1 Product composition and supply from thermal pyrolysis of recycled PE.

Table 2 Annual supply constraints of primary raw materials for the network model in base case

Results & Discussion

Waste PE utilisation

Figure 1 shows the amount of waste PE utilized via the recycling process in the network model for Cases 1 and 2. In Case 1 (orange squares in Figure 1), the network model does not utilize any waste PE when the process cost is at or above 122 cents/lb. When the cost decreases below 122 cents/lb, the recycling process starts to be adopted and 0.4 Mt/year of waste PE is used.The “process cost” of PE recycling referred to here is the cost point at which the recycling process gets adopted by the system.  If the adoption cost is greater than the costs associated with acquiring the waste PE and operating the process, then the recycling could be done profitably.  As the process adoption cost further decreases to 114 cents/lb, an additional 0.4Mt/year of waste PE is utilized by the system. The use of waste PE increases to1.3 Mt/year with several steps as the process cost further decreases to 48cents/lb. When the cost is lower than 46 cents/lb, the utilization of waste PE starts to increase sharply. The system utilizes all the available waste PE (5 Mt/year)when the cost is below 38 cents/lb.


In Case 2 (more supply of natural gas and NGLs), utilization of PE recycling starts at a higher cost (right-shifted) beginning when the cost decreases to 146 cents/lb. In Case 2, with excess supply of natural gas and NGLs, ethane cracking is the initial process displaced by PE recycling, while in Case1, methane to olefins processes are the processes initially displaced.   More detailed mappings of these changes are provided in the sections that follow. InCase 2, there is a sharp increase in the utilization of waste PE starting at 52cents/lb and Case 2 reaches full utilization (5 Mt/year) when the cost decreases to 36 cents/lb, which is similar to the cost point for full utilization in Case 1.

Ethylene production

The gross production of ethylene from different processes in Case 1 is shown in Figure 2. When waste PE is not utilized, ethylene is mainly produced by methanol-to-olefins (MTO) processes and ethane cracking. PE recycling displaces production of ethylene from MTO processes by 1.59 Mt/year, as shown in Table 3.

The changes in ethylene production lead to a complex cascade of additional changes in the chemical manufacturing network because the changesin the use of MTO processes lead to changes in production of byproducts. These complex transformations can be mapped using the framework shown in Figure 3. The detailed description of the visualization procedure and meaning of the nodes and edges can be found in the SI. In Figure 3, raw materials for chemical manufacturing (natural gas, NGLs, crude oil, lignite, and coal) are shown as nodes that have circles with red borders. Raw materials that have no supply constraints, such as minerals, solvents, initiators, and catalysts, are shown as nodes that have circles with orange borders. Products are shown as nodes that have circles with green borders. Intermediates are shown as nodes in the middle of the diagram. Raw materials, intermediates and products are connected by lines representing processes that transform raw materials and intermediates into products with the thickness of the lines representing the extent of use of the process. Figure 3 represents process utilization for all the processes in Case 1.

Figure 4 shows the effect of adding a supply of waste polyethylene (upper left of diagram) to raw material supplies. Figure 4 shows the differences between the flows with no plastics recycling and the flows with plastics recycling at zero cost (all 5 Mt of available waste PE is used). Only lines representing processes with changes in flow are shown and the thicknesses of the lines indicate the magnitude of the process use with full utilization of waste PE. Dashed lines indicate the processes that are no longer used with plastics recycling at zero cost. The colours of the lines indicate the magnitude of the change. The changes shown in Figure 4 result from the fact that the ethylene and propylene resulting from PE recycling displace MTO processes. The decrease in use of methanol leads to the decrease in use of methane from natural gas. The displaced MTO process also leads to a change in the sources of C4 products, which, in turn, leads to a change in refinery product slates. For example, production of methyl tert-butyl ether switches to a route that uses mainly isobutane and n-butane instead of isobutylene and n-butylene and the changes in product distributions then propagate throughout the manufacturing network.  

Figure 4 maps the changes between no use of waste PE and 5 Mt/year of waste PE use. The process changes that occur at intermediate levels of wastePE use (i.e., 122 to 48 cents/lb in Figure 1) are not simple linear interpolations between the process utilizations between no waste PE use and 5 Mt/year of waste PE use, however, the main features remain consistent. The reduced use of methanol leads to a decrease in use of methane from natural gas. Production of methyl tert-butyl ether switches to isobutane and n-butane from NGLs as raw materials. Additional mappings of the flows shown in Figure 4 and detailed characterizations of the changes at various cost points are provided in FiguresS1-S9 in the SI.

To examine the features of a network where PE recycling might displace ethane and propane cracking, the second base case with expansion of natural gas and NGLs availability was simulated. As shown in Table 4 andFigure 5, for this case, waste PE leads to displacement of ethane cracking, rather than MTO processes, since ethane cracking dominates ethylene production in the base case. Figure 6 shows the network changes driven by the availability of waste PE for Case 2.  

In cases where the waste PE streams largely displace ethane cracking, the use of NGLs decreases. This results in changes in the supply of propane, isobutane, and n-butane, and these changes propagate throughout the network.

A defining feature of Figures 4 and 6 is the number of processes that are impacted by the availability of a relatively small number of products from polyethylene recycling. If a goal is to minimize the extent to which the use of waste polyethylene disrupts downstream manufacturing flows, in principle, a strategy could be to match the product yields obtained from ethane cracking. If, for example, the polyethylene pyrolysis yields could be changed so that they matched the most extensively used ethane cracking technology, it might be assumed that the network effects would be minimal, however, this is not the case. If the yield of ethylene from PE recycling increases from 35% to 55%, and the yields of ethane, propylene, and propane become 8%, 2%, and 1%, respectively, the yields match those from ethane cracking. Figure 7 shows the changes in networks flows for a case where the light gases produced by thermal pyrolysis of PE are assumed to be almost exclusively ethylene. Matching the yields of the recycling process to the yields of ethane cracking changes the processes that see modified use, but the number of processes that have changes is still extensive.  The case study summarized inFigure 7 demonstrates the difficulty associated with attempting to perform drop-in displacement of processes in chemical manufacturing networks.

‍

Implications

The recycling of waste PE has the potential to be utilized by the chemical manufacturing industry if a process can be built and operated at a cost below 122 to 148cents/lb of waste PE. More widespread utilization of waste PE feedstocks occurs if costs are less than 36 to 38 cents/lb. The supply of recycled materials (mainly ethylene and propylene)displaces the supply of those from conventional olefins manufacturing processes and leads to additional complex changes in the chemical manufacturing network because the displacement leads to changes in byproduct production.

The large number of process flows impacted by the introduction of large-scale chemical recycling of plastics may impact the assessment of the net benefits of plastics recycling and it may impact the degree to which feedstocks from chemical recycling of plastics is utilized. For example, if a goal is to reduce greenhouse gas (GHG) emissions in the chemical manufacturing sector, then the changes in flows of all affected processes should be accounted for. TableS5 in the SI lists the changes in process utilization shown in Figure 4 along with the greenhouse gas emissions associated with each process. The greenhouse gas emission factors were drawn from Chen, et al.29 The net change in greenhouse gas emissions for this scenario is dominated, not by the changes in primary ethylene production, but instead by changes in refinery processes.

Another impact of the large number of changes in process utilization is the need for increased capacity for some processes and idled capacity for other processes. For the case where all 5 Mt/yr of waste PE is utilized, the total capital required to increase capacity of the processes that expand production is $11 billion. Idled capacity is valued at $14 billion.

These capital costs can be compared to the capital required for the PE pyrolysis process.  Capital investment for a plastics pyrolysis plant with processing capacity of 100,000 tons/yr is approximately $43 million.30 In another study, the capital investment required for a plastics pyrolysis process with processing capacity of 365,000 tons/yr is estimated to be $118 million.31 This suggests that, to process 5Mt/yr of waste PE by pyrolysis, the total capital required is about $1.6 billion to $2.2 billion.  The capital required to enable the optimal use of pyrolysis products ($11 billion) is estimated to be roughly five times the capital required for the pyrolysis facility. Further, the idled capital investment ($14 billion) by adopting PE recycling is more than six times than the total capital investment required for processing PE wastes by pyrolysis. Overall, these costs suggest that the use of feedstocks from PE recycling is most likely to be adopted in regions where existing capacity required for the use of the feedstocks is available.


Conclusions

Chemical manufacturing involves networks of highly interconnected processes, and changes such as large-scale chemical recycling of plastics have the potential to propagate change throughout the network. These changes have the potential to influence the overall costs and benefits of plastics recycling. This work has demonstrated a framework for evaluating these network changes, and has shown that for some impacts, such as the capital cost implications of adopting plastics recycling, the consequences of network changes can be large compared to relatively simple assessments of change.

Scroll to keep reading

Fig. 1 Utilisation of recycled PE by the network model at different recycling cost points in Case 1 (orange square) and Case 2 (blue circle).

Table 3 Comparison of production of ethylene from different processes in Case 1.

Table 4 Comparison of production of ethylene from different processes in Case 2.

Figure 2 Gross production of ethylene by different processes vs. process cost of recycling PE in Case 1.

Figure 3 Network of chemical manufacturing in base Case 1 (recycling PE not introduced).

Fig. 4 Network of chemical manufacturing: changes between no use of recycled PE and 5 Mt/year of recycled PE use.

Fig. 5 Production of ethylene by different processes vs. process cost of recycling PE in the case with large scale supply of natural gas and NGLs (Case 2).

Fig. 6 Network of chemical manufacturing: changes between no use of recycled PE and 5 Mt/year of recycled PE use; the base case for this analysis has more extensive NGLs availability than the cases modeled in Figures 3 and 4.

Fig. 7 Network of chemical manufacturing: changes between no use of recycled PE and 5 Mt/year of recycled PE use; the base case for this analysis has more extensive NGL availability than the cases modeled in Figures 3 and 4 and the PE recycling technology process yields are assumed to match typical yield for ethane cracking.

references
  1. Geyer, R.; Jambeck, J. R.; Law,K. L. Production, Use, and Fate of All Plastics Ever Made. Sci. Adv. 2017,3 (7). https://doi.org/10.1126/sciadv.1700782.
  2. Advancing Sustainable Materials Management: Facts and FiguresReport | US EPAhttps://www.epa.gov/facts-and-figures-about-materials-waste-and-recycling/advancing-sustainable-materials-management(accessed 2022 -07 -25).
  3. Mohammadi, M.; Jämsä-Jounela, S. L.; Harjunkoski, I. OptimalPlanning of Municipal Solid Waste Management Systems in an Integrated SupplyChain Network. Comput. Chem. Eng. 2019, 123, 155–169.https://doi.org/10.1016/j.compchemeng.2018.12.022.
  4. Ma, J.; Tominac, P. A.; Aguirre-Villegas, H. A.; Olafasakin, O.O.; Wright, M. M.; Benson, C. H.; Huber, G. W.; Zavala, V. M. EconomicEvaluation of Infrastructures for Thermochemical Upcycling of Post-ConsumerPlastic Waste. Green Chem. 2022, 25 (3), 1032–1044.https://doi.org/10.1039/d2gc04005k.
  5. Anuar Sharuddin, S. D.; Abnisa, F.; Wan Daud, W. M. A.; Aroua,M. K. A Review on Pyrolysis of Plastic Wastes. Energy Conversion andManagement. Pergamon May 1, 2016, pp 308–326.https://doi.org/10.1016/j.enconman.2016.02.037.
  6. Dogu, O.; Pelucchi, M.; Van de Vijver, R.; Van Steenberge, P. H.M.; D’hooge, D. R.; Cuoci, A.; Mehl, M.; Frassoldati, A.; Faravelli, T.; VanGeem, K. M. The Chemistry of Chemical Recycling of Solid Plastic Waste viaPyrolysis and Gasification: State-of-the-Art, Challenges, and FutureDirections. Progress in Energy and Combustion Science. Pergamon May 1,2021, p 100901. https://doi.org/10.1016/j.pecs.2020.100901.
  7. Abbas-Abadi, M. S.; Ureel, Y.; Eschenbacher, A.; Vermeire, F.H.; Varghese, R. J.; Oenema, J.; Stefanidis, G. D.; Van Geem, K. M. Challengesand Opportunities of Light Olefin Production via Thermal and CatalyticPyrolysis of End-of-Life Polyolefins: Towards Full Recyclability. Progressin Energy and Combustion Science. Pergamon May 1, 2023, p 101046.https://doi.org/10.1016/j.pecs.2022.101046.
  8. Suresh Kumar Reddy, K.; Kannan, P.; Al Shoaibi, A.;Srinivasakannan, C. Thermal Pyrolysis of Polyethylene in Fluidized Beds: Reviewof the Influence of Process Parameters on Product Distribution. J. EnergyResour. Technol. Trans. ASME 2012, 134 (3).https://doi.org/10.1115/1.4006790.
  9. Mastral, F. J.; Esperanza, E.; Garciía, P.; Juste, M. Pyrolysisof High-Density Polyethylene in a Fluidised Bed Reactor. Influence of theTemperature and Residence Time. J. Anal. Appl. Pyrolysis 2002, 63(1), 1–15. https://doi.org/10.1016/S0165-2370(01)00137-1.
  10. Williams, P. T.; Williams, E. A. Fluidised Bed Pyrolysis of LowDensity Polyethylene to Produce Petrochemical Feedstock. J. Anal. Appl.Pyrolysis 1999, 51 (1), 107–126.https://doi.org/10.1016/S0165-2370(99)00011-X.
  11. Elordi, G.; Olazar, M.; Lopez, G.; Artetxe, M.; Bilbao, J. ProductYields and Compositions in the Continuous Pyrolysis of High-DensityPolyethylene in a Conical Spouted Bed Reactor. Ind. Eng. Chem. Res. 2011,50 (11), 6650–6659. https://doi.org/10.1021/ie200186m.
  12. Onwudili, J. A.; Insura, N.; Williams, P. T. Composition ofProducts from the Pyrolysis of Polyethylene and Polystyrene in a Closed BatchReactor: Effects of Temperature and Residence Time. J. Anal. Appl. Pyrolysis2009, 86 (2), 293–303.https://doi.org/10.1016/j.jaap.2009.07.008.
  13. Chen, Z.; Zhang, X.; Che, L.; Peng, H.; Zhu, S.; Yang, F.; Zhang,X. Effect of Volatile Reactions on Oil Production and Composition in Thermaland Catalytic Pyrolysis of Polyethylene. Fuel 2020, 271,117308. https://doi.org/10.1016/j.fuel.2020.117308.
  14. Li, H.; Aguirre-Villegas, H. A.; Allen, R. D.; Bai, X.; Benson, C.H.; Beckham, G. T.; Bradshaw, S. L.; Brown, J. L.; Brown, R. C.; Cecon, V. S.;Curley, J. B.; Curtzwiler, G. W.; Dong, S.; Gaddameedi, S.; García, J. E.;Hermans, I.; Kim, M. S.; Ma, J.; Mark, L. O.; Mavrikakis, M.; Olafasakin, O.O.; Osswald, T. A.; Papanikolaou, K. G.; Radhakrishnan, H.; Sanchez Castillo,M. A.; Sánchez-Rivera, K. L.; Tumu, K. N.; Van Lehn, R. C.; Vorst, K. L.;Wright, M. M.; Wu, J.; Zavala, V. M.; Zhou, P.; Huber, G. W. Expanding PlasticsRecycling Technologies: Chemical Aspects, Technology Status and Challenges. GreenChemistry. Royal Society of Chemistry November 28, 2022, pp 8899–9002.https://doi.org/10.1039/d2gc02588d.
  15. Derosa, S. E.; Allen, D. T. Impact of Natural Gas and Natural GasLiquids Supplies on the United States Chemical Manufacturing Industry:Production Cost Effects and Identification of Bottleneck Intermediates. ACSSustain. Chem. Eng. 2015, 3 (3), 451–459.https://doi.org/10.1021/sc500649k.
  16. United States Natural Gas Industrial Price (Dollars per ThousandCubic Feet) https://www.eia.gov/dnav/ng/hist/n3035us3a.htm (accessed 2023 -04-21).
  17. U.S. Natural Gas Liquid Composite Price (Dollars per Million Btu)https://www.eia.gov/dnav/ng/hist/ngm_epg0_plc_nus_dmmbtuA.htm (accessed 2023-04 -21).
  18. Cushing, OK WTI Spot Price FOB (Dollars per Barrel)https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=RWTC&f=A(accessed 2023 -04 -21).
  19. Natural Gas Annual Respondent Query System. U.S. EnergyInformation Administration.https://www.eia.gov/dnav/ng/ng_cons_sum_dcu_nus_a.htm (accessed 2022 -08 -23).
  20. Skone, T. J.; Littlefield, J.; Marriott, J. Life Cycle GreenhouseGas Inventory of Natural Gas Extraction, Delivery and Electricity Production. 2011.https://doi.org/10.2172/1515238 (accessed May 21, 2023).
  21. Shires, Theresa M.; Loughran, Christopher J.; Jones, Stephanie;Hopkins, E. Compendium of Greenhouse Gas Emissions Methodologies for the Oiland Natural Gas Industry. Prep. by URS Corp. Am. Pet. Inst. (API). API,Washingt. DC 2009 (https://ww2.arb.ca.gov/sites/default/files/2020-04/API%20Compendium%202009.pdf accessed 2022 -09 -16).
  22. What are natural gas liquids and how are they used. U.S. EnergyInformation Administration.https://www.eia.gov/todayinenergy/detail.php?id=5930 (accessed 2022 -09 -15).
  23. U.S. Natural Gas Plant Field Production. U.S. Energy InformationAdministration. https://www.eia.gov/dnav/pet/pet_pnp_gp_dc_nus_mbbl_m.htm(accessed 2022 -08 -23).
  24. U.S. Refinery and Blender Net Input of Crude Oil. U.S. EnergyInformation Administration.https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=MCRRIUS1&f=A(accessed 2022 -08 -26).
  25. U.S. API Gravity (Weighted Average) of Crude Oil Input toRefineries. U.S. Energy Information Administration.https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=MCRAPUS2&f=A(accessed 2022 -08 -26).
  26. U.S. Refinery Yield. U.S. Energy Information Administration.https://www.eia.gov/dnav/pet/pet_pnp_pct_dc_nus_pct_a.htm (accessed 2022 -08-26).
  27. Coal Data - U.S. Energy Information Administration (EIA)https://www.eia.gov/coal/data.php (accessed 2022 -09 -16).
  28. Subbituminous and bituminous coal dominate U.S. coal production.U.S. Energy Information Administration.https://www.eia.gov/todayinenergy/detail.php?id=2670 (accessed 2022 -09 -15).
  29. Chen, Q.; Dunn, J. B.; Allen, D. T. Mapping Greenhouse GasEmissions of the U.S. Chemical Manufacturing Industry: The Effect of FeedstockSourcing and Upstream Emissions Allocation. ACS Sustain. Chem. Eng. 2022,10 (18), 5932–5938. https://doi.org/10.1021/acssuschemeng.2c00295.
  30. Kulas, D. G.; Zolghadr, A.; Chaudhari, U. S.; Shonnard, D. R.Economic and Environmental Analysis of Plastics Pyrolysis after SecondarySortation of Mixed Plastic Waste. J. Clean. Prod. 2023, 384,135542. https://doi.org/10.1016/j.jclepro.2022.135542.
  31. Almohamadi, H.; Alamoudi, M.; Ahmed, U.; Shamsuddin, R.; Smith, K.Producing Hydrocarbon Fuel from the Plastic Waste: Techno-Economic Analysis. KoreanJ. Chem. Eng. 2021, 38 (11), 2208–2216.https://doi.org/10.1007/s11814-021-0876-3.

Explore more Deep Dives

Deep Dive 1
Model-based energy and emission analysis of ammonia production for improved process sustainability

Ammonia production is highly unsustainable. Our model-based analysis reveals that in the US, emissions from the ammonia industry could be reduced by 10% by taking measures to enhance the efficiency of existing plants. To reduce emissions to near-zero by 2050, more fundamental changes to production pathways are necessary. Low-carbon technologies such as water electrolysis and carbon capture and storage will need to be developed and deployed to achieve this near-zero target with minimum cost.

Deep Dive 2
Global mass flow of nitrogen fertilisers and mitigation options

Nitrogen fertilisers account for 5% of global greenhouse gas (GHG) emissions. Our global mass flow model uses a whole-systems approach to explore combinations of mitigation interventions. If all the interventions we consider were applied together, GHG emissions from synthetic fertilisers could be reduced by 84% by 2050. The single most effective intervention is increasing the efficiency of nitrogen fertiliser application to plants during their use phase.

Deep Dive 3
Recycled polymers as a feedstock for chemical manufacturing supply chains in the United States: A network analysis for polyethylene pyrolysis

If 5 Mt of PE waste were chemically recycled each year, and the resulting ethylene and propylene were put to use by the chemical industry, other sources of these chemicals would be displaced. The knock-on effects of this would span a large number of products and processes. These wider repercussions need to be considered in any evaluation of chemical recycling as an emissions-reduction strategy.

Deep Dive 5
The economic context

We need to consider the potential wider economic impacts of climate change mitigation actions taken within the petrochemical sector. Therefore, we have developed an integrated macroeconomic model that combines both material and economic flows. The model is calibrated for the US context so that it represents the real-world stocks and flows of mass, energy, population, and money.

Deep Dive 6
Why is a Network Analysis of the Petrochemical Industry useful?

A network analysis of the petrochemical industry reveals the complexity and interconnectedness of the sector and indicates how changes made in one plant or location would have implications elsewhere. This knowledge helps policymakers to target strategic locations or companies for emissions mitigation.

Deep Dive 4
Material flow analysis of petrochemicals production in the UK

We use a new methodology to map the emissions and energy use of the petrochemical sector in greater detail than existing studies. An effective, targeted emissions mitigation strategy requires both a big-picture view of the sector and an understanding of the granular detail so that broad goals can be translated into specific interventions. Here, we demonstrate the use of semantic web technologies, which allow us to navigate multiple levels of detail and aggregation with the same tool. This has not been possible before now.

Deep Dive 1
Model-based energy and emission analysis of ammonia production for improved process sustainability

Ammonia production is highly unsustainable. Our model-based analysis reveals that in the US, emissions from the ammonia industry could be reduced by 10% by taking measures to enhance the efficiency of existing plants. To reduce emissions to near-zero by 2050, more fundamental changes to production pathways are necessary. Low-carbon technologies such as water electrolysis and carbon capture and storage will need to be developed and deployed to achieve this near-zero target with minimum cost.

Deep Dive 2
Global mass flow of nitrogen fertilisers and mitigation options

Nitrogen fertilisers account for 5% of global greenhouse gas (GHG) emissions. Our global mass flow model uses a whole-systems approach to explore combinations of mitigation interventions. If all the interventions we consider were applied together, GHG emissions from synthetic fertilisers could be reduced by 84% by 2050. The single most effective intervention is increasing the efficiency of nitrogen fertiliser application to plants during their use phase.

Deep Dive 5
The economic context

We need to consider the potential wider economic impacts of climate change mitigation actions taken within the petrochemical sector. Therefore, we have developed an integrated macroeconomic model that combines both material and economic flows. The model is calibrated for the US context so that it represents the real-world stocks and flows of mass, energy, population, and money.

Deep Dive 6
Why is a Network Analysis of the Petrochemical Industry useful?

A network analysis of the petrochemical industry reveals the complexity and interconnectedness of the sector and indicates how changes made in one plant or location would have implications elsewhere. This knowledge helps policymakers to target strategic locations or companies for emissions mitigation.

Download this Deep Dive

Download pdf

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat.

Back to Start of Deep Dive