RES EARCH

PLASTICS of CO2-equiv (95%) are reduced by a biomass

pathway, where biomass uptake recycles CO2,
Achieving net-zero greenhouse gas emission compensating for the emissions primarily due
to plastic waste incineration and the produc-
plastics by a circular carbon economy tion of fossil-based feedstocks such as naphtha.
Whereas plastic waste and biomass provide
Raoul Meys1,2*, Arne Kätelhön1,2, Marvin Bachmann1, Benedikt Winter1,3, Christian Zibunas1, carbon and sufficient energy for conversion,
Sangwon Suh4, André Bardow1,3,5* CCU technologies require electricity with a low
carbon footprint to reduce GHG emissions,
Mitigating life-cycle greenhouse gas emissions of plastics is perceived as energy intensive and mainly to produce hydrogen by water elec-
costly. We developed a bottom-up model that represents the life cycle of 90% of global plastics to trolysis (19). For the current global average
examine pathways to net-zero emission plastics. Our results show that net-zero emission plastics carbon footprint of electricity (20), CCU would
can be achieved by combining biomass and carbon dioxide (CO2) utilization with an effective increase plastics’ GHG emissions. However, by
recycling rate of 70% while saving 34 to 53% of energy. Operational costs for net-zero emission using electricity with the current footprint of
plastics are in the same range as those for linear fossil-based production with carbon capture and wind power, commercialized CCU technologies
storage and could even be substantially reduced. Realizing the full cost-saving potential of can reduce up to 4.4 Gt of CO2-equiv (94%).

Downloaded from https://www.science.org at Indian Institute of Technology, Kharagpur on September 30, 2022
288 billion US dollars requires low-cost supply of biomass and CO2, high-cost supply of oil, and Overall, plastics solely based on recycling, bio-
incentivizing large-scale recycling and lowering investment barriers for all technologies that use mass utilization, or CCU fail to reach net-zero
renewable carbon feedstock. GHG emissions, even for wind-based electricity
production. Assuming wind-based electricity

B
supply, between 0.2 Gt and 1.7 Gt of CO2-equiv
etween 1950 and 2015, plastic production Our study shows that by combining recy- would have to be abated by negative-emission
increased from 2 to 380 million metric cling, biomass utilization, and CCU, net-zero technologies, such as direct air capture (DAC)
tons (Mt) year−1 (1), leading to the ever- GHG emission plastics could be achieved with with CO2 storage (21), to achieve net-zero plastics.
increasing plastic pollution of natural lower energy demands and lower operational In the single-technology pathways, waste in-
environments (2–5). The production of costs than those associated with current fossil- cineration, biomass, and renewable electric-
plastics is expected to increase its share of based production technologies combined with ity supply are the residual GHG sources that
global oil consumption from 6% today to 20% carbon capture and storage (CCS). For this prevent fully net-zero emission plastics: Even
by 2050 (6). Additionally, by 2050, plastics are purpose, we built and used a global bottom- though recycling rates are maximized, all re-
expected to reach more than 1100 Mt per year up model for plastics production and waste cycling processes produce residual wastes. These
and claim 15% of the yearly greenhouse gas treatment based on >400, mostly industrially residual wastes are incinerated, leading to un-
(GHG) emissions allotment to keep global validated, life-cycle assessment compliant and avoidable GHG emissions even for maximal
warming below 1.5°C (6). However, meeting harmonized technology datasets representing recycling rates. Additionally, waste incinera-
global climate targets will require net-zero the life cycle of >90% of global plastics. tion emits small amounts of non-CO2 emis-
GHG emissions by the second half of the Using this model, we project five pathways sions, such as carbon monoxide or methane,
century (7), so the life-cycle GHG emissions for life-cycle GHG emissions of plastic from increasing residual GHG emissions for biomass
of plastics must be reduced. Strategies to “cradle-to-grave” in the year 2050. The recycling utilization and CCU. Biomass cultivation emits
mitigate GHG emission include decarbonization pathway allows maximal recycling of all plastic noncarbon GHGs, such as nitrous oxide, that the
of the energy supply of the plastics supply chain wastes, on the basis of a 6% minimal landfilling CO2 uptake cannot counterbalance. Even in the
from oil extraction to plastic production, and the rate projected by Geyer et al. (1). By contrast, the most ambitious scenarios of the International
implementation of circular technologies (8–12)— biomass and CCU pathways assume that plastic Energy Agency (IEA) (19), renewable electricity
such as (i) chemical and mechanical recycling, waste is primarily incinerated. The resulting production is not entirely net zero by 2050, with
(ii) biomass utilization, and (iii) carbon capture CO2 emissions are circulated via biomass uptake ~13.5 g of CO2-equiv per kWh (22), leading to
and utilization (CCU)—to exchange the fossil or CCU. The circular carbon pathway optimally residual GHG emissions from electricity supply
carbon feedstock. combines recycling, biomass utilization, and in the CCU pathway. Even wind-based electricity
Recent literature on individual or partly com- CCU. The GHG emissions of all circular path- (i.e., electricity production with the lowest current
bined circular technologies shows large-scale ways are benchmarked to state-of-the-art plastic GHG emissions) is not fully net zero (23).
projected reductions in GHG emissions (13–18). production and waste incineration, framed as By contrast, the circular carbon pathway
However, no study identifies how circular tech- the linear carbon pathway. The current fossil- that optimally combines recycling, biomass
nologies can be combined to achieve net-zero based industry already includes recycling. Be- utilization, and CCU reduces GHG emission
emission plastics. Furthermore, utilization of cause landfill of plastics will increasingly fade of plastics by as much as 4.73 Gt of CO2-equiv,
circular technologies is constrained because (1), the remaining options for plastic waste assuming wind-based electricity (Fig. 1A). Thus,
circular technologies are generally associated treatment are energy recovery (represented by the plastics’ life cycle would even be slightly
with higher energy demands and costs (15). the linear carbon pathway) and the recycling net-negative. The shift is achieved because the
pathway. Thus, these two scenarios depict the combination of all circular technologies mini-
1
Institute for Technical Thermodynamics, RWTH Aachen
complete range of potential fossil-based futures mizes residual GHG emissions (see materials
University, Schinkelstr. 8, 52062 Aachen, Germany. 2Carbon for plastic waste treatment. and methods, as well as fig. S5 for a detailed
Minds GmbH, 50933 Cologne, Germany. 3Energy and Sankey diagram).
Process Systems Engineering, ETH Zürich, 8092 Zürich, Achieving net-zero emission plastics Biomass takes up CO2 and thereby offsets
Switzerland. 4Bren School of Environmental Science and
Management, University of California, Santa Barbara, CA, Our results show that the recycling pathway, CO2 emitted from waste incineration and pro-
USA. 5Institute of Energy and Climate Research–Energy via mechanical and chemical recycling, reduces duction processes. The respective biomass is
Systems Engineering (IEK-10), Forschungszentrum Jülich GHG emissions by 3.0 billion tons (Gt) of CO2 then gasified to generate synthesis gas, a mix-
GmbH, Jülich, Germany.
*Corresponding author. Email: abardow@ethz.ch (A.B.); raoul. equivalent (CO2-equiv) or 64% compared with ture of hydrogen and carbon monoxide, for
meys@ltt.rwth-aachen.de (R.M.) the linear carbon pathway (Fig. 1A). Up to 4.5 Gt methanol production. CO2 produced during

Meys et al., Science 374, 71–76 (2021) 1 October 2021 1 of 6

RES EARCH | R E S E A R C H A R T I C L E

Downloaded from https://www.science.org at Indian Institute of Technology, Kharagpur on September 30, 2022

Fig. 1. Potential reductions of global GHG emissions of plastics by four decreased electricity climate impacts, because all electricity is supplied
circular pathways in 2050. (A) Reductions of life-cycle GHG emissions from from the energy recovery via plastic waste incineration. (C) Optimal carbon
“cradle-to-grave” of the four pathways—recycling (Rec.), biomass (Bio), input in percentage of the circular carbon pathway, depending on the
carbon capture and utilization (CCU), and the circular carbon pathway carbon intensity of electricity. The IEA report (28) provides only regionalized
(Optimal)—depending on the electricity carbon footprint. av., average. electricity impacts until 2040. (D) Feedstock supply and waste treatment
(B) Remaining GHG emissions of the linear carbon pathways and the four of the circular carbon pathway for wind-based electricity production with
circular pathways, depending on the carbon intensity of electricity. The GHG 7 g of CO2-equiv per kWh. Line width and corresponding values represent the
emissions of the linear carbon pathway are not altered as a result of carbon content (million tons of C) of the flows.

Meys et al., Science 374, 71–76 (2021) 1 October 2021 2 of 6

RES EARCH | R E S E A R C H A R T I C L E

biomass gasification is captured by means of biomass, and CCU lead to varying synergies switch from partly bio-based to entirely CO2-
established technologies (e.g., the Rectisol pro- and electricity consumptions (Fig. 1B). As the based plastics reduces GHG emissions by only
cess; see materials and methods). No DAC is carbon intensity of electricity decreases, the 1.5%. Thus, using either biomass or CO2 as
employed. The captured CO2 from biomass circular carbon pathway with minimal GHG carbon feedstock results in net-zero emission
gasification is directly converted to methanol emissions employs more and more CCU tech- plastics if combined with large-scale plastic
by thermal hydrogenation. Subsequently, meth- nologies. This trade-off leads to break-even recycling.
anol can be used to produce ethylene and points for the carbon footprint of electricity
propylene, as well as benzene, toluene, and supply, at which CCU becomes climate ben- Renewable resource demands
xylene, which are the primary raw materials eficial compared with the recycling (39.5 g of Our analysis shows that combining CCU
for the 14 largest-volume plastic materials. The CO2-equiv per kWh) and biomass (5.8 g of and/or biomass with large-scale recycling can
related process technologies for conversion of CO2-equiv per kWh) pathways. The electricity achieve net-zero emission plastics. However,
methanol to olefins and aromatics are already supply’s carbon intensity also dictates the the actual feasibility will strongly depend on
industrialized. In 2017, 28% of the global meth- utilized carbon feedstock in the circular car- renewable resource availability. To this end,
anol production was used for ethylene and bon pathway (Fig. 1C): For electricity carbon two questions arise: Are there sufficient re-
propylene production (24). Thus, producing footprints above 8.6 g of CO2-equiv per kWh, newable resources to meet the global plastic
renewable methanol from biomass and cap- the optimal pathway is solely based on biomass demand? How does the circular carbon economy

Downloaded from https://www.science.org at Indian Institute of Technology, Kharagpur on September 30, 2022
tured CO2 offsets the primary source of residual and plastic waste as carbon feedstock (Fig. 1C). perform against other pathways for net-zero
GHG emissions in the recycling pathway: the Thus, these carbon inputs would be optimal for emission plastics (e.g., CCS)?
incineration of residual wastes. 2040 electricity impact predictions for China, The circular carbon pathway recycles 70% of
At the same time, recycling reduces the the United States, and Europe, which globally the plastic waste back to plastics and uses
overall demand for biomass and renewable produce 57% of all plastics (24, 28). 19.3 EJ of biomass and 9.9 PWh of renewable
electricity and the corresponding residual For electricity with carbon intensities below electricity (Fig. 1D). The effective recycling
GHG emissions. Two classes of recycling tech- 8.6 g of CO2-equiv per kWh, some CCU tech- rate of 70% is the maximum achievable,
nologies achieve the reduction of demand for nologies become beneficial to minimize GHG owing to losses in the recycling processes
biomass and renewable electricity. First, me- emissions (Fig. 1C). If wind-based electricity is in the best contemporary technologies and
chanical recycling is used to generate recycled employed, the optimal circular carbon pathway the residual landfilling of 6%. Thus, the effective
resins from plastic packaging waste. Each uses biomass, recycling, and CCU to produce recycling rate includes only the actually recycled
plastic resin produced via this process does 357 Mt of mechanically recycled plastic and material. For instance, material ultimately in-
not have to be produced from biomass or CO2. 814 Mt of virgin plastic. The virgin plastic cinerated during mechanical and chemical re-
Second, plastics are treated by pyrolysis to production consumes 429 Mt of chemical feed- cycling is not counted as recycled.
produce naphtha feedstock. This naphtha stock from chemical recycling, 2148 Mt of biomass, The resource demands can be shifted be-
feedstock is then used in steam crackers and and 949 Mt of CO2. Conversion of the inert tween biomass and renewable electricity because
solvent extraction processes to produce ethylene, molecule CO2 requires 9.9 PWh of renewable both CCU and biomass can achieve net-zero
propylene, benzene, toluene, and xylene. These electricity (Fig. 1D). emission plastics in combination with recycling.
chemical raw materials can be converted to A complete switch to CO2-based products By increasing biomass supply and thus reduc-
plastics by conventional, industrialized tech- requires electricity with a carbon intensity ing the supply of CO2 and vice versa, the
nologies, reducing the need for biomass uti- below 6 g of CO2-equiv per kWh. Today, not electricity demand can vary between 1.6 and
lization or CCU technologies. As a result of even wind power plants can provide such 18.1 PWh (Fig. 2, red line). These electricity
recycling, the overall demand for biomass low-emission electricity (29). However, the demands correspond to 59 to 670% of the
utilization or CCU technologies decreases, as
do their residual GHG emissions. Fig. 2. Demand for biomass
Beyond the reduction of residual GHG emis- and electricity to achieve
sions, part of the carbon taken up by biomass net-zero emission plastics
and CCU technologies is stored in landfilled in 2050. Data for the circular
plastics, given an unavoidable landfilling rate carbon pathway are shown
of 6% (1). This permanent carbon storage together with those for
compensates for the residual GHG emissions the linear carbon and recycling
that still occur in the circular carbon pathway. pathways with CCS. The
However, permanent carbon storage in land- range of CCS reflects differ-
fills cannot be seen as sustainable (15) because ent sources of CO2, such as
landfilling, managed or mismanaged, is the pri- ammonia plants (34) or
mary cause of plastic pollution (25). Nevertheless, ambient air (35). The energy
a certain amount of plastic leakage seems un- demands for the linear carbon
avoidable, even in the most ambitious policy scenarios are based on fossil
scenarios (26, 27). A very conservative assump- resources, which are
tion that the carbon in landfilled plastics en- converted to biomass and
tirely turns into CO2 would lead to additional electricity on an energy basis.
emissions of 0.24 Gt of CO2-equiv. In this case, The marker annotated with
the circular carbon pathway would lead to net “Fig. 1D” refers to the Sankey
emissions of 0.21 Gt of CO2-equiv. diagram in Fig. 1. In both
The technologies and carbon feedstocks that cases, the same amount
minimize GHG emissions of plastics depend of electricity and biomass
on the electricity supply’s carbon intensity, are used.
because different combinations of recycling,

Meys et al., Science 374, 71–76 (2021) 1 October 2021 3 of 6

RES EARCH | R E S E A R C H A R T I C L E

electricity predicted to be supplied to the pathway that would need 4.7 Gt of CO2 storage tion. Energy recovery is inefficient because of
chemical industry (2.7 PWh) in the IEA’s (e.g., using CCS) to reach net-zero emission unavoidable losses from thermal energy to
most ambitious scenario for 2050 (30). How- plastics. In this case, 76.9 EJ of fossil-based en- electricity conversion that result from thermo-
ever, the electricity supply calculated by the ergy and 1.9 to 33.9 EJ of additional electricity dynamic limitations. Therefore, plastic waste
IEA follows a cost-optimal demand in the for CCS would be required (Fig. 2, gray area). incineration will inevitably never suffice to close
petrochemical industry, to achieve a specific The range of electricity reflects the nature of the the energy loop and maintain 100% of the
emission-reduction target. Thus, the calculated CO2 source, ranging from a high-concentration energy content. By contrast, recycling essen-
electricity supply approximates the electricity stream [e.g., produced during ammonia synthe- tially conserves the energy content of plastics
amount that will be supplied, not the amount sis (34)] to the dilute scenario of DAC (35). The by reusing plastic waste and avoiding its in-
that could potentially be supplied to the chem- industrial-scale separation of CO2 from the gas cineration, thus lowering energy demands. A
ical industry. Furthermore, the plastics industry streams during ammonia and hydrogen pro- major difference is that the linear pathways can
is not entirely covered by the IEA model, and duction is well established and frequently per- exploit fossil energy generated over millennia,
thus the electricity demands for plastics are not formed by the Rectisol process considered here whereas the biomass and CCU pathways need
fully represented. However, ~80% of the overall (36, 37). However, currently, the separated CO2 to generate their energy now.
mass of the major bulk chemicals (e.g., ethylene, is not stored and most often is simply released To achieve these reduced energy demands,
propylene, benzene, toluene, and xylene) are to the atmosphere. For DAC, recently published the circular carbon economy can rely on com-

Downloaded from https://www.science.org at Indian Institute of Technology, Kharagpur on September 30, 2022
used for plastics. Thus, we assume the 2.7 PWh data for the Climeworks plants in Hellisheiði, mercialized recycling technologies. All tech-
of electricity supplied to the chemical industry Iceland, and Hinwil, Switzerland, are used, as- nology datasets used in the main paper are
as a reasonable estimate for a possible amount suming wind-based electricity supply (35). based on already-commercialized technologies
of electricity that can be supplied to plastics life Overall, the linear carbon pathway plus CCS using industrially validated data or process
cycles. As a result, we denote the combination would consume between 78.8 and 110.8 EJ simulations. Sorting household plastics and
of 2.7 PWh electricity with 42.6 EJ of biomass of energy. Thus, the circular carbon pathway mechanical recycling are already industrial
to achieve net-zero emission plastics as the at the feasible point, with a total energy con- practice in western European countries such
“feasible point” (Fig. 2). sumption of 42.6 EJ of biomass and 9.7 EJ of as Germany and Austria (38). Whereas plastic
Estimates of biomass availability vary widely electricity, could potentially save between 34 packaging can be efficiently recycled mechan-
from 30 EJ to >1000 EJ per year (31). An expert and 53% of the total energy demand. ically, mixed and other plastic wastes lack
consortium found that an estimate of 100 EJ is Energy demands can be further reduced by mechanical recyclability (39, 40). In this case,
supported by “high agreement” in literature combining the recycling pathway with CCS to pyrolysis offers a promising large-scale ave-
(32). The International Renewable Energy achieve net-zero emission plastics (Fig. 2, orange nue to increase recycling rates (41).
Agency (IRENA) estimates that an additional area). This combination requires 40.6 to 52.2 EJ Additionally, early-stage technologies to con-
287 EJ are available (22, 33) from using while achieving net-zero emission plastics, re- vert plastic waste to respective monomers are
lignocellulosic biomass and food wastes as well ducing energy by an additional 1 to 12% com- currently under development (42, 43). To high-
as land made available by technology and pared with the circular carbon pathway. Thus, light the potential benefits of technology de-
farming improvements. Considering the future the total life-cycle energy demand differs by up velopment, we include emerging technologies
biomass demand (104 EJ) of all other sectors to 12% between CCS, biomass, and CCU. How- in the supplementary materials (see materials
(30), IRENA’s estimated untapped biomass sup- ever, in all cases, recycling plastic waste reduces and methods for modeling principles and tech-
ply equals 183 EJ. Thus, the biomass demand of the energy demand relative to the fossil-based nology pathways). By leveraging potential chem-
42.6 EJ at the feasible point would represent benchmark with CCS. ical recycling, biomass utilization, and CCU
23% of this remaining untapped biomass po- Although the decreased energy demand may technologies with low technology readiness
tential. However, the access to this untapped appear counterintuitive because of the antici- levels (i.e., the technology has been maximally
biomass relies on boosting crop yield, reduc- pated lower efficiency of biomass and CO2 validated in the lab), the circular carbon path-
ing food waste, afforestation, and improving conversion, it can be rationalized by energy way could reduce the energy demand by 83%
livestock management to free pastureland conservation over the complete life cycle and relative to the linear carbon pathway with CCS
(22, 33). recycling: The pathways based on fossil sources, (see materials and methods and figs. S3 and S4).
To put the renewable energy demand into biomass, and CO2 can recover the energy con- However, even promising recycling technol-
perspective, we compare it to the linear carbon tained in plastics only during waste incinera- ogies at low technology readiness levels cannot

Table 1. Operational costs of linear and circular carbon pathways at the feasible point. Operational costs are shown for oil, biomass, CO2, and
electricity, as well as for the amount of waste treated by mechanical (Mech.) recycling, chemical (Chem.) recycling, and energy recovery (left to right). Prices
are provided in table S12. / indicates zero operating cost.

Oil Biomass CO2 Electricity Mech. recycling Chem. recycling Energy recovery Total
Prices, pathway
(billion USD) (billion USD) (billion USD) (billion USD) (billion USD) (billion USD) (billion USD) (billion USD)
............................................................................................................................................................................................................................................................................................................................................
Low prices, linear
675 / / / / / 164 839
carbon pathway
............................................................................................................................................................................................................................................................................................................................................
High prices, linear
946 / / / / / 164 1110
carbon pathway
............................................................................................................................................................................................................................................................................................................................................
Low prices, circular
18 212 1 54 82 413 42 822
carbon pathway
............................................................................................................................................................................................................................................................................................................................................
High prices, circular
25 639 3 162 82 413 42 1366
carbon pathway
............................................................................................................................................................................................................................................................................................................................................

Meys et al., Science 374, 71–76 (2021) 1 October 2021 4 of 6

RES EARCH | R E S E A R C H A R T I C L E

leverage their potential unless sufficient plastic bon cost of 112 USD per kg of CO2-equiv is in the problem in shifting from one stage of the life
waste is collected and made available. A recent predicted range of prices for CO2 certificates in cycle to another. Including plastic waste in-
study predicted that, in 2040, 88% of the plastic Europe or mid-century social costs of carbon cineration in emission pricing schemes through,
demand will still be lost in managed landfills (48, 49), indicating that, with regard to opera- e.g., extended producer responsibility policy
and waste incinerators (32% combined) and tional costs, current carbon price scenarios would would be a step forward to incentivize recycling
through waste mismanagement (56%) (27). To suffice to reach net-zero emission plastics. (53). For carbon pricing to effectively reduce
increase collection rates, landfill bans have The abatement costs are based only on the GHG emissions throughout the plastics’ life
proven highly effective in European countries operational costs and do not include capital cycle and improve the circularity of plastics,
such as Germany, the Netherlands, and Sweden. expenditures. However, an increase in capital the entire life cycle of plastics must be covered
In Europe, landfilling dropped by 34%, whereas expenditures can be expected to provide plastic, within the scope of carbon pricing.
recycling increased by 64% between 2006 biomass, or CO2 as resources. These investments Additionally, local municipalities often con-
and 2014 (44). Once the collected waste is in are needed for the infrastructure to provide the duct waste management as a service, particu-
a managed system, recycling can be fostered renewable carbon feedstocks to the production larly collection and sorting, for residents and/or
by implementing recycling quotas. However, facilities of the chemical industry and to convert local companies. Because the residents pay
a proper waste management infrastructure is the renewable feedstock to plastic products. for the service, incentives to generate a usable
missing in many low- and middle-income coun- Although a full assessment of capital expendi- waste stream are limited. In return, munici-

Downloaded from https://www.science.org at Indian Institute of Technology, Kharagpur on September 30, 2022
tries that would offer the largest potential to tures is not possible, owing to the lack of data in palities often receive postconsumer waste of
access plastic waste (26, 45). To gain access to this publication, the increase can be estimated low value and poor quality. These postcon-
these untapped resources, millions of households from a recent calculation by the IEA. To reduce sumer wastes are then treated at the lowest
must be connected to waste management ser- the GHG emissions of the petrochemical sector, potential costs (e.g., waste incineration or, even
vices. Lau et al. argue that this monumental task which is closely connected to the plastic industry worse, landfilling) instead of being properly
requires linking local service chains [e.g., the (50), the IEA estimated an investment need of sorted. Thus, collection and sorting are commonly
informal sector (4)] to the value chain (e.g., 1.5 trillion USD (51) to save 0.9 Gt of CO2-equiv the bottlenecks of recycling industries (39). Policy-
recycling) by increasing the profitability of yearly after 2050. Although these investment makers should aim to incentivize value addition at
material recycling through investments in amounts do not fully represent those required the beginning of the waste value chain: the plastic
waste management infrastructure and improved to achieve net-zero emission plastics, they clear- consumer. In this scenario, a deposit system for
coordination for collection, sorting, and manage- ly indicate that the additional capital expendi- plastic materials could provide a potential avenue
ment of plastic wastes (27). tures will increase abatement costs. Assuming for consumers to provide valuable plastic waste
the same capital expenditures as those of the feedstock directly from the start. These deposit
Operational cost IEA, an average production plant lifetime of systems have been very successful in combina-
Operational costs differ between the linear 30 years, and 0.9 Gt of CO2-equiv savings for tion with landfill bans in Europe (44).
and circular carbon pathways in two major these years, the abatement cost would increase Although increasing the availability of plastic
aspects: (i) the costs for feedstock and energy by ~56 USD per ton of CO2. However, a more in- waste is one important point to be addressed
to produce plastics and (ii) the costs for “end- depth assessment of capital expenditures is de- through policy-making. Current policies also
of-life” treatment of postconsumer plastics (e.g., sirable to guide future investment decisions and subsidize oil exploration and production of
landfilling, energy recovery, and chemical and properly define the respective CO2 abatement fossil products and thereby offer a cheap and
mechanical recycling). Taking into account the costs to achieve net-zero emission plastics. abundant alternative to plastic waste, biomass,
expected price ranges (see materials and meth- and CO2 as renewable carbon feedstocks. In the
ods and table S11) for biomass, CO2, electricity, Policies for circular and net-zero emission oil and gas industry, investments in the exten-
and oil (30, 46) as well as the operational cost plastics sive infrastructure have already amortized.
for mechanical recycling (47) chemical recy- Our results indicate that net-zero emission By contrast, plastic waste, biomass, and CO2
cling (36), and waste incineration (47), the plastics can be achieved by using technolo- utilization are in their infancy, and investment
operational costs of the circular carbon path- gies that are already available and commer- possibilities currently do not satisfy private in-
way vary between 822 and 1366 billion USD cialized. To materialize the potential, policies vestors because of their lower return on invest-
on a global basis in 2050 (Table 1). Thus, the that foster the deployment of circular carbon ment. As a result, the initial capital investment
circular carbon pathway lies in the same cost technologies need to be designed and imple- largely disincentives potential investors, partic-
range as the linear carbon pathway, which has mented. We identify two crucial technolog- ularly in the recycling area (54). However, in-
operational costs between 839 and 1110 billion ical changes that will be necessary to achieve vestment firms such as BlackRock are becoming
USD. Assuming costs at the low end for all net-zero emission plastics: (i) increase plastic increasingly interested in including climate im-
resources would lead to nearly identical cost for recycling rates and supply more plastic waste pact in their investment decisions (55).
the linear carbon pathway (839 billion USD) feedstock, and (ii) deploy CCU or biomass tech- Thus, the use of globally agreed policy in-
and the circular pathway (822 billion USD). For nologies, depending on local availabilities of struments (56) to increase the availability of
high oil prices (~70 USD per barrel) as well as renewable electricity and biomass. Fostering plastic waste as a resource and provide eco-
low-cost supply of biomass, CO2, and renewable such changes requires economic incentives. nomic incentives for increased investment in
electricity (~5 USD per GJ, 30 USD per ton, and Economic incentives can play a crucial role biomass and CO2 utilization can advance the
2 USD per kWh, respectively), the circular in increasing plastics’ circularity and achieving pathway toward net-zero emission plastics.
carbon pathway would save 288 billion USD net-zero emission plastics. However, under the However, the improvements regarding energy
in operational cost. current structure, pricing carbon emissions consumption and GHG emissions must be
Depending on the resource prices, the respec- would have, at best, limited impact in in- carefully balanced with other environmental
tive CO2 abatement of 4.7 Gt of CO2-equiv would centivizing plastic circularity. Current emis- impacts known to arise from large-scale usage
cost −61 to 112 USD per kg of CO2-equiv abated. sions trading schemes (such as the EU ETSs) of biomass or renewable electricity. These en-
This range has considerable uncertainty, owing focus on production processes and exclude vironmental impacts include an increase of
to the broad price ranges for oil, biomass, CO2, end-of-life management processes (52) such terrestrial acidification and water eutrophication
and electricity. However, even the high-end car- as incineration from the scope, leading to a due to biomass utilization (57) and elevated

Meys et al., Science 374, 71–76 (2021) 1 October 2021 5 of 6

RES EARCH | R E S E A R C H A R T I C L E

metal depletion and water consumption due 24. ICIS Supply and Demand Database (2017); www.icis.com/ 52. European Commission, “Guidance on Interpretation of Annex I
to the increase of solar and wind-based explore/services/analytics/supply-demand-data/icis-supply- of the EU ETS Directive (excl. aviation activities)” (DG Climate,
and-demand-database/. 2010); https://ec.europa.eu/clima/sites/clima/files/ets/
power plants (58). Looking ahead, the circular 25. P. Gabrielli, M. Gazzani, M. Mazzotti, Ind. Eng. Chem. Res. 59, docs/guidance_interpretation_en.pdf.
carbon economy could hold its promise to 7033–7045 (2020). 53. W. Leal Filho et al., J. Clean. Prod. 214, 550–558 (2019).
redesign plastics production systems such that 26. S. B. Borrelle et al., Science 369, 1515–1518 (2020). 54. W. Gao, T. Hundertmark, T. Simons, J. Wallach, C. Witte,
27. W. W. Y. Lau et al., Science 369, 1455–1461 (2020). “Plastics recycling: Using an economic-feasibility lens to select
decoupling from fossil carbon resources achieves 28. IEA, “Tracking power 2020” (IEA, 2020); www.iea.org/reports/ the next moves” (McKinsey & Company, 2021); www.mckinsey.
net-zero emission plastics with lower energy tracking-power-2020. com/industries/chemicals/our-insights/plastics-recycling-
demands while requiring reduced operational 29. S. Schlömer et al., in “Climate change 2014: Mitigation of using-an-economic-feasibility-lens-to-select-the-next-moves.
climate change. Contribution of Working Group III to the Fifth 55. BlackRock, “Larry Fink’s 2021 letter to CEOs” (2021);
costs and, thus, combining economic and en- Assessment Report of the Intergovernmental Panel on Climate www.blackrock.com/corporate/investor-relations/larry-fink-
vironmental well-being (6). Change,” O. Edenhofer et al., Eds. (Cambridge Univ. Press, ceo-letter.
2014), p. 1335. 56. N. Simon et al., Science 373, 43–47 (2021).
RE FE RENCES AND N OT ES 30. IEA, “Energy technology perspectives 2020” (IEA, 2020); 57. S. Walker, R. Rothman, J. Clean. Prod. 261, 121158 (2020).
www.iea.org/reports/energy-technology-perspectives-2020. 58. G. Garcia-Garcia, M. C. Fernandez, K. Armstrong, S. Woolass,
1. R. Geyer, J. R. Jambeck, K. L. Law, Sci. Adv. 3, e1700782 (2017).
31. M. B. Jones, F. Albanito, Glob. Change Biol. 26, 5358–5364 P. Styring, ChemSusChem 14, 995–1015 (2021).
2. R. G. Santos, G. E. Machovsky-Capuska, R. Andrades, Science
(2020). 59. R. Meys, Bene94/Plastic-supply-chain-model-including-matlab-
373, 56–60 (2021).
32. F. Creutzig et al., GCB Bioenergy 7, 916–944 (2015). code-: V1.0, version 1.0, Zenodo (2021); https://doi.org/10.
3. M. MacLeod, H. P. H. Arp, M. B. Tekman, A. Jahnke, Science
33. International Renewable Energy Agency, “Boosting biofuels: 5281/zenodo.5118762.
373, 61–65 (2021).
Sustainable paths to greater energy security” (IRENA, 2016).
4. S. Kaza, L. C. Yao, P. Bhada-Tata, F. van Woerden, What a

Downloaded from https://www.science.org at Indian Institute of Technology, Kharagpur on September 30, 2022
34. N. von der Assen, L. J. Müller, A. Steingrube, P. Voll, A. Bardow, AC KNOWLED GME NTS
Waste 2.0: A Global Snapshot of Solid Waste Management to
Environ. Sci. Technol. 50, 1093–1101 (2016). A.B. and R.M. would like to thank N. Tigu for technical support.
2050 (World Bank, 2018).
35. S. Deutz, A. Bardow, Nat. Energy 6, 203–213 (2021). Funding: This work was supported by US Environmental Protection
5. A. Stubbins, K. L. Law, S. E. Muñoz, T. S. Bianchi, L. Zhu,
36. IHS Markit, Ed., Process Economics Program (PEP) Yearbook Agency grant 83557907 (S.S.); EU Horizon 2020 Framework
Science 373, 51–55 (2021).
(IHS Markit, 2018).
6. Ellen MacArthur Foundation, Ed., “The new plastics economy: Programme grant 720695 (B.W.); the Bundesamt für Energie
37. B. Elvers, F. Ullmann, Eds., Ullmann’s Encyclopedia of Industrial SWEET PATHFNDR project (A.B.); and the Ministry of Economic
Rethinking the future of plastics” (Ellen MacArthur Foundation, 2016).
Chemistry (Wiley, ed. 7, 2011). Affairs, Innovation, Digitalization and Energy of the State of
7. J. Rogelj et al., Nat. Clim. Chang. 8, 325–332 (2018).
38. A. Reichel, M. Schoenmakere, J. Gillabel, “Circular economy in North Rhine Westphalia Sci4Climate project (C.Z.). Author
8. L. T. J. Korley, T. H. Epps 3rd, B. A. Helms, A. J. Ryan, Science
Europe: Developing the knowledge base” (EEA report no. 2/ contributions: Conceptualization: R.M., S.S., and A.B.
373, 66–69 (2021).
2016, Publications Office of the European Union, 2016). Methodology: R.M. and A.K. Data curation: R.M., M.B., B.W., C.Z.,
9. R. Altman, Science 373, 47–49 (2021).
39. K. Ragaert, L. Delva, K. Van Geem, Waste Manag. 69, 24–58 and A.K. Supervision: A.B. Writing – original draft: R.M. and A.B.
10. S. Kakadellis, G. Rosetto, Science 373, 49–50 (2021).
(2017). Writing – review & editing: R.M., A.K., S.S., A.B., M.B., B.W., and
11. A. M. Bazzanella, F. Ausfelder, “Low carbon energy and
40. K. Kümmerer, J. H. Clark, V. G. Zuin, Science 367, 369–370 C.Z. Competing interests: A.B. served on review committees for
feedstock for the European chemical industry” (DECHEMA
(2020). research and development at ExxonMobil and TotalEnergies, oil
Gesellschaft für Chemische Technik und Biotechnologie, 2017).
41. H. Jeswani et al., Sci. Total Environ. 769, 144483 (2021). and gas companies that are also active in polymer production.
12. J. B. Zimmerman, P. T. Anastas, H. C. Erythropel, W. Leitner,
42. R. Meys et al., Resour. Conserv. Recycling 162, 105010 (2020). Data and materials availability: All data are publicly available.
Science 367, 397–400 (2020).
13. J. Zheng, S. Suh, Nat. Clim. Chang. 9, 374–378 (2019). 43. M. Hong, E. Y.-X. Chen, Green Chem. 19, 3692–3706 However, some cases require a user license from IHS Markit to
14. B. G. Hermann, K. Blok, M. K. Patel, Environ. Sci. Technol. 41, (2017). access the underlying data. To gain access, IHS Markit can be
7915–7921 (2007). 44. PlasticsEurope, “Plastics – the Facts 2018. An analysis of contacted via the following website: https://ihsmarkit.com/
15. A. Kätelhön, R. Meys, S. Deutz, S. Suh, A. Bardow, Proc. Natl. European plastics production, demand and waste data” products/chemical-technology-pep-index.html. All data that do not
Acad. Sci. U.S.A. 116, 11187–11194 (2019). (PlasticsEurope, 2018); www.plasticseurope.org/application/ require an additional license are archived at Zenodo (59). The
16. I. D. Posen, P. Jaramillo, A. E. Landis, W. M. Griffin, Environ. files/6315/4510/9658/Plastics_the_facts_2018_AF_web.pdf. Zenodo repository also includes the MATLAB code to use the data
Res. Lett. 12, 034024 (2017). 45. J. R. Jambeck et al., Science 347, 768–771 (2015). and generate the results.
17. A. E. Schwarz et al., Waste Manag. 121, 331–342 (2021). 46. IEA, “World Energy Outlook 2018” (IEA, 2018); www.iea.org/weo/.
18. Á. Galán-Martín et al., One Earth 4, 565–583 (2021). 47. N. S. Rudolph, C. Aumanate, R. Kiesel, Understanding Plastics
19. J. Artz et al., Chem. Rev. 118, 434–504 (2018). Recycling: Economic, Ecological, and Technical Aspects of SUPPLEMENTARY MATERIALS
20. G. Wernet et al., Int. J. Life Cycle Assess. 21, 1218–1230 (2016). Plastic Waste Handling (Hanser, 2017).
science.org/doi/10.1126/science.abg9853
21. S. Fuss et al., Environ. Res. Lett. 13, 063002 (2018). 48. European Commission, “Energy Roadmap 2050 - Impact
Materials and Methods
22. International Renewable Energy Agency, “Global energy assessment and scenario analysis” (European Commission, 2011).
Fig. S1 to S6
transformation. A roadmap to 2050” (IRENA, 2019). 49. K. Ricke, L. Drouet, K. Caldeira, M. Tavoni, Nat. Clim. Chang. 8,
Tables S1 to S13
23. IPCC, “Climate change 2013: The physical science basis. 895–900 (2018).
References (60–122)
Contribution of Working Group I to the Fifth Assessment 50. D. Gielen, J. Newman, M. K. Patel, MRS Bull. 33, 471–477 (2008).
Report of the Intergovernmental Panel on Climate Change,” 51. IEA, “The future of petrochemicals” (IEA Publications 23 April 2021; accepted 20 August 2021
T. F. Stocker et al., Eds. (Cambridge Univ. Press, 2013). International Energy Agency, 2018). 10.1126/science.abg9853

Meys et al., Science 374, 71–76 (2021) 1 October 2021 6 of 6

 Achieving net-zero greenhouse gas emission plastics by a circular carbon
economy
Raoul MeysArne KätelhönMarvin BachmannBenedikt WinterChristian ZibunasSangwon SuhAndré Bardow

Science, 374 (6563), • DOI: 10.1126/science.abg9853

Downloaded from https://www.science.org at Indian Institute of Technology, Kharagpur on September 30, 2022
Reducing net emission
The great majority of plastics in current use are sourced from fossil fuels, with additional fossil fuels combusted
to power their manufacture. Substantial research is focused on finding more sustainable building blocks for next-
generation polymers. Meys et al. report a series of life cycle analyses suggesting that even the current varieties of
commercial monomers could potentially be manufactured and polymerized with no net greenhouse gas emissions.
The cycle relies on combining recycling of plastic waste with chemical reduction of carbon dioxide captured from
incineration or derived from biomass. —JSY

View the article online

https://www.science.org/doi/10.1126/science.abg9853
Permissions
https://www.science.org/help/reprints-and-permissions

Use of this article is subject to the Terms of service

Science (ISSN ) is published by the American Association for the Advancement of Science. 1200 New York Avenue NW, Washington, DC
20005. The title Science is a registered trademark of AAAS.
Copyright © 2021 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim
to original U.S. Government Works