Climate change mitigation potential of carbon capture and utilization in the chemical

industry
Author(s): Arne Kätelhön, Raoul Meys, Sarah Deutz, Sangwon Suh and André Bardow
Source: Proceedings of the National Academy of Sciences of the United States of America ,
June 4, 2019, Vol. 116, No. 23 (June 4, 2019), pp. 11187-11194
Published by: National Academy of Sciences

Stable URL: https://www.jstor.org/stable/10.2307/26706428

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

National Academy of Sciences is collaborating with JSTOR to digitize, preserve and extend access to
Proceedings of the National Academy of Sciences of the United States of America

This content downloaded from

111.68.102.25 on Mon, 07 Oct 2024 03:21:41 UTC
All use subject to https://about.jstor.org/terms
Climate change mitigation potential of carbon capture
and utilization in the chemical industry
Arne Kätelhöna, Raoul Meysa, Sarah Deutza, Sangwon Suhb, and André Bardowa,c,1
a
Institute of Technical Thermodynamics, RWTH Aachen University, 52062 Aachen, Germany; bBren School of Environmental Science and Management,
University of California, Santa Barbara, CA 93117; and cInstitute of Energy and Climate Research–Energy Systems Engineering (IEK-10), Forschungszentrum
Jülich, GmbH, 52425 Jü lich, Germany

Edited by Alexis T. Bell, University of California, Berkeley, CA, and approved April 9, 2019 (received for review December 10, 2018)

Chemical production is set to become the single largest driver of structure as conventional production. In this case, there is no benefit
global oil consumption by 2030. To reduce oil consumption and from the temporary storage of carbon within the chemicals during the
resulting greenhouse gas (GHG) emissions, carbon dioxide can be use phase because the same amount of carbon is stored for the same
captured from stacks or air and utilized as alternative carbon time (see SI Appendix for further discussion).
source for chemicals. Here, we show that carbon capture and While climate benefits of some CCU technologies have been
utilization (CCU) has the technical potential to decouple chemical reported individually, the global potential of CCU for climate
production from fossil resources, reducing annual GHG emissions change mitigation is still under debate (13), hampering the in-
by up to 3.5 Gt CO2-eq in 2030. Exploiting this potential, however, tegration of CCU into policy frameworks and technology road-
requires more than 18.1 PWh of low-carbon electricity, corre- maps (14, 15). Previous estimates (13) of the climate change
sponding to 55% of the projected global electricity production in mitigation potential of CCU relied on the amount of CO2 stored
2030. Most large-scale CCU technologies are found to be less effi- in products. However, the amount of CO2 stored can differ sub-

ENVIRONMENTAL
cient in reducing GHG emissions per unit low-carbon electricity stantially from the potential GHG emission reductions due to
when benchmarked to power-to-X efficiencies reported for other

SCIENCES
CCU; using, e.g., 1 kg CO2 for the production of CO2-based
large-scale applications including electro-mobility (e-mobility) and
polymers can reduce up to 3 kg CO2-eq emissions due to the
heat pumps. Once and where these other demands are satisfied,
avoided conventional production (16). Thus, a comprehensive
CCU in the chemical industry could efficiently contribute to climate
analysis of the climate change mitigation potential of CCU re-
change mitigation.
quires a detailed assessment from a life cycle perspective (17, 18).
In this study, we examine the technical potential for climate
|
climate change carbon capture and utilization | chemicals | change mitigation by CCU in the chemical industry. The tech-
|

SUSTAINABILITY
renewable energy circular economy
nical potential is defined here as the maximum GHG emission

SCIENCE
reductions that are technically feasible through full deployment
he chemical industry consumes ∼28% of the industrial and
T ∼10% of the global final energy (1). Most of this energy is
obtained from oil and other fossil resources, which are either
of CCU technologies (19) with wind power as low-carbon elec-
tricity input. We establish a bottom-up model covering the pro-
duction of 20 large-volume chemicals accounting for over 75% of
used as carbon feedstock (58%) or to generate process energy the chemical industry’s GHG emissions (20). Based on this
(42%) (1). While other uses of fossil resources can be directly model, we analyze potential disruptive changes through large-
replaced by renewable energy, the need for a carbon feedstock is
scale CO2 utilization and provide an industry-wide analysis of
expected to make the chemical industry the largest driver for the
future increase in global oil demand (2). The chemical industry
Significance
already reduced the use of fossil resources and the resulting
greenhouse gas (GHG) emissions substantially through process
intensification and energy efficiency measures, but the potential Carbon dioxide (CO2) drives climate change when released to
of these measures has largely been exhausted (3). Consequently, the atmosphere. Alternatively, CO2 could be captured and
deeper cuts in GHG emissions require more disruptive change utilized as carbon source for chemicals. Here, we provide a
such as the use of biomass feedstock (4), electrification (5), global assessment of the technical climate change mitigation
potential of carbon capture and utilization (CCU) in the chem-
carbon capture and storage (6), and carbon capture and utili-
ical industry. We develop an engineering-level model of the
zation (CCU) (3). Here, we analyze the technical potential of
global chemical industry representing 75% of current green-
CCU to mitigate GHG emissions.
house gas (GHG) emissions. The model allows us to analyze the
CCU technologies can substitute the conventional production
potential disruptive changes through large-scale CO2 utiliza-
of various chemicals including basic chemicals, fine chemicals, and
tion and resulting emission reductions. Our study shows that
polymers (7–11). These technologies capture carbon dioxide (CO2)
CCU has the technical potential to lead to a carbon-neutral
from industrial point sources or ambient air and use the captured
chemical industry and decouple chemical production from
CO2 as alternative carbon source for chemical production. Replacing fossil resources. This transition, however, would cause largely
conventional chemical production through CCU affects GHG emis- increased mass flows and demand for low-carbon electricity.
sions throughout the production chain (i.e., from cradle-to-gate; Fig.
1) by (i) removing CO2 from air or from stacks, (ii) avoiding the Author contributions: A.K., R.M., S.S., and A.B. designed research; A.K., R.M., S.D., and
GHG emissions necessary for conventional chemical production in- A.B. performed research; A.K., R.M., S.D., S.S., and A.B. analyzed data; and A.K., R.M., S.S.,
cluding the supply of raw materials and energy, and (iii) adding GHG and A.B. wrote the paper.

emissions necessary for CO2 capture and the conversion of CO2 into The authors declare no conflict of interest.
chemicals. Many CCU technologies use hydrogen from electricity- This article is a PNAS Direct Submission.
driven water electrolysis to activate the inert CO2 molecule, offer- Published under the PNAS license.
ing a route to integrate renewable electricity more significantly into 1
To whom correspondence should be addressed. Email: [email protected].
chemical life cycles (12). In contrast, CCU technologies do not affect This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
the use phase and end-of-life (i.e., gate-to-grave) GHG emissions of 1073/pnas.1821029116/-/DCSupplemental.
chemicals if they produce the chemicals with an identical molecular Published online May 13, 2019.

www.pnas.org/cgi/doi/10.1073/pnas.1821029116 PNAS | June 4, 2019 | vol. 116 | no. 23 | 11187–11194

This content downloaded from

111.68.102.25 on Mon, 07 Oct 2024 03:21:41 UTC
All use subject to https://about.jstor.org/terms
 carbon source production this scenario is denoted “low-TRL scenario.” The direct syn-
CO2 thesis of chemicals from CO2 reduces the amount of CO2 and
conventional production

hydrogen needed to 2.77 and 0.38 Gt, respectively, but mass

CO2
flows are still about 182% larger compared with the conven-
tional scenario. Due to the early development stages of these
use phase end-of-life technologies, the realization of this scenario would require

chemically identical product

substantial research and development efforts and novel pro-
CO2 duction facilities. Since this scenario is based on low-TRL
CO2 technologies, all calculations are subject to increased data un-
C certainty. The employed methodology to estimate life cycle
assessment (LCA) results for low-TRL technologies is pre-
CO2
CO2
sented in SI Appendix.
In the CCU scenarios, hydrogen produced via water electrolysis
CO2-based production

CO2 CO2 CO2

CO2
CO2

CO2
CO2 represents the major source of energy for the chemical industry.
CO2
CO2

CO2

CO2
Electricity demands amount to about 32.0 PWh in the high-TRL
CO2
(temporary) carbon storage scenario and 18.1 PWh in the low-TRL scenario. Electricity de-
mands could be reduced through direct use of hydrogen from
water electrolysis. For example, the direct use of hydrogen in
ammonia production could reduce the electricity demand in the
C low-TRL scenario by 7.5% by avoiding the steps for methanation
and steam reforming in the production of ammonia from CO2-
product life cycle from cradle-to-grave based methane (details in SI Appendix). However, these pro-
duction pathways are not based on CO2 and therefore excluded
from the CCU scenarios. For the CCU scenarios, the electricity
Fig. 1. Changes in the life cycle of established chemicals through the
demands in the high- and low-TRL scenario correspond to about
implementation of CCU technologies.
97% and 55% of the expected world electricity production in
2030 according to the new policy scenario of the International
CCU technologies’ technical potential to mitigate GHG emis- Energy Agency (IEA) (31), respectively. Hence, CCU in the
sions from a life cycle perspective. chemical industry could only reduce GHG emissions on the large
scale with the joint massive expansion of electricity production
Results capacities. As a result, the carbon footprint of electricity from the
Scenarios for a CCU-Based Chemical Industry. The bottom-up model technologies used to expand the electricity production capacities
of the chemical industry yields future production pathways for will determine the climate benefits of CCU.
the production of the 20 large-volume chemicals in 2030 (20) for
Climate Change Mitigation Depending on Electricity Supply. We
three scenarios. In the “conventional scenario” (Fig. 2A), the
found that the GHG emission reductions of replacing conven-
chemical industry converts 1.12 Gt of fossil resources into
tional chemical production through CCU varies widely depend-
chemicals and generates about 2.0 Gt CO2-eq cradle-to-gate
ing on the carbon footprint of electricity (Fig. 3). Our results
emissions using conventional production technologies that are
show that CCU can substantially reduce GHG emissions in both
already commercialized today. Due to increased demand, this
CCU scenarios, but only if low-carbon electricity is available. In
estimate is higher than current GHG emissions of the chemical the high-TRL scenario, CCU technologies reduce life cycle
industry [about 1.5 Gt CO2-eq in 2013 (19)] and is in line with GHG emissions compared with conventional technologies once the
previous estimates for 2030 (20). employed electricity has a carbon footprint of 260 g CO2-eq/kWh or
In a second scenario, where the production of the 20 chemicals lower. The life cycle GHG emissions of the chemical industry de-
relies on CO2-based methanol (21) and methane (22), feedstock cline nonlinearly as function of the carbon footprint of electricity
mass flows increase substantially by 287% due to the CO2 used since our optimization approach minimizes GHG emissions and
and water coproduced (Fig. 2B): 3.72 Gt CO2 and 0.59 Gt hy- different CCU technologies start to outcompete the corresponding
drogen from water electrolysis are annually converted into conventional technologies in terms of their life cycle GHG emis-
methanol and methane. Methanol is further processed into sions at different carbon footprints of electricity. Electricity carbon
olefins and BTX, which are the basis of the other large-volume footprints where CCU technologies enter the production mix in the
chemicals. Methane substitutes natural gas to produce both heat model range from 74 g CO2-eq/kWh (mixed xylene) to 260 g CO2-
and ammonia. CO2 can be captured from two different sources eq/kWh (methanol; SI Appendix, Table S1). For all technologies,
representing the upper and lower bound for the climate impact capturing CO2 from the industrial point source leads to higher
of CO2 supply: a highly concentrated industrial point source emission reductions than air capture, because less energy is required
(∼100% CO2) and ambient air with about 400 ppm CO2 (23). for capture (23). GHG emissions are most reduced by using CO2
This scenario leads basically to the methanol economy envi- from the industrial point source and electricity from wind power,
sioned by Asinger (24) and promoted by Olah (25). Since all because wind power is currently the energy generation technology
technologies in this scenario are at technology readiness levels with the lowest carbon footprint. The resulting maximum GHG
(TRLs) of 7 and higher, this scenario is denoted “high-TRL emission reduction is the difference between the GHG emissions of
scenario.” The integration of CO2 via methane and methanol the conventional and of the CCU scenario. The life cycles of con-
would require comparably low research and development efforts ventional and CCU technologies differ only from cradle-to-gate,
and allows use of wide parts of the existing petrochemical in- while emissions during the use phase and end of life, as well as
frastructure and technologies. potential benefits from carbon storage are identical in all scenarios
The third scenario further includes CCU technologies for the since the same chemicals are produced. The maximum GHG
direct conversion of CO2 into olefins (26, 27), BTX (28), carbon emission reductions amount to 3.4 Gt CO2-eq per year and repre-
monoxide (3), ethylene oxide (29), and styrene (30) (SI Ap- sent the technical GHG mitigation potential of the scenario.
pendix, Fig. S1). These technologies are currently at early re- Emissions are reduced by replacing emission-intensive processes
search and development stages with TRLs below 7. Therefore, within the chemical industry and its supply chain, and by capturing

11188 | www.pnas.org/cgi/doi/10.1073/pnas.1821029116 Kätelhön et al.

This content downloaded from

111.68.102.25 on Mon, 07 Oct 2024 03:21:41 UTC
All use subject to https://about.jstor.org/terms
 A conventional scenario
waste & energy recovery

methanol
secondary chemicals:
O2, air, N2, other
vinyl chloride C2-chemicals

ethylene oxide ethylene glycol

ethylene
polyethylene
polyolefins

propylene polypropylene
acrylonitrile C3-chemicals

fossil
propylene oxide acetone & phenol
resources
(oil, gas, coal) cumene
styrene
aromatic chemicals
terephthalic acid
BTX aromatics

ammonia

air and water

carbon dioxide

ENVIRONMENTAL
B high-TRL scenario

SCIENCES
waste water, other wastes &
energy recovery
secondary chemicals:
O2, air, N2, other

SUSTAINABILITY
water

SCIENCE
water
carbon dioxide
water
water

methanol
methanol
C2-chemicals

vinyl chloride, ethylene glycol

ethylene oxide
ethylene polyethylene polyolefins
propylene polypropylene

C3-chemicals
acetone & phenol
acrylonitrile, propylene oxide, cumene
styrene aromatic chemicals
terephthalic acid
BTX aromatics

methane ammonia

hydrogen

carbon dioxide
air and water
methane for heat

Fig. 2. Mass flows within the chemical industry in (A) the conventional scenario and (B) the high-TRL scenario for CCU implementation. In both scenarios, the
projected final demand for 20 large-volume chemicals in 2030 is produced. Note, mass flows are scaled differently in the conventional and the high-TRL
scenario. The mass flows of the final demand (on the Right) are identical.

CO2 outside the boundaries of the chemical industry. The resulting CCU technologies become beneficial for electricity carbon
structure of the chemical industry is presented in Fig. 2B. footprints between 44 g CO2-eq/kWh (carbon monoxide) and
In the low-TRL scenario, moderate climate impact reductions 334 g CO2-eq/kWh (paraxylene; SI Appendix, Table S1). In the
are achieved regardless of the carbon footprint of electricity, low-TRL scenario, the annual emissions are reduced from 4.2 Gt
because this scenario also includes CCU technologies that do not CO2-eq for the conventional scenario to 0.7 Gt CO2-eq leading
require electricity for hydrogen production, i.e., synthesis of to a maximum climate change mitigation potential of 3.5 Gt
ethylene from CO2 and methane, and ethylene oxide from CO2 CO2-eq per year.
and ethylene. However, also in the low-TRL scenario, emissions To contextualize the required carbon footprints of electricity, we
can be substantially reduced only with low-carbon electricity. show the current carbon footprint of grid electricity for different

Kätelhön et al. PNAS | June 4, 2019 | vol. 116 | no. 23 | 11189

This content downloaded from

111.68.102.25 on Mon, 07 Oct 2024 03:21:41 UTC
All use subject to https://about.jstor.org/terms
 4.5 sions until the demands for the chemicals from such technologies
Cradle-to-grave GHG emissions (in Gt CO2-eq/year)

Cradle-to-gate GHG emissions (in Gt CO2-eq/year)

2 are exhausted (Materials and Methods). Additional electricity is
4 thus progressively allocated to CCU technologies reducing GHG
1.5
emissions less efficiently. As a result, the slope of the GHG
3.5 emissions decreases with the amount of additional electricity
available. For carbon-free electricity, efficiencies of CCU tech-
1
3 nologies range from 340 g CO2-eq avoided per kilowatt hour for
the production of paraxylene to 140 g CO2-eq avoided per
0.5
2.5 kilowatt hour for the production of methane. Hence, higher
emission reductions per kilowatt hour can be achieved, if the
0

Wind
2
limited low-carbon electricity is used for the most efficient
CCU technologies.
-0.5
1.5 Discussion
-1 Considering both the large potential of CCU in the chemical
1 industry to mitigate climate change and the large amount of low-
Conventional Switzerland -1.5 carbon electricity needed to exploit this potential, we should ask

Solar PV
Belgium
Austria

France
0.5 High TRL ourselves two questions: (i) Can we provide that much low-carbon
Low TRL electricity, and, if yes, (ii) should we use it in the chemical
-2
0 industry?
500 400 300 200 100 0
Carbon footprint of electricity (in g CO2-eq/kWh) According to a recent Intergovernmental Panel on Climate
Change (IPCC) report (32), the amount of renewable energy
Fig. 3. Cradle-to-grave and cradle-to-gate GHG emissions of the chemical that could be produced by full implementation of demonstrated
industry producing the final demand for 20 large-volume chemicals in 2030 as technologies exceeds even the most ambitious scenarios for re-
function of the carbon footprint of electricity. Cradle-to-grave emissions in- newable energy deployment in 2050 by more than one order of
clude all emissions throughout the life cycles of the 20 large-volume chemicals,
magnitude. Thus, in theory, it is possible to produce sufficient
while cradle-to-gate emissions cover only the production stage and the supply
of all raw materials and energy needed for production. The vertical dashed low-carbon electricity. However, the actual implementation of
lines illustrate the climate impact of grid electricity in selected countries or of renewable energy capacities is limited by practical constraints
electricity from selected renewable energy technologies. including costs, competition for land, public acceptance of en-
ergy infrastructure, and limits in the uptake of intermittent elec-
tricity supply by the electricity grid (33). Despite the remarkable
countries and technologies in Fig. 3. While most countries and
technologies do not provide electricity with a sufficiently low
carbon footprint to realize substantial GHG emission reductions
5
Cradle-to-grave GHG emissions (in Gt CO2-eq/year)

by CCU, there are countries where CCU could reduce GHG

Cradle-to-gate GHG emissions (in Gt CO2-eq/year)

emissions substantially using grid electricity already today. In 2.5
4.5
France, for example, the introduction of high-TRL CCU tech- Conventional 500 2
nologies would roughly halve the climate impact of chemicals 4 300
over the entire life cycle. 200 1.5
Even though individual countries have already demonstrated 3.5
175
that low-carbon electricity can be provided at scale, the trans- 1
3 150
formation of the chemical industry toward CCU would cause a
substantial demand for additional electricity that would require a 125 0.5
2.5
Heat

large-scale expansion of electricity production capacities. This 100 0

expansion would require large shares of renewable (or nuclear) 2
pump

energy to achieve a sufficiently low carbon footprint. 75 -0.5

Sol
If all additional electricity were provided by renewable energy, 1.5 ar P
E-mo

E-

V
vs. n

50
bo

the amount of renewable energy required for the full-scale in-

E-m

-1
ile

Wi
bility

1
rv

troduction of CCU would correspond to 126% and 222% of nd

atura

25
obil

s.

current targets [sustainable development scenario of IEA (31)] -1.5

na
vs. g

ity v

0.5
l gas

tu

for the global renewable electricity production in 2030 for the 0

ra
s. d

lg

g CO2-eq/ -2
a

low-TRL and the high-TRL scenario, respectively. Even these

as
s

0
boile
oline

kWh
iese

bo

current targets are already known not to be achievable based on -2.5

ile
r

existing and announced policies (31). Thus, the need for a fur-
l

-0.5
0 5 10 15 20 25
ther expansion of renewable electricity production capacities is
Additional electricity (in PWh)
likely to be a limiting factor for CCU in the chemical industry.
The climate change mitigation potential of CCU technologies Fig. 4. Cradle-to-grave and cradle-to-gate GHG emissions for the chemical
depends on both the amount of additional electricity available industry in the low-TRL scenario producing the final demand for 20 large-
and the carbon footprint of electricity (Fig. 4). The amount of volume chemicals in 2030 as function of the amount of additional electricity
additional electricity available determines the scale of CCU available and its carbon footprint in grams of CO2 equivalent/kilowatt hour.
implementation and resulting GHG emission reductions. For a Cradle-to-grave emissions include all emissions throughout the life cycles of
the 20 large-volume chemicals, while cradle-to-gate emissions cover only the
given electricity carbon footprint, the climate impact of the
production stage and the supply of all raw materials and energy needed for
chemical industry depends nonlinearly on the amount of elec- production. The black solid line represents the GHG emissions in the con-
tricity available, because individual CCU technologies differ in ventional scenario. The light gray lines illustrate potential GHG emission
their efficiencies in using renewable energy to achieve emission reductions by using the additional electricity for e-mobility to substitute
reductions. Our model allocates additional electricity to the gasoline or diesel cars, or for heat generation in a heat pump or an e-boiler
CCU technologies that save the largest amount of GHG emis- to substitute heat from natural gas boilers.

11190 | www.pnas.org/cgi/doi/10.1073/pnas.1821029116 Kätelhön et al.

This content downloaded from

111.68.102.25 on Mon, 07 Oct 2024 03:21:41 UTC
All use subject to https://about.jstor.org/terms
technological progresses and substantial reduction in cost, modern the world’s most favorable production regions (37). Using
nonhydro renewable power, including wind, solar, biomass, geo- hourly wind and solar geospatial data, Fasihi et al. (38)
thermal, and ocean power, accounts for only 1.7% of the global identified wide areas, where a combination of one-axis sun-
final energy demand (34). Hence, dramatically increasing low- tracking solar photovoltaic capacities and modern wind turbines
carbon electricity to the extent that it can sufficiently power could enable water electrolysis with load factors higher than 50%.
CCU for chemicals in the next decade is likely to be a challenge, Many of these areas are located in Africa, Australia, and South
and renewable energy is expected to remain a limited resource for America, where the amount of available renewable energy re-
the coming decades. sources is more than 50 times higher than the current total primary
In this case, how much of the limited, low-carbon electricity energy demand (32). Even though our study does not quantify
resource should be allocated to CCU technologies compared potential climate benefits from using CO2-based energy vectors
with other potential uses for low-carbon electricity? For this outside the chemical industry, the trade of renewable energy
question, it is useful to differentiate between on- and off-grid use through CO2-based products could be highly beneficial especially
of renewable electricity. For on-grid use, the additional low- for densely populated regions such as Germany, Japan, or Korea
carbon electricity for CCU is distributed via the electricity grid. that have fewer renewable energy resources and may in practice fall
In this case, the electricity could also be used by other GHG short of employing sufficient capacities to satisfy their future re-
mitigation technologies connected to the grid such as electro- newable energy demand (32).
mobility (e-mobility) or heat pumps. From the point of climate The presented analysis focuses on the technical potential for
change mitigation, limited low-carbon electricity should be used climate change mitigation by CCU in the chemical industry in
for those technologies that reduce the climate impact the most 2030, i.e., the maximum reduction of climate impacts that is
(5, 35). In Fig. 4, as an example, we show the climate impact technically feasible in case of full deployment of CCU (Fig. 2).
reduction by substituting natural gas boilers with either heat The actual deployment rate, however, is likely to be lower. First,
pumps or electric boilers, as well as substituting diesel or gaso- achieving full deployment of novel CCU technologies by
line cars through e-mobility. Power-to-heat and power-to-e- 2030 would imply an average annual increase in market pene-

ENVIRONMENTAL
mobility are both very large-scale uses of renewable electricity tration of 10% between 2020 and 2030. This increase in market

SCIENCES
and would thus compete with CCU for electricity. The climate penetration substantially exceeds current replacement rates of
impact reductions have been calculated based on power-to-X chemical production plants: Assuming an average plant life time
efficiencies by Sternberg and Bardow (5) for the use of of 25 y (39), only 4% of existing production capacities are
emission-free electricity from surplus power. The power-to-X replaced annually. Consequently, even though parts of the
efficiency represents the climate impact reductions per electric- existing chemical infrastructure could still be used, a rapid
ity used (details in Materials and Methods). In comparison with implementation of CCU technologies would require substantially
all hydrogen-based CCU technologies considered in this work, e- higher investment volumes, as well as the phase out of existing

SUSTAINABILITY
mobility and heat pumps reduce the climate impact more production plants before the end of their technically feasible

SCIENCE
strongly per kilowatt hour of electricity used. Hence, from a lifetime. Second, the full implementation of CCU technologies
climate perspective, the implementation of these technologies can only be achieved with a massive expansion of low-carbon
should be prioritized over the hydrogen-based CCU technologies electricity production capacities that substantially exceeds cur-
considered in this work until their demand for renewable energy rent targets for low-carbon electricity production for 2030 (31).
is fully exhausted. However, hydrogen-based CCU implementa- This expansion is likely to represent a bottleneck for large-scale
tion in the chemical industry still provides a later option for hydrogen-based CCU implementation within the next decades.
climate change mitigation, since large-scale emission reductions Furthermore, the expansion will be based on a mix of electricity
will eventually be needed in all relevant sectors (19). Similarly, generation technologies leading to a higher carbon footprint
even the installation of electric boilers to substitute natural gas than wind power assumed for the technical potential in this
boilers reduces the climate impact more than all hydrogen-based study. Third, in addition to CCU, other measures have been
CCU technologies except the CO2-based production of para- proposed for reducing climate impacts in the chemical industry
xylene and styrene. Still, climate benefits are achieved by CCU including the use of biomass feedstock (4), electrification (5),
technologies that do not require electricity for hydrogen pro- and carbon capture and storage (6). Thus, climate impacts of the
duction such as the production of ethylene and ethylene oxide. chemical industry are likely to be reduced by a mix of measures.
However, the perspective changes in remote areas without The technical climate change mitigation potential of CCU is
grid connection but with large unused renewable energy re- affected by the CO2 source. Capturing CO2 from highly con-
sources. Here, off-grid CCU technologies could provide climate centrated CO2 sources such as industrial point sources reduces
benefits without competing with other technologies for renew- emissions more, because less energy is required for capture than
able energy. Even more, by transforming regionally bound re- for dilute sources such as air. The emissions reductions capture
newable energy into transportable commodities, CCU technologies how much emissions are changing as consequence of introducing
could enable global trade of renewable energy, for example, in the CCU technologies. However, for the goal of an overall zero-
form of CO2-based methanol or methane. Also, ammonia could be emission society, the industrial point source itself would have
used as renewable fuel (36, 37). However, ammonia is produced to be de-fossilized or end-of-life emissions from the CCU
more efficiently by the direct conversion of renewable hydrogen product would need to be avoided, e.g., by permanent carbon
with nitrogen than by CCU (SI Appendix). Setting up production storage. In contrast, direct air capture and biogenic point sources
plants in remote areas will result in additional costs for in- (e.g., biobased ethylene production) would allow for a closed
frastructure, feedstock supply (e.g., CO2 and water), and the carbon cycle even if the CCU product is incinerated. Switching
transportation of the CO2-based product (e.g., methanol) to toward lower concentrated sources will increase costs due to
established chemical production sites for further processing. higher energy demands for CO2 capture. In a net-zero emission
For the CO2-based production of methanol in remote areas, for economy, CCU could then provide the carbon feedstock needed
example, additional transportation costs for the CO2 supply for the production of a wide range of products (e.g., chemicals,
from industrial point sources and the transportation of meth- plastics, and pharmaceuticals).
anol have been estimated to about 100 USD/t, i.e., about 25% While a detailed economic assessment of CCU in the chemical
of current methanol production costs based on fossil resources industry is currently not possible due to the early development
(37). These additional costs, however, will partly be offset by stage of most CCU technologies, the magnitude of additional
lower renewable electricity prices and higher load factors in operational cost can be estimated based on major cost factors

Kätelhön et al. PNAS | June 4, 2019 | vol. 116 | no. 23 | 11191

This content downloaded from

111.68.102.25 on Mon, 07 Oct 2024 03:21:41 UTC
All use subject to https://about.jstor.org/terms
 Table 1. Technologies for CCU implementation in the high- and low-TRL scenario
Chemical High-TRL scenario Low-TRL scenario

Benzene Methanol-to-aromatics (42) 9 CO2 + 27 H2 → C6H6 + 18 H2O + 3 CH4 (28)

Carbon dioxide Capture from highly concentrated Capture from highly concentrated
point source or ambient air (23) point source or ambient air (23)
Carbon monoxide CO2 + H2 → CO + H2O (3, 49)
Ethylene Methanol-to-olefins (42) 2 CO2 + 2 CH4 → C2H4 + 2 CO + 2 H2O (26)
2 CO2 + 6 H2 → C2H4 + 4 H2O (27, 54)
Ethylene oxide C2H4 + CO2 → C2H4O + CO (29)
Hydrogen From water electrolysis (50) From water electrolysis (50)
Methane From CO2 (Sabatier reaction) (51, 52) From CO2 (Sabatier reaction) (51, 52)
Methanol From CO2 and H2 (direct hydrogenation) (53) From CO2 and H2 (direct hydrogenation) (53)
Propylene Methanol-to-olefins (42) 3 CO2 + 9 H2 → C3H6 + 6 H2O (27, 54)
Styrene C8H10 + CO2 → C8H8 + CO + H2O (30)
Toluene Methanol-to-aromatics (55) 9 CO2 + 26 H2 → C7H8 + 18 H2O + 2 CH4 (28)
Xylene, mixed Methanol-to-aromatics (55) 9 CO2 + 25 H2 → C8H10 + 18 H2O + CH4 (28)
Xylene, para Methanol-to-aromatics (55) 9 CO2 + 25 H2 → C8H10 + 18 H2O + CH4 (28)

Low-TRL technologies are represented by the chemical equations of the underlying reactions.

including the avoided annual consumption of crude oil (4.21 should be interpreted with caution. GHG emission reductions due
billion barrel) and natural gas (6.59 PWh) and the production of to substitution have been demonstrated for individual CCU tech-
hydrogen from low-carbon electricity in the CCU scenarios (0.38 nologies (3), but not for a large-scale CCU implementation in the
Gt and 0.59 Gt in the low- and high-TRL scenario). With global chemical industry. Finally, our analysis highlights the need to
expected prices in 2030 of 92.3 USD per barrel oil [current policy determine not only emission reductions compared with conven-
scenario (31)], 0.017 USD per kWh natural gas (40), and 1.8–3.5 tional technologies, but to also compare climate change mitigation
USD per kg hydrogen (41) from renewable energy in most measures regarding their power-to-X efficiency.
beneficial production regions (lower limit) and regions with good The presented analysis focuses on the climate change mitiga-
but not excellent solar and wind resources (upper limit; Europe tion potential of CCU and neglects CCU services beyond climate
or similar), the additional operating cost for the full imple- change mitigation (35). For example, CO2-based fuels could
mentation of CCU technologies in the low- and high-TRL sce- reduce the import dependency of countries that do not produce
nario amount to 185–833 and 564–1,570 billion USD annually, fossil fuels by enabling a diversification of energy supply, and
respectively. These additional costs correspond to 19–87% and hence increase energy security. Furthermore, CCU products can
59–164% of the market value of the produced chemicals reduce other environmental impacts, e.g., NOx and particular
according to 2017 prices (42). The corresponding climate change matter emissions from fuels (45). Such benefits need to be
mitigation costs range between 52–235 USD per t CO2-eq and considered to comprehensively assess the potential of CCU. For
168–467 USD per t CO2-eq in the low- and high-TRL scenario. climate change mitigation, our results provide a basis for the
For the low-TRL scenario, the lower bound of the climate sound integration of CCU into research agendas, technology
change mitigation costs from this simplified calculation are thus roadmaps, and future policy frameworks.
in the range of expected carbon prices for the year 2030 (19).
This estimate, however, neglects all required investment cost Materials and Methods
within the chemical industry except for hydrogen production by Scope of the Model and Key Assumptions. The bottom-up model of the
water electrolysis. In the CCU scenarios, investment costs are chemistry industry yields future production pathways for 20 major bulk
chemicals: acrylonitrile, ammonia, benzene, caprolactam, cumene, ethylene,
likely to increase, because larger mass flows need to be moved:
ethylene glycol, ethylene oxide, methanol, mixed xylenes, phenol, poly-
The mass flow of hydrogen and CO2 needed is 1.8 and 2.9 times ethylene, polypropylene, propylene, propylene oxide, p-xylene, styrene,
larger than the mass flow of fossil inputs in the low- and high- terephthalic acid, toluene, and vinyl chloride. Production pathways are
TRL scenario, respectively. The larger mass flows will lead to represented by more than 160 processes based on engineering-level data.
larger production facilities. The increase in investment costs will Thereby, flows of energy and materials are determined in detail throughout
add to the CO2 mitigation costs of CCU. entire supply chains. The key assumptions of the model are illustrated below,
Previous studies have discussed the potential role of CCU for while the computational structure is shown in the following section.
climate change mitigation based on the amount of CO2 utilized Technology choices. Technologies are chosen to minimize GHG emissions,
(15, 43). The High Level Group of Scientific Advisors to the enabling the determination of the theoretical maximum climate change
mitigation potential by CCU compared with a non-CCU scenario. For each
European Commission, for example, recently reported an esti-
chemical, both conventional and CO2-based production technologies are
mated long-term utilization potential of 1–2 Gt per year (15). considered. For the conventional production pathways, only the best-
Alternatively, the amount of CO2 stored in chemicals has been available technologies (BATs) from a climate perspective are included, i.e.,
used as simplified indicator for the climate change mitigation the industrialized technologies with the lowest climate impact per kilogram
potential of CCU (13). In contrast, the modeling approach product. The focus on BATs implies a conservative estimate of the climate
proposed in this study provides a detailed assessment of GHG change mitigation potential of CCU technologies.
emission reductions due to disruptive changes through large- CO2 supply. For the CO2 supply, we consider two potential sources repre-
scale CCU implementation. The lack of models for assessing senting the upper and lower bound for the climate impact of CO2 supply: a
highly concentrated industrial point source (∼100% CO2) and ambient air
disruptive changes has recently been identified as priority re-
with about 400 ppm CO2 (23).
search need for LCA of CCU (44). Our results show that the Use phase and end of life. Emissions during the use phase are neglected. As end-
climate change mitigation potential of CCU in the chemical in- of-life scenario, we assume the incineration of all chemicals in a waste in-
dustry is neither determined by the amount of CO2 used, nor stored cinerator, leading to 2.2 Gt CO2-eq emissions. The assumption of waste in-
in chemicals, but by the substitution of conventional technologies. cineration is a worst-case assumption for end-of-life emissions. Assuming
Therefore, estimates based on the amount of carbon used or stored waste incineration neglects that parts of the chemicals may also be stored in

11192 | www.pnas.org/cgi/doi/10.1073/pnas.1821029116 Kätelhön et al.

This content downloaded from

111.68.102.25 on Mon, 07 Oct 2024 03:21:41 UTC
All use subject to https://about.jstor.org/terms
 Table 2. Key assumptions and parameters of the calculations of the power-to-X efficiencies of
e-mobility and power-to-heat by Sternberg and Bardow (5)
E-mobility Value Ref.

Battery electric vehicle with lithium ion battery

Battery efficiency 80%
Electricity consumption of battery electric vehicle 0.14–0.2 kWh·km−1 56, 57
Fuel consumption of gasoline car 0.52 kWh·km−1 58
Fuel consumption of diesel car 0.38 kWh·km−1 58
Power-to-heat
Electricity consumption of heat pump 2.1–4.9 kWh·kWhheat−1 59
Electricity consumption of electric boiler 1 kWh·kWhheat−1
Natural gas demand of natural gas boiler 1.1 kWh·kWhheat−1 58

materials for substantially long times. Also, the waste incineration plant impact can be expressed as CO2-equivalents of GHG emissions according to
could add facilities for CO2 capture and sequestration. However, both the the following equation:
use phase and the end of life of chemicals are assumed to be identical,
because all CCU technologies considered produce the identical chemicals as h = QBs. [2]
conventional production. Therefore, potential benefits from temporary
carbon storage or sequestration will be the same in the conventional and Based on these definitions, the linear program of the bottom-up model of the
the CCU scenarios, and not affect the difference in GHG emissions, i.e., the chemical industry is defined by the following equations (Eqs. 3–6):
climate change mitigation potential.
Min h = QBs, [3]
CCU scenarios. The CCU scenarios include all technologies from the conven-

ENVIRONMENTAL
tional scenario and technologies for CCU implementation. The technologies
s. t. As = y, [4]

SCIENCES
for CCU implementation are illustrated in Table 1. The methodology for data
collection and a list of all processes and data sources is provided in SI Ap- s ≤ c, [5]
pendix. In all scenarios, processes for the disposal of wastes and for energy
integration and recovery are included. sj ≥ 0, [6]

Computational Structure of the Bottom-Up Model. The bottom-up model of where h represents the climate impact of the production system, A specifies
the chemical industry builds on the technology choice model (TCM) (46). the technology matrix, y is the final demand, and c is potential upper bounds
Following the general definitions of LCA (47), the technical production sys- for scaling vector entries. An upper bound has been set for the amount of

SUSTAINABILITY
tem is described by four basic entities: technologies, intermediate flows, electricity available in the calculations for Fig. 4. Eq. 3 expresses the climate

SCIENCE
elementary flows, and the final demand. Technologies are defined as ac- impact minimization objective under the constraints that the final demand is
tivities transforming inputs into outputs. Input and output flows of energy, produced (Eq. 4) and that the scaling vector entries are between zero and
materials, or products are categorized into intermediate flows and ele- the upper bound (Eqs. 5 and 6).
mentary flows. Intermediate flows are flows between technologies, such as
intermediate products or raw materials. Elementary flows are exchanged Determination of Power-to-X Efficiencies of E-Mobility and Power-to-Heat. The
between technologies and the environment, such as emissions of GHGs or climate impact reductions for the benchmark power-to-X technologies are
resources extracted from the environment. The output of products for final determined based on power-to-X efficiencies by Sternberg and Bardow (5).
consumption is specified by the final demand. The power-to-X efficiencies are defined as climate impact reductions per
The data on the physical production system are structured according to the
megawatt hour of electricity used and represent the slopes of the gray
generalized calculus for LCA (47). The transformation of intermediate flows
lines in Fig. 4. In Sternberg and Bardow (34), the power-to-X efficiencies
by technologies is described by the technology matrix A. In this matrix A,
GWreduction are determined based on the following equation:
columns represent production technologies, while rows indicate intermediate
flows. A coefficient aij of the technology matrix A shows the intermediate flow GWreduction = Mproduct GWconv − GWESS . [7]
i that is produced (for aij > 0) or consumed (for aij < 0) by technology
j. Analogously, the consumption and emission of elementary flows by tech- The power-to-X technology uses 1 MWh of electricity to produce the amount
nologies is represented by the elementary flow matrix B. In this matrix, a co- Mproduct of its product. Thereby, the conventional production of this amount
efficient bej describes the elementary flow e that enters (for bej < 0) or leaves of product is avoided, reducing global warming impacts by (Mproduct
(for bej > 0) technology j. The final demand is represented by the final demand GWconv), where GWconv is the global warming impact of conventional pro-
vector y, where an element yi describes the final demand for intermediate duction of one product unit. GWESS represents the global warming impact of
flow i. the power-to-X technology using 1 MWh of electricity. For both power-to-X
For the production of various chemicals in the model, more than one and substituted technologies, emissions due to operation and construction
production technology is included. Choices between technologies are de- are considered. Key assumptions and parameters of the calculations of the
termined by the minimization of the climate impact of the entire production power-to-X efficiencies of e-mobility and power-to-heat are illustrated in
system. To quantify the climate impact, a scaling vector s is defined, which
Table 2.
scales the amount of inputs and outputs of each technology, thus describing
the production level. Based on this scaling vector, the cumulated elementary
ACKNOWLEDGMENTS. We thank Patrick Eichstädt, Valentin Dinges, and
flows g of the production system that are exchanged with the environment
Lucia Baltz for the technical support. This work has been carried out in part
throughout the entire life cycle of chemicals are determined by the fol- within the project “enCO2re: enabling CO2 re-use,” which is funded by the
lowing: Climate Knowledge and Innovation Community by the European Institute of
Innovation and Technology. A.B. and A.K. gratefully acknowledge support
g = Bs. [1] by the German Academic Exchange Service through its thematic network
“ACalNet” funded by the German Federal Ministry of Education and Re-
Following leading guidelines for life cycle impact assessment (48), these search. S.S. was supported by the US Environmental Protection Agency
cumulated elementary flows g are further characterized regarding their (EPA) through Science to Achieve Results Program Grant 83557907. Any
contribution to climate change using the characterization vector Q. An el- opinions, findings, and conclusions or recommendations expressed in this
ement qe of Q represents the 100-y global warming potential of elementary material are those of the authors and do not necessarily reflect the views
flow e according to IPCC (19), illustrating the relative contribution of this of US EPA. This work has not been subjected to US EPA review, and no
flow to climate change over a 100-y time horizon. By this means, the climate official endorsement should be inferred.

Kätelhön et al. PNAS | June 4, 2019 | vol. 116 | no. 23 | 11193

This content downloaded from

111.68.102.25 on Mon, 07 Oct 2024 03:21:41 UTC
All use subject to https://about.jstor.org/terms
 1. International Energy Agency (2017) Tracking Clean Energy Progress 2017 (IEA Publications, 30. Wang Q, Li X, Li W, Feng J (2014) Promoting effect of Fe in oxidative de-
Paris). hydrogenation of ethylbenzene to styrene with CO2 (I) preparation and performance
2. International Energy Agency (2018) The Future of Petrochemicals: Towards More of Ce1−xFexO2 catalyst. Catal Commun 50:21–24.
Sustainable Plastics and Fertilisers (IEA Publications, Paris). 31. International Energy Agency (2017) World Energy Outlook 2017 (IEA Publications, Paris).
3. Artz J, et al. (2018) Sustainable conversion of carbon dioxide: An integrated review of 32. Edenhofer O, et al., eds (2012) Renewable Energy Sources and Climate Change Mit-
catalysis and life cycle assessment. Chem Rev 118:434–504. igation: Special Report of the Intergovernmental Panel on Climate Change (Cam-
4. Chen G-Q, Patel MK (2012) Plastics derived from biological sources: Present and fu- bridge Univ Press, Cambridge, UK).
ture: A technical and environmental review. Chem Rev 112:2082–2099. 33. International Energy Agency (2017) Getting Wind and Sun onto the Grid: A Manual
5. Sternberg A, Bardow A (2015) Power-to-what?—environmental assessment of energy for Policy Makers (IEA Publications, Paris).
storage systems. Energy Environ Sci 8:389–400. 34. REN21 (2018) Renewables 2018 Global Status Report (REN21 Secretariat, Paris).
6. Bui M, et al. (2018) Carbon capture and storage (CCS): The way forward. Energy 35. Schakel W, Fernández-Dacosta C, van der Spek M, Ramírez A (2017) New indicator for
Environ Sci 3:1645. comparing the energy performance of CO2 utilization concepts. J CO2 Util 22:278–288.
7. Klankermayer J, Leitner W (2015) CHEMISTRY. Love at second sight for CO2 and H2 in 36. Grinberg Dana A, Elishav O, Bardow A, Shter GE, Grader GS (2016) Nitrogen-based
organic synthesis. Science 350:629–630. fuels: A power-to-fuel-to-power analysis. Angew Chem Int Ed Engl 55:8798–8805.
8. Cuéllar-Franca RM, Azapagic A (2015) Carbon capture, storage and utilisation tech- 37. Cédric Philibert (2017) Renewable Energy for Industry: From Green Energy to Green
nologies: A critical analysis and comparison of their life cycle environmental impacts. J Materials and Fuels (IEA Publications, Paris).
CO2 Util 9:82–102. 38. Fasihi M, Bogdanov D, Breyer C (2016) Techno-economic assessment of power-to-liquids
9. Al-Mamoori A, Krishnamurthy A, Rownaghi AA, Rezaei F (2017) Carbon capture and (PtL) fuels production and global trading based on hybrid PV-wind power plants. Energy
utilization update. Energy Technol (Weinheim) 5:834–849. Procedia 99:243–268.
10. Kondratenko EV, Mul G, Baltrusaitis J, Larrazábal GO, Pérez-Ramírez J (2013) Status 39. Ecoinvent Association (2017) Ecoinvent, Version 3.4 (Ecoinvent, Zurich). Available at
and perspectives of CO2 conversion into fuels and chemicals by catalytic, photo- www.ecoinvent.org/. Accessed October 26, 2018.
catalytic and electrocatalytic processes. Energy Environ Sci 6:3112. 40. Statistica (2018) U.S. and European natural gas price 1980–2030. Available at https://
11. Centi G, Quadrelli EA, Perathoner S (2013) Catalysis for CO2 conversion: A key tech- www.statista.com/statistics/252791/natural-gas-prices. Accessed June 6, 2018.
nology for rapid introduction of renewable energy in the value chain of chemical 41. Cédric Philibert (2017) Producing Ammonia and Fertilizers: New Opportunities from
industries. Energy Environ Sci 6:1711. Renewables (International Energy Agency, Paris).
12. Ampelli C, Perathoner S, Centi G (2015) CO2 utilization: An enabling element to move 42. Markit IHS (2018) Process Economics Program (PEP) Yearbook (IHS Markit, London).
to a resource- and energy-efficient chemical and fuel production. Philos Trans A Math 43. Aresta M, Dibenedetto A, Angelini A (2013) The changing paradigm in CO2 utiliza-
Phys Eng Sci 373:20140177.
tion. J CO2 Util 3-4:65–73.
13. Mac Dowell N, Fennell PS, Shah N, Maitland GC (2017) The role of CO2 capture and
44. National Academies of Sciences, Engineering, and Medicine (2019) Gaseous Carbon
utilization in mitigating climate change. Nat Clim Change 7:243–249.
Waste Streams Utilization (National Academies Press, Washington, DC).
14. European Commission (2018) Pathways to Sustainable Industries: Energy Efficiency
45. Deutz S, et al. (2018) Cleaner production of cleaner fuels: Wind-to-wheel—Environ-
and CO2 Utilisation (Publication Office of the European Union, Luxemburg).
mental assessment of CO2-based oxymethylene ether as a drop-in fuel. Energy Envi-
15. High Level Group of Scientific Advisors to the European Commission (2018) Novel
ron Sci 11:331–343.
Carbon Capture and Utilization Technologies (Publication Office of the European
46. Kätelhön A, Bardow A, Suh S (2016) Stochastic technology choice model for conse-
Union, Luxemburg).
quential life cycle assessment. Environ Sci Technol 50:12575–12583.
16. von der Assen N, Bardow A (2014) Life cycle assessment of polyols for polyurethane
47. Heijungs R, Suh S (2002) The Computational Structure of Life Cycle Assessment
production using CO2 as feedstock: Insights from an industrial case study. Green Chem
(Springer Netherlands, Dordrecht, The Netherlands).
16:3272–3280.
48. European Commission, Joint Research Centre (2011) International Reference Life Cycle
17. von der Assen N, Voll P, Peters M, Bardow A (2014) Life cycle assessment of CO2
Data System (ILCD) Handbook-Recommendations for Life Cycle Impact Assessment in
capture and utilization: A tutorial review. Chem Soc Rev 43:7982–7994.
the European Context (Publication Office of the European Union, Luxemburg)
18. Hellweg S, Milà i Canals L (2014) Emerging approaches, challenges and opportunities
49. Aresta M (2010) Carbon Dioxide as Chemical Feedstock (Wiley-VCH, Weinheim, Germany).
in life cycle assessment. Science 344:1109–1113.
50. U.S. Department of Energy (2011) Fuel Cell Technologies Program Multi-Year Re-
19. Edenhofer O, et al., eds (2014) Climate Change 2014: Mitigation of Climate Change:
search, Development, and Demonstration Plan (US Department of Energy, Wash-
Working Group III Contribution to the IPCC Fifth Assessment Report (Cambridge Univ
ington, DC).
Press, Cambridge, UK).
51. Müller B, Müller K, Teichmann D, Arlt W (2011) Energiespeicherung mittels Methan
20. International Energy Agency/DECHEMA/ICCA (2013) Technology Roadmap. Energy and
GHG Reductions in the Chemical Industry via Catalytic Processes (IEA Publications, Paris). und energietragenden Stoffen–Ein thermodynamischer Vergleich. Chemieingenieurtechnik
21. Rihko-Struckmann LK, Peschel A, Hanke-Rauschenbach R, Sundmacher K (2010) As- (Weinh) 83:2002–2013 (German).
sessment of methanol synthesis utilizing exhaust CO2 for chemical storage of elec- 52. Saint Jean Md, Baurens P, Bouallou C (2014) Parametric study of an efficient re-
trical energy. Ind Eng Chem Res 49:11073–11078. newable power-to-substitute-natural-gas process including high-temperature steam
22. Rönsch S, et al. (2016) Review on methanation—from fundamentals to current pro- electrolysis. Int J Hydrogen Energy 39:17024–17039.
jects. Fuel 166:276–296. 53. Pérez-Fortes M, Tzimas E (2016) Techno-Economic and Environmental Evaluation of
23. von der Assen N, Müller LJ, Steingrube A, Voll P, Bardow A (2016) Selecting CO2 CO2 Utilisation for Fuel Production (Joint Research Centre, European Commission,
sources for CO2 utilization by environmental-merit-order curves. Environ Sci Technol Luxembourg).
50:1093–1101. 54. Yang H, et al. (2017) A review of the catalytic hydrogenation of carbon dioxide into
24. Asinger F (1986) Methanol–Chemie- und Energierohstoff: Die Mobilisation der Kohle value-added hydrocarbons. Catal Sci Technol 7:4580–4598.
(Springer, Berlin). German. 55. Bazzanella AM, Ausfelder F (2017) Low Carbon Energy and Feedstock for the European
25. Olah GA, Goeppert A, Prakash GKS (2011) Beyond Oil and Gas: The Methanol Econ- Chemical Industry (DECHEMA, Frankfurt).
omy (Wiley-VCH, Weinheim, Germany). 56. Majeau-Bettez G, Hawkins TR, Strømman AH (2011) Life cycle environmental assess-
26. Yabe T, Kamite Y, Sugiura K, Ogo S, Sekine Y (2017) Low-temperature oxidative ment of lithium-ion and nickel metal hydride batteries for plug-in hybrid and battery
coupling of methane in an electric field using carbon dioxide over Ca-doped LaAlO3 electric vehicles. Environ Sci Technol 45:4548–4554.
perovskite oxide catalysts. J CO2 Util 20:156–162. 57. Metz M, Doetsch C (2012) Electric vehicles as flexible loads—a simulation approach
27. Gao J, Jia C, Liu B (2017) Direct and selective hydrogenation of CO2 to ethylene and using empirical mobility data. Energy 48:369–374.
propene by bifunctional catalysts. Catal Sci Technol 7:5602–5607. 58. Thinkstep (2018) GaBi LCA Databases (Thinkstep, Leinfelden-Echterdingen, Germany).
28. Wang H, et al. (2014) Carbon dioxide hydrogenation to aromatic hydrocarbons by Available at www.gabi-software.com/international/databases/gabi-databases. Accessed
using an iron/iron oxide nanocatalyst. Beilstein J Nanotechnol 5:760–769. October 26, 2018.
29. Mobley PD, et al. (2017) Utilization of CO2 for ethylene oxide. Energy Procedia 114: 59. Braungart S, Günther D, Miara M, Wapler J, Weißing W (2013) Electrically Driven Heat
7154–7161. Pumps (FIZ Karlsruhe, GmbH, Karlsruhe, Germany).

11194 | www.pnas.org/cgi/doi/10.1073/pnas.1821029116 Kätelhön et al.

This content downloaded from

111.68.102.25 on Mon, 07 Oct 2024 03:21:41 UTC
All use subject to https://about.jstor.org/terms