Chemical Engineering Research and Design 182 (2022) 194–206

Available online at [Link]

Chemical Engineering Research and Design

journal homepage: [Link]/locate/cherd

Electrochemical conversion of carbon dioxide to

ethylene: Plant design, evaluation and prospects
]]
]]]]]]
]]

for the future

Lotte Berkelaar, Joram van der Linde, Julia Peper, Aditya Rajhans,
⁎
Daniël Tiemessen, Louis van der Ham , Henk van den Berg
Department of Science and Technology, University of Twente, Enschede, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: In support of the energy and feedstock transition, new concepts of producing chemical
Received 28 November 2021 building blocks are necessary. While currently ethylene is mainly produced from fossil
Received in revised form 10 February feedstocks, this paper explores its production via the electrochemical conversion of CO2.
2022 Different process routes were reviewed and a most promising route was selected based on
Accepted 21 March 2022 Faradaic efficiency, current density, overpotential, and material availability. In this pre­
Available online 25 March 2022 ferred route, syngas is produced electrochemically from CO2, which is then converted into
methanol before the final conversion into ethylene and other high-value chemicals. On
Keywords: this preferred route, a design has been made which includes process design, unit design
Ethylene production and a techno-economic evaluation, with CAPEX and OPEX included. Also the CO2 emission
Electrochemistry is evaluated and it was found that a net consumption of 2.5 ton CO2 per ton high-value
Syngas chemical can be achieved. As expected, the investment and operational costs are very
Carbon utilization high with the applied price sets. However, when the electricity price decreases due to
developments in renewable energy and a tax is implemented on CO2 emissions, this
process can become economically viable and can contribute to CO2 emission reduction in
view of climate change.
© 2022 The Author(s). Published by Elsevier Ltd on behalf of Institution of Chemical
Engineers. This is an open access article under the CC BY license ([Link]
[Link]/licenses/by/4.0/).

technologies to bridge this gap. In this work, a further step is

1. Introduction
taken to not only reduce CO2 emissions, but to also discard
fossil fuel use to actively reduce CO2 levels in the atmo­
Global energy consumption has been rising steadily over the
sphere.
years. The International Energy Agency puts this increase at
One of the ways of reducing CO2 emissions is by reusing it
2.3% for 2018 - the biggest rise of the decade (IES, 2019). This
as a building block for chemicals with industrial value, for­
rise, mostly fueled by the increased use of fossil fuels, re­
mally known as Carbon Capture and Utilization (CCU)
sulted in an annual 1.7% increase in CO2 emissions (IES, 2019)
(Carbon Capture, 2020). Currently employed methods of up­
CO2 is a well-known contributor to the greenhouse effect and
grading CO2 require the use of fossil-based energy. This is
reducing its emissions is now more important than ever. In
counter-productive (Pappijn et al., 2020).
response, the European Union (EU) aims to reduce 80% of CO2
In this work, based on literary sources, a design is for­
emissions by 2050 (COM, 2011). The current mismatch be­
mulated for the transformation of CO2 extracted from a di­
tween production and use of CO2 has prompted both re­
rect emission source towards a valuable product, by
searchers and companies to look into new and alternative
employing electricity from a sustainable source. At the heart
of the process, an electrochemical conversion reactor is lo­
⁎ cated.
Corresponding author.
E-mail address: [Link]@[Link] (L.v.d. Ham).
[Link]
0263-8762/© 2022 The Author(s). Published by Elsevier Ltd on behalf of Institution of Chemical Engineers. This is an open access article
under the CC BY license ([Link]
 Chemical Engineering Research and Design 182 (2022) 194–206 195

pathways were reviewed. An extensive literature search was

Nomenclature
conducted on electrochemical reactor set-ups for these
A Area [m2] conversion steps and their viability was judged based on
ar Specific reactor area [m2 m−3] corresponding experimental results. Subsequently, basic
FEi Faradaic Efficiency of compound i [-] process designs were made for each of the considered elec­
F Faradaic constant [C mol−1] trochemical conversion steps and the most promising route
I Current [A] was selected, based on various ranking criteria. For the route
j Current density [A m−2] that was judged to be most promising, a detailed process
m Mass [kg] design was made, followed by a techno-economic evaluation
Mi Molecular weight of species i [kg mol−1] to judge the viability of this process as an alternative to
mi Mass of species i [kg] conventional production routes of ethylene, such as naphtha
Ni Molar flux [mol m−2 s−1] and ethane steam cracking (Ghanta, 2014), and other new
ni Number of electrons of species i [-] sources (McKinsey & Company, 2012).
Q Electrical charge [C] Basic process design and the following detailed design in
V Potential [V] this work were performed following a systematic metho­
Y Space time yield [kg m−3 s−1] dology, based on the Douglas Method (Douglas, 1988). Basic
zi Electrical charge of species i [-] designs covered gray box analysis with only conversions and
selectivities of reaction steps taken into account. The final
design consisted of a detailed process design, with optimi­
As a building block of polyethylene, one of the world’s zation of separation steps and analysis of possibilities for
most-produced chemicals (Patel, 2016), this design for a heat integration. Required capital expenditures (CAPEX) of
sustainable ethylene production facility taps into a market the proposed process were estimated based on bare equip­
dominated by energy-intensive and hence often CO2 in­ ment costs of the main process units. Operational ex­
tensive production methods. Besides, ethylene is used to penditures (OPEX) were estimated based on CAPEX and
form other bulk chemicals including ethylene oxide, ethanol, required utilities. Finally, the main technical aspects of the
vinyl acetate and various (co-)polymers. The reduction of designed process were discussed and the proposed design
emissions from the use and production of the world’s most was compared with alternative production routes based on
produced petrochemical, ethylene, is vital because its sheer costs and sustainability.
production volume of over two hundred million metric The direct result of this research was an extensive in­
tonnes per year causes a notable contribution to the green­ ternal report of over 150 pages, describing the full design
house effect. An overview of important parameters that are methodology and all assumptions, decisions, and calculation
used is given in Table 1. The plant capacity is chosen to be results. This report can be requested from the corresponding
smaller than a full-scale plant to approach a demo-scale author. The research objective of this paper is to highlight
plant. When this does not work, scaling-up would not make the main limitations and opportunities for using electro­
sense either. Other values are standards or found in litera­ chemistry for producing sustainable ethylene. Furthermore,
ture. The feasibility of the suggested design is reviewed later the paper focuses on the design methodology for such a
in this paper. process, starting from solely an electrochemical reduction
Carbon dioxide (CO2) can be converted electrochemically reaction.
into ethylene. A variety of reduction reactions of CO2 exist
that could form intermediate compounds for the production
of ethylene (Zhang et al., 2017). Besides direct electro­ 2. Chemistry
chemical conversion of CO2 into ethylene, electrochemical
reduction of CO2 to CO or to methanol was also considered as The overall chemical reaction in the process is shown in Eq.
possible viable alternatives for the electrochemical step of 1. Eqs. 2, 3, and 4 show the reduction reactions of CO2 to­
upgrading CO2 towards ethylene. These three reaction wards ethylene or the considered intermediates.

Table 1 – Overview of important process parameters.

Parameter Value Unit Source

Plant capacity 100 kta –

Product purity > 99 wt% –
Operational hours 8000 h/year –
Electricity price 45 $/MWh Energiemarktinformatie (2021)
Natural gas price 9 $/MWh Administration, U.E.I. Natural Gas, 2020
Economic values
ISBL factor 3.3 – (Towler and Sinnott (2013)
CEPCI (2018) 610 – Chemengonline (2020)
Total fixed capital cost factor 1.82 – (Towler and Sinnott (2013)
Depreciation period 10 year –
Component prices
CO2 80 $/t Irlam (2017)
H2O 1 $/t Bhattacharyya et al. (2012)
C2H4 1050 $/t Platts Global Ethylene Price Index (2020)
C3H6 900 $/t Fibre2fashion (2020)
O2 14 $/t Intratec (2007); Ycharts (2019)
196 Chemical Engineering Research and Design 182 (2022) 194–206

2CO2 (g) + 2H2 O(l ) C2 H4 (g) + 3O2 (g) (1) process was analyzed and all design activities were mapped.
Then, the design process started with identifying a func­
2CO2 (g) + 12H+ + 12e C2 H 4 (g) + 4H2 O(l ) (2) tional design in which raw materials are converted to the
desired product. This resulted in the formulation of five
CO2 (g) + 2H+ + 2e CO(g) + H2 O(l) (3) process alternatives, each a combination of new and con­
ventional technologies. The least attractive process alter­
CO2 (g) + 6H+ + 6e CH3 OH(l) + H2 O(l) (4) native was rejected based on efficiency, selectivity, electricity
The applied prices of all components are given in Table 1. requirements, and complexity. The remaining process al­
With these prices, a black box was constructed to get an ternatives were subjected to multiple design iterations until
upper bound on the profitability of the process. Based on the one process alternative remained. This process alternative
overall reaction (Equation 1) and the component prices was worked out in full and was further developed using
(Table 1), the black box break-even point is obtained at a yield Aspen Plus process simulation software.
of 23% for ethylene. CO2 captured from industrial emissions The detailed design of this process alternative includes
was considered (Irlam, 2017). equipment design, heat integration, process control, safety, a
sustainability analysis, an economic perspective, CAPEX,
2.1. Electrochemistry OPEX, and concludes with a techno-economic evaluation.

The electrochemical cell is the basic building block of the 3.1. Process alternatives
proposed reactor. Three parameters determine the perfor­
mance of an electrochemical reactor (ECR): Faradaic At the start, all imaginable process alternatives containing
Efficiency (FE), current density (j), and space-time yield (Y) electrochemistry were on the table. Both direct and indirect
(Fuller Thomas and Harb John, 2018). The mathematical electrochemical conversion of CO2 to ethylene was in­
formulas are represented in Eqs. 5, 6, and 7. The goal is to vestigated.
maximize the space-time yield with the FE and the current For the direct route, a literature study was conducted. On
density being the key variables. To that effect, higher FE the basis of electrocatalyst development stage, production
means more energy in the form of electrons being trans­ capacity, and scalability, a selection of five different elec­
ferred to the product, while a higher current density results trocatalysts was made out of tens of published experimental
in requiring less membrane surface. The applied potential studies. These five electrocatalysts were studied in more
can be divided into the theoretical potential and the over­ detail and the most promising electrochemical setup was
potential. The total overpotential is a combination of re­ chosen as the basis of the direct electrochemical conversion
sistance in the electrocatalysts, electrolyte solution, of CO2 to ethylene.
membrane and the catalytic overpotential. In general, for For the process alternatives based on indirect electro­
CO2 reduction reactions, substantial overpotential can be chemical CO2 conversion, a suitable intermediate chemical
expected (Kortlever et al., 2015). had to be chosen. Both syngas and methanol were deemed
m suitable intermediates, based on industrial preferences and
n Mi F
i
FEi = (5) on the fact that they are already established chemical
Q theoretical
building blocks (Mar, 2009). A literature search was con­
I ducted to find an experimental basis for the electrochemical
j= =F zi Ni (6) cell for producing each considered intermediate. Per process
A
alternative, again roughly five electrochemical technologies
jar Mi were compared and the most promising technology was
Y = FEi × (7)
nF chosen. Various schematic drawings and the combination of
currently employed technologies resulted in the addition of
The CO2 molecule has an energetically favorable linear
four more process alternatives to be investigated further. In
structure. Therefore, for reactions to occur, this structure
total, the five process alternatives depicted in Fig. 1 were
either needs to be bent or to be reformed towards a CO2 ra­
selected.
dical (Hernandez et al., 2017). The first requires a lot of en­
The electrochemical step in two of the five process alter­
ergy input and the thermodynamic potential towards the
natives was (currently) deemed unready for large-scale pro­
latter is high (E∘= - 1.9 V vs. NHE).
duction. Three remaining process alternatives were
A detailed investigation into the electrochemical reduc­
subjected to a first economic analysis and a functional design
tion of CO2 gave insight into the challenges plaguing its
was made. This resulted in the choice for one final preferred
successful operation. One major challenge is the selection of
process alternative, which was designed in detail.
a suitable electrocatalyst that gives sufficient current den­
sities and Faradaic efficiencies while maintaining cost-ef­
fectivity and stability during the process cycle. Additional 3.2. Index flowsheet
hurdles can be found in tweaking operating conditions such
as temperature and pressure. All these factors hinder the An index flowsheet was developed for the desired process
scale-up from a single cell lab setup to multi-cell commercial alternative. All functional steps from the basic design were
production unit. evaluated thoroughly and suitable units to fulfill the reac­
tion- and separation steps required were selected. The se­
3. Methodology paration steps were chosen according to the method of
Barnicki and Fair (1990). Subsequently, optimal process
A systematic approach based on the process design princi­ conditions were determined and the process including heat
ples of Douglas (1988) was taken. In a first step, the overall exchangers and compressors was modeled in Aspen Plus.
 Chemical Engineering Research and Design 182 (2022) 194–206 197

was process selection by elimination of economically un­

favorable routes. The first screening was done solely based
on the electrochemical performance. Table 2 provides the
critical parameters used as the [Link] 3.
Hori et al. (1985) discovered that copper electrodes can be
successfully used for the reduction of CO2 to C1-C2 products,
including ethylene. In this case, CO2 and H2O could directly
be converted into ethylene, rather than through an inter­
mediary product. The direct conversion of CO2 to ethylene,
Fig. 1 – Overview of reaction routes. (Abbreviations: route 1, although deemed an attractive alternative, suffers
Methanol-to-Olefins [MTO], Fisscher-Tropsch [FT], Syngas- three major drawbacks which prevent any application on an
to-Methanol [STM]). industrial scale: Low Faradaic efficiency towards ethylene,
low current density, and catalyst stability issues (Kas et al.,
2015). Recent study by Huang et al. (2021), indicates that
Selection of an appropriate thermodynamic model was done using Copper-based GDE cells can enable higher current
following the selection scheme of Carlson Eric (1996). densities (up to 1 A/cm2). However, with Faradaic efficiency
to ethylene only reaching 28%, this route was rejected from
3.3. Equipment design further consideration.
Route 2 (indirect conversion via methanol) has the third-
In Aspen Plus, all process units were dimensioned. Major lowest Faradaic efficiency of the electrochemical reactor.
units were designed in detail and reactors were kinetically Furthermore, the electrochemical cell in route 2 produces a
modeled or designed based on literature when no kinetic liquid mixture of methanol, formic acid, water, and a ternary
data was available. For the electrochemical cell, an extensive electrolyte ([Bmim]PF6/CH3CN/H2) (Lu et al., 2018) and there­
literature search was performed and a full-scale reactor was fore requires an extensive separation train before methanol
designed based on the design and results of a laboratory- can be fed to the MTO process. All these aspects combined
scale reactor. also make this route economically less attractive. Lastly, the
electrochemical cell of route 2 is the furthest away from in­
3.4. Economics dustrial implementation. It has the lowest current density of
the indirect routes and operates completely in the liquid
Required capital expenditure (CAPEX) of the proposed design phase. For all of the reasons mentioned above, route 2 was
was estimated using a factor method. The factors as men­ deemed unviable.
tioned in Table 7.5 from Towler and Sinnott (2013). The base For route 3 (indirect conversion via syngas to ethylene),
value is the estimated total bare equipment cost, which is the high syngas yield (as high as 99% Verma et al. (2016)) is
indexed using the CEPCI value of 2018 of 610. The total bare promising, but getting the final product via Fischer-Tropsch
equipment cost is then multiplied with an ISBL factor of 3.3, is problematic. The output of Fischer-Tropsch gives a mix­
which is typical for fluids processing plants (Towler and ture of a large number of different hydrocarbons which have
Sinnott, 2013), to obtain the inside battery limits (ISBL) costs. to be either purified or cracked. Ethylene selectivity is very
The ISBL is converted into the total fixed capital costs in­ low with only 10% (de Klerk, 2013). This low selectivity in­
cluding offsites, design and engineering, and contingency troduces multiple reaction and separation steps which not
costs, by multiplying the ISBL costs with a factor of 1.82 only increase the process complexity but also lead to carbon
(Towler and Sinnott, 2013). Lastly, working capital is added as losses. Thus, the large diversity of products and low ethylene
10% of the total fixed capital costs to find the total capital yield is the reason for also discontinuing research into
expenditure (CAPEX) required. For all units, except the re­ route 3.
actors, the bare equipment costs were estimated using cost Route 4, the reaction route with syngas and methanol as
correlations of Towler and Sinnott (2013) based on unit di­ intermediary products, makes use of the same electro­
mensions. An estimation of the equipment costs of the re­ chemical set-up (Dufek et al., 2011) as discussed for the Fi­
actors was made based on Aspen Plus Economic Analyzer or scher-Tropsch route. Though, unlike the wide range of
on economic data from the literature. products resulting from Fischer-Tropsch, the methanol-to-
Fixed operational expenditure was determined by esti­ olefins process can produce the desired olefins more selec­
mating operating labor, maintenance and the annualized tively. It is therefore deemed a suitable reaction route for
capital charge. Variable operational expenditure was esti­ further investigation.
mated based on raw materials and plant utilities. Product The last reaction route, route 5, has mostly the same
income of ethylene, propylene, and oxygen was considered process approach as route 4 but does not include the elec­
to determine the profitability of this process. trochemical reduction of CO2. Only hydrogen is produced in
the electrolyzer, which is based on proven technology
4. Decision on reaction route (Buttler and Spliethoff, 2018). However, feeding CO2 instead
of CO to the syngas-to-methanol reactor results in a lower
As described in the process alternatives section, the initial single-pass conversion and a larger water fraction in the
literature review resulted in the selection of five process product of this reactor, in comparison with the previous
routes. These routes are visualized in Fig. 1. The next step route, route 4 (Roode-Gutzmer et al., 2019). The decision for
 198

Table 2 – Overview of the reaction routes’ electrochemical performance.

Route Cathode T [∘ C] P [atm] Electrolyte Current density mA/cm2 Faradaic efficiency [%] By-products Ref

1 Cu nanoparticles 70–90 1 KHCO3 2.5 44 H2, HCOOH, CH3OH Kas et al. (2015)
2 Cu1.63Se ambient 1 PF/CH3CN/H2O 41.5 75 H2, HCOOH, CO Lu et al. (2018)
3&4 Ag 70–90 1 K2SO4 120 100 – Dufek et al. (2011)
5 Ni alloy 70–90 1 KOH 400 70 – Buttler and Spliethoff (2018)

Table 3 – Stream table of process diagram shown in Fig. 2.

Stream no. 1 2 4 5 6 7 8 9 10 11 12 13 16 17 23 24 25 26
From Mixer Mixer R-101 S-101 S-101 S-102 Mixer S-102 R-102 S-103 S-103 R-103 D-T D-T D-T D-T
To Mixer R-101 Mixer R-101 S-101 S-102 Mixer Mixer R-102 Mixer S-103 Splitter R-103 S-104 CH4 out C2H4 out C3H6 out C3 + out

Mass flow [t/h] – – – – – – – – – – – – – – – – –

CO2 – 0.01 77.66 352.96 282.37 282.35 0.01 7.06 33.49 275.30 28.52 26.97 1.56 1.56 – – – –
CO – – – – 44.93 44.93 – 44.93 63.87 – 19.35 19.32 0.02 0.02 0.02 0 – –
H2 – – – – 7.52 7.52 – 7.52 27.98 – 20.89 20.88 0.01 0.01 0.01 0 – –
Chemical Engineering Research and Design 182 (2022) 194–206

CH3OH – – – – – – – – 3.70 – 58.25 3.78 54.47 – – – – –

H2O 67.40 96.10 – – 28.90 0.02 28.88 0.02 0.02 – 2.31 0.05 2.25 30.61 – – – –
O2 – – – – – – – – – – – – – 2.01 1.94 0 – –
C2H4 – – – – – – – – – – – – – 12.16 0.12 12.5 0.01 –
C3H6 – – – – – – – – – – – – – 7.39 0 – 8.88 0.04
C3 + – – – – – – – – – – – – – 2.38 – – – –
CH4 – – – – – – – – – – – – – 1.90 1.80 0.01 – –
C2H6 – – – – – – – – – – – – – 0.12 – 0.01 0 –
C3H8 – – – – – – – – – – – – – 0.12 – – 0 0.12
Total 67.40 96.11 77.66 352.96 363.72 334.82 28.89 59.53 129.06 275.30 129.32 71 58.31 58.28 3.89 12.52 8.89 0.16
 Chemical Engineering Research and Design 182 (2022) 194–206 199

Fig. 2 – Simplified process flow diagram of route 4. Corresponding stream table is shown in Table 3.

discontinuing route 5 is based on the thermodynamic lim­ based gas diffusion cathode, a Ru-based anode, and a Nafion-
itations that result from feeding CO2 in a methanol reactor 424 cation exchange membrane. In the original experiments
instead of CO. by Dufek, K2SO4 is used as catholyte and KOH as an anolyte.
So in conclusion, route 4 is deemed the most viable fol­ However, recent studies by Huang et al. (2021) have shown
lowing Douglas’ method, based on efficiency, selectivity, that local alkaline conditions present near the anode, due to
process complexity, and thermodynamic limitations. KOH, lead to consumption of the electrolyte. Leading to the
formation of carbonate ions (CO3^2-), thus decreasing the
5. Process design results overall carbon efficiency of the cell. Thus, to avoid this, acidic
media is decided to be implemented. To simplify the ionic
We will now describe the full process design of the selected balances, sulfuric acid is used both as a catholyte and as an
route. A flowsheet is shown in Fig. 2. The number between anolyt. The inflow of reactants and outflow of products are
brackets indicates the number of the corresponding stream. indicated in Fig. 3. The stack is operated at atmospheric
First, CO2 (#4) and water (#1) are mixed with recycle streams conditions and 70 ∘C.
#11 and #8 and fed to the electrochemical reactor (R-101). In Experiments by Dufek et al. (2011) on this setup showed
this reactor, CO2 and water are converted into syngas and that the application of 3.5 V yielded a current density of
oxygen. In literature, the single-pass conversion is barely 1200 A/m2. If a lower voltage was applied the current density
described. We have estimated a 20% single-pass conversion fell, and a larger reactor would be required for the same
for CO2 based on Ripatti et al. (2019). This single-pass con­ production goal. This trade-off is emphasized in Fig. 4. The
version is also used in the mass balance calculations. Oxygen simultaneous reactions in this co-electrolysis cell produce
(#3) leaves the reactor as a product stream. Then two se­ syngas, unlike other alternatives with higher current den­
paration steps are in place. The first separation unit is a sities (Huang et al., 2021). The total electrical duty of oper­
simple flash drum to remove water and the second is an ating the reactor can be calculated by multiplying the applied
absorber to remove unreacted CO2. CO2 (#11) and water (#8) cell potential, the current density, and the electrode area. A
are fed back into the electrolyzer while produced syngas (#9) cost basis of 45 $ per MWh operating power is used
is fed to the syngas-to-methanol reactor (R-102). A flash (Energiemarktinformatie, 2021), alongside a capital cost price
drum after the syngas-to-methanol reactor is used to sepa­ of 920 $/m2 electrode (as will be discussed in the economic
rate methanol from unreacted syngas. The syngas (#13) is evaluation chapter).
recycled into the inlet of the reactor R-102 after purging a Fig. 4 shows the electrochemical reactor costs as a func­
small amount and methanol (#16) is fed to the methanol-to- tion of the applied cell potential. The annualized capital costs
olefins reactor (R-103). In this reactor, ethylene and propy­ are calculated by dividing the CAPEX of the total plant by the
lene are produced with 100% conversion and 80% selectivity depreciation period (10 years). The economic optimum is
(Vora et al., 1997). To purify the produced ethylene, the re­ found at a cell potential of 2.3 V, which yields a current
actor is followed by a separation train, which includes an density of 400 A/m2. The total electric duty of the reactor is
olefin cracker (OCP=R-104) to increase the ethylene yield then roughly 660 MW.
(Minoux et al., 2009). Finally, 99% pure ethylene is obtained in In this design, two hundred individual cells are assembled
stream #24. Stream #25 contains 99% propylene and stream in a single stack. The design parameters are based on avail­
#26 multiple C3 + components. Stream #23 consists of me­ able technology from several electrolyzer suppliers
thane, which is burned. The heat is utilized for heat in­ ([Link], 2017). The 2200 stacks are connected to the
tegration in the process. Below, the unit design is discussed power supply in parallel. The distance between the elec­
in more detail. trodes is 35 mm, which results in a stack length of roughly
7 m. It is assumed that the entire electrode is active and
5.1. Electrochemical reactor equals 1.8 m2 electrode material per cell. The total active
electrode area required encompasses roughly 790,000 m2.
The electrochemical reactor in this design is a stacked-cell The total volume taken by the electrolyzers is 28,000 m3. A
system, linearly scaled-up from the experimental setup laid large volume but not unrealistic for this capacity of 660 MW
out in the work of Dufek et al. (2011) It comprises an Ag- (see Section 6.3).
200 Chemical Engineering Research and Design 182 (2022) 194–206

The reactor was modeled in Aspen Plus as a kinetic plug

flow reactor, following the kinetics as described by Van Den
Bussche and Froment (1996). The work of Chen et al. (2011)
describing the modeling of an industrial Lurgi reactor was
taken as the basis for this design. Optimization of the reactor
length and the feed ratio of hydrogen to CO and CO2 (slightly
above stoichiometric) led to a single-pass carbon conversion
of 49% and an overall carbon conversion of 96%.

5.2.2. Methanol to olefins (MTO) reactor

Methanol to olefins has grown to be a well-established pro­
cess in synfuel technology since its inception in the 1970s
(Gogate, 2019). Out of the three most used reactor designs
currently in practice, it was decided to adapt UOP/Hydro’s
advanced MTO Honeywell UOP A Different Game Plan (2021).
Fig. 3 – Layout of a single electrochemical cell. The reactor comprises a twin bed fluidized catalytic cracker
and regenerator setup, operating at a temperature of 425 ∘C
and 2 bar pressure. Modified SAPO-34 is used as a catalyst
achieving a conversion of > 99%. As alluded by its name, the
products are a mixture of olefins with carbon efficiency to­
wards ethylene/propylene (C2 =/C3 =) of around 80%. This
can be further improved to 85–90% by cracking the higher
olefins (like butene and pentene) in an Olefin Cracking Pro­
cess (OCP) reactor.
For the Aspen Plus implementation of the reactors, instead
of a rigorous design, a simplified approach was used based
on the work of Vora et al. (1997) Thus, a stoichiometric re­
actor with predefined conversions (100% conversion and 80%
selectivity towards ethylene and propylene) and previously
mentioned reactor conditions was used for the MTO reactor.
Similarly for the OCP reactor, a stoichiometric reactor at 620 ∘
C and 2 bar with a 100% conversion and a 4:1 ratio of pro­
pylene-to-ethylene in the product (Minoux et al., 2009) was
simulated.

5.3. Separation units

Fig. 4 – Annual costs of the electrochemical reactor as a
function of the applied cell potential. The separation units are divided into three sections. The first
section is between the electrolyzer and the syngas-to-me­
Three important assumptions regarding the electro­ thanol reactor. The second section is located between the
chemical reactor are made: (1) The reaction rate is in­ syngas-to-methanol reactor and the methanol-to-olefins re­
dependent of the CO2 concentration (inlet concentration is actor and the third section is the separation train after the
35 mol/m3). (2) The mass transport of H+ through the mem­ methanol-to-olefins reactor (see figure 2). All separation
brane is not rate-limiting (3). The reactant flow is equally units were dimensionalized in Aspen Plus.
divided among the active stacks. The first section consists of a flash drum for the syngas/
water split and an absorber to remove CO2. In the flash drum,
water and syngas are split with less than 1 wt% water in the
5.2. Other reactors vapor phase. This flash drum (S-101) actually consists of a
train of three compressors, coolers and phase separators in
5.2.1. Syngas-to-methanol reactor series, to prevent compressor overheating while removing
The reactor type used for the design of the syngas-to-me­ most of the water in the product stream. The absorber (S-102)
thanol reactor is a quasi-isothermal boiling water reactor. is used to remove unreacted CO2. A maximum of 3 mol% CO2
The reaction takes place in the gaseous phase at the tube is allowed in the product stream (#9) for optimal feed con­
side of the reactor, which is filled with a Cu/ZnO/Al2O3 cat­ ditions (stoichiometric amount of hydrogen with respect to
alyst as is typical in industry (Palma et al., 2018). The shell CO and CO2) for the syngas-to-methanol reactor. With a
side of the reactor is filled with water at boiling point to re­ single-pass conversion of 20% in the electrochemical reactor,
move excess heat produced by the formation of methanol this means that 97.5% of the unreacted CO2 needs to be re­
o
( H298K = −90 kJ/mol) as pressurized steam (Palma et al., 2018; moved. Selexol®is used as solvent due to its proven use, en­
Bozzano and Manenti, 2016). High temperature and pressure ergy efficiency, and ease of regeneration (Padurean et al.,
are beneficial as reaction conditions to favor the formation of 2012). The solvent regeneration cycle is not shown in detail
methanol in the equilibrium reaction, but the operating in the simplified process flow diagram (figure 2), but it is
temperature is limited to 300 ∘C to prevent catalyst sintering modeled in the extended Aspen Plus flowscheme.
(Palma et al., 2018; Roode-Gutzmer et al., 2019). The reactor is The second section consists of a flash separator (S-103).
operated at 70 bar and between 225 and 300 ∘C. Due to a single-pass conversion of 49% in the syngas-to-
 Chemical Engineering Research and Design 182 (2022) 194–206 201

methanol reactor, the reactor outlet stream consists of a

significant fraction of unreacted syngas. The flash separator
(S-103) purifies this product stream to 91 mol% methanol,
6 mol% water, and 3 mol% CO2. This is pure enough for the
MTO reactor and further purification is a waste of energy.
The third section is the separation train for ethylene
purification. The sequence of this is based on Barnicki’s
method (Barnicki and Fair, 1990) with phase separation as a
first step. A flash drum (S-104) is used to remove the water
from the gases. Another flash drum (S-111) is placed after the
olefin cracker to remove the water formed in this reactor.
After the first water removal, CO2 is removed. Due to water
presence, absorption with MEA is used. The amount of CO2 in
the outgoing stream is negligible. The remaining hydro­ Fig. 5 – Pie chart of ISBL costs for the process presented in
carbons are separated by distillation, as is common practice Fig. 2. Values represent MM$.
in industry (Salkuyeh and Adams, 2015). In stream #24, 99 wt
% pure ethylene is obtained. Stream #25 contains 98 wt%
pure propylene, which can also be sold as a valuable by­ Production Model of the National Renewable Energy
product. Finally, in stream #26 another range of C3 + products Laboratory (U.S. Department of Energy, 2009). This approach
is present, which also have value, and stream #27 is sent to was also suggested by Jouny et al. (2018) Another approach
an OCP reactor for cracking the higher olefins. could be to estimate the electrolyzer costs based on capital
cost estimates per kW for large-scale water electrolysis sys­
6. Techno-economic evaluation tems. Using a targeted bare capital cost of 242 $ per kW and
an installment factor of 1.37 as reported by the Office of En­
The selected reaction route (route 4) is evaluated on two ergy Efficiency & Renewable Energy of the United States
main aspects: economics and CO2 reduction. First, an eco­ Government (U.S. Department of Energy, 2020), this results in
nomic analysis was performed and a future projection is an estimated installed cost for the electrochemical reactor of
made. The designed route is then compared to competing 253 MM$. However, this approach does not take into account
existing technologies. However, with reducing CO2 emissions the large difference in required electrode area for both types
as an important goal of the designed plant, the environ­ of systems caused by a difference in the current densities
mental aspects of the proposed process cannot be ignored. In (400 A/m2 for R-101 versus 10,000–20,000 A/m2 for commer­
the second part of this evaluation, CO2 reduction aspects of cial PEM electrolyzers (Buttler and Spliethoff, 2018). A third
this plant are described. approach, using the costs per tonne of Cl2 produced in a
chloralkali process as given by Brinkmann et al. (2014) re­
6.1. Economic evaluation sulted in an estimated capital cost of 1876 MM$. Here, it is
assumed that the costs of producing a mole of Cl2 are the
Fig. 5 shows the breakdown of the ISBL costs of the proposed same as for producing a mole of CO or H2, as they require the
plant, with a total of 1095 MM$. The corresponding unit same electron transfer. This last estimate is assumed to be
numbers can be found in the process flow diagram in Fig. 2. far too high, since the chloralkali process requires all mate­
Multiplying the ISBL with a factor of 1.82 to obtain the total rials used to be corrosive resistant, which is not required for
fixed capital costs including offsites, design and engineering, this process. Summarizing, this shows how difficult it is to
and contingency costs, and adding 10% of the total fixed make a proper estimate of the costs of an electrochemical
capital costs as working capital, this adds up to an enormous reactor, whereas it also shows how important it is to be able
expected CAPEX of 2191 MM$. However, the extremely high to make a good estimate as it is a huge part of the capital
installed costs of the electrochemical reactor (848 MM$, 77% costs of an electrochemical process. For the base case of the
of the total ISBL costs) may make the conventional formula economic analysis, a capital cost of 848 MM$ for R-101 and a
for estimating the costs of offsites, design and engineering, total CAPEX of 2191 MM$ were considered.
contingency costs, and working capital as fraction of the ISBL Fig. 6 gives an overview of the OPEX of the proposed
costs out of proportion. A lower limit is obtained by first process. The electrical duty of R-101 is by far the biggest
calculating the CAPEX costs of the plant on basis of the plant contributor to the OPEX costs, requiring a duty of 658 MW. Of
ISBL costs excluding the installed costs of the electro­ this electrical duty, 230 MW is lost as heat. As the reactor
chemical reactor. In the last step, the installed costs of the operates at 70 ∘C, this energy can be considered as useless
electrochemical reactor are directly added to the CAPEX costs waste heat. This shows a severe limitation of this reactor,
of the plant excluding this reactor. Following this approach, a while if this generated heat was at a higher temperature, it
lower limit for the CAPEX of this process of 1340 MM$ is would have residual value (for instance by steam produc­
obtained. This shows how influential the costs of just the tion), decreasing the OPEX of the process.
electrochemical reactor are for the CAPEX of the plant and Taking capital recovery and sales of propylene and
raises the question whether such a general factor method is oxygen as side products into account, the cost of production
suitable for such a process. More in-depth analysis is re­ is 1865 $/t ethylene. This production cost is not competitive
quired to make a more accurate CAPEX estimation. with the considered current ethylene price (1050 $/t) or any
The installed costs of the electrochemical reactor (848 MM of the potential new ethylene sources, which have estimated
$) have been estimated based on the capital costs per square costs between 920 and 1470 $/t ethylene (McKinsey&
meter electrode area of a commercial water electrolysis Company, 2012). However, several parameters have sig­
system, as described by the H2A Central Hydrogen nificant impact on the production costs of ethylene for this
202 Chemical Engineering Research and Design 182 (2022) 194–206

Fig. 6 – Pie chart of OPEX. Values represent MM$/y.

Fig. 8 – Plant’s CAPEX costs as a function of the plant

capacity.

not benefit from the economy of scale. For the OPEX, this
value is between 76% and 89%, depending on the CO2 price
(80 $/ton or 0 $/ton). These two percentages are based on the
reactant and electricity prices. Therefore, a careful statement
can be made that the effect of economy of scale on the
plant’s economy is limited when the electrolyzer costs will
increase approximately linearly with scale.
Fig. 8 shows the effect of an increase in current density of
the electrolyzer on the CAPEX of the plant, assuming this
current density can be reached using the same cell potential,
due to a reduction in cell resistances. The resulting CAPEX is
Fig. 7 – Plant’s CAPEX costs as a function of the plant calculated using the same two methods as described pre­
capacity. viously. A doubling of the current density reduces the plant’s
CAPEX by 30–40%. If the electrolyzer’s current density could
process and could, in favorable conditions, significantly re­ be increased a five fold to 0.2 A/cm2 (reaching 10% of the
duce the estimated costs. maximum current density of commercial PEM electrolyzers),
the plant’s CAPEX could even drop to 660 MM$.
6.1.1. Economic sensitivity analysis Regarding operational expenses, Fig. 6 show that the
In the previous section, it was already shown that the cal­ electricity price is extremely dominant on the operational
culation method of the plant’s CAPEX with respect to the costs. Costs of CO2 contribute the second most to the op­
enormous costs of the electrolyzer can lead to a significant erational expenditure. Regarding the electricity price, their is
difference in estimated plant’s CAPEX (2191 MM$ when the a large uncertainty for the coming years. On the short term it
electrolyzer costs are included when calculating all addi­ is expected to rise (U.S. Energy Information Administration,
tional costs to the ISBL that make-up the plant’s CAPEX, 2020), whereas on the long term it is expected to fall (IRENA,
versus 1340 MM$ when the electrolyzer is not included when 2020). When CO2 emission is not taxed, the price of CO2 can
calculating these additional expenses). Two other factors can be as high as 80 $/t, which was assumed in this case. Dif­
have a large impact on the plant’s CAPEX: the plant scale and ferently, in the European Union, CO2 is taxed (Autoriteit,
the current density of the electrolyzer. The designed process 2021), effectively giving it a price tag of − 30 ?/t CO2 in 2020
has a scale of only 100 kta. Industrial size ethylene plants can and it increases by around 10 ?/t annually until − 125
reach ten times this production. Fig. 7 shows the effect of a ?/ton in 2030. The European Union also introduced an
plant capacity increase assuming a typical cost correlation emission trade system (EU ETS) (Emissieautoriteit, 2021) in
coefficient of 0.6 (Towler and Sinnott, 2013). Whether the which emission rights can be traded between companies that
electrolyzer also benefits from economies of scale to this emit too much and companies that emit less than allowed,
extent is doubtful. Fig. 7 shows the effect of a capacity in­ effectively giving the emission rights a price.
crease in case also the electrolyzer benefits from the econo­ The effects of the uncertainty in CAPEX and OPEX on the
mies of scale and the case in which the electrolyzer cost estimated costs per ton of ethylene are shown in Fig. 9. The
remains linear with the electrode area required. In both cases used ranges for the parameters are either given as + 50 and
huge CAPEXes are obtained for industrial sized ethylene − 50% of the base value or as [optimistic value, base value,
plants. pessimistic value (if considered)]. For these calculations,
The difference in the two CAPEX graph lines in Fig. 7 can the electrolyzer was assumed to also benefit from econo­
be explained by the influence of the electrolyzer. If it is as­ mies of scale and the electrolyzer was included in the ISBL
sumed that the electrolyzer will not benefit from the when calculating the CAPEX based on the ISBL. Further­
economies of scale, over 78% of installed equipment costs do more, capital recovery and sales of oxygen and propylene
 Chemical Engineering Research and Design 182 (2022) 194–206 203

For comparison, conventional ethylene plants emit over

0.7 ton CO2-eq per ton of ethylene (Ghanta, 2014). It must be
noted that heating is not yet included in the CO2 calculation
of the proposed design and the restriction is applied that all
electricity used in the process should come from renewables.
For heating, methane burning is assumed to be the source,
which results in emissions due to heating of 0.59 ton CO2 per
ton HVC. The net CO2 intake is thus 2.5 ton CO2 per ton HVC,
or 4.8 ton CO2-eq per ton ethylene. If the reaction has an
efficiency of 100% towards ethylene, the CO2 intake without
heat taken into account is 3.1 ton CO2-eq per ton ethylene
produced. This number is lower than the 4.8 ton mentioned
before because there is no CO2 that goes into other products.
Finally, if the electrolyzer is fueled by gray, non-renewable
energy, the CO2 emission rises to almost 19 ton CO2 emitted
Fig. 9 – Influence of most important parameters on per ton HVC. Due to its high energy usage (658 MW) it is thus
ethylene production costs. of great importance that the electrolyzer uses renewable
energy.
as side products were taken into account while calculating If all energy, except that for heating purposes, comes from
the production costs for ethylene. It is evident that indeed renewables, a CO2 intake of 2.5 ton per ton HVC is realized.
the electricity price and the costs/revenues of CO2 are the Further electrification of this heat supply is possible, but
most important factor for the resulting costs per ton of cumbersome, and deemed beyond the scope of this design.
ethylene. Furthermore, also the plant’s CAPEX has a very As stated, 52% of this HVC is ethylene. If extrapolated to the
significant influence on this figure. None of the evaluated world production of 283 million metric tons, expected in 2024
parameters can solely result in a profitable process, even in (Statistica, 2021), it means 1.3 billion tons of CO2 can be taken
the most favorable case. However, a future scenario in in per year. This is 4% of the annual CO2 emission.
which multiple of the evaluated parameters take on fa­
vorable values can most definitely result in a viable busi­ 6.3. Review on electrolyzer size
ness case.
An example of such a favorable future scenario, which To test whether the size of the electrolyzer is realistic, the
considers an electricity price of 20 $/MWh and − 120 $/t CO2, volume and area of the electrolyzer are checked, as well as
results in an ethylene production cost of around 600 $/tonne the amount of solar panels or wind turbines that have to be
and a positive operational result and gives a payback period used. The total volume of the electrolyzer is 28,000 m3. To see
between 17 and 18 years. The net present value is 1317 MM$ whether this is realistic or not, a comparison is made with
and is calculated using an initial investment sum of 2191 MM hydrogen electrolyzers due to the fact that no commercial
$ and a life-time of 30 years, with a scrap value of 540 MM$ CO2 electrolyzers exist yet. Although also for hydrogen no
and assumed product prices as specified in table 1. The de­ large scale electrolyzers are currently in operation (currently
preciation period is 10 years and a discount rate of 2% is the maximum is 20 MW [Link] (2021)), a study
used. The corporate tax rate is 25% of the taxable income. by the Institute of Sustainable Process Technology (ISPT) in
Improved cost efficiency for the electrolyzer can reduce the the Netherlands and IRENA (IRENA, 2020) showed that a
payback period quickly. To give an estimation of this decline, 1 GW electrolyzer would have a footprint up to 17,000 m2. If
we looked at hydrogen production via electrolysis and found the same footprint is assumed for the set-up of this paper, it
that the cost of producing hydrogen is expected to drop by would mean that the electrolyzer would be 1.65 m high,
three to four times in the near future (SP, 2021). which is realistic. A 660 MW electrolyzer will take 66 ha of
solar panels (based on 1 kW per m2 Linquip (2022)), which is
6.2. CO2 reduction evaluation four times the area of our electrolyzer. Instead of solar pa­
nels, 220 wind turbines would do as well (based on 3 MW per
This route to make ethylene is interesting because it has the turbine Good energy (2022)), which is also not unrealistic
potential to reduce CO2 emissions by using CO2 as a reactant. compared to the world amount of wind turbines (341,000 in
The Sankey diagram in Fig. 10 shows the carbon efficiency 2017 already CNBC (2017)). It must be noted however that this
and makes clear where the carbon of CO2 ends up in the is for a 100 kta plant. For a more realistic plant size, the
process. This shows that 51.5% of the carbon fed to the numbers must be ten-folded, so 170 ha for the electrolyzer
process in the form of CO2 end up in the desired product, area and 660 ha of solar panels or 2200 windmills. These
ethylene. Only 5.5% of the carbon does not end up in a va­ numbers are quite large but not unfeasible in places where
luable product. In our design, 78 ton of CO2 is fed every hour. large open spaces are available.
This translates in a CO2 intake of 3.1 ton CO2 per ton high-
value chemicals (HVC), of which 52% is ethylene. This cal­ 6.4. Qualitative review of key assumptions
culation was made by taking the mass of CO2 intake per hour,
correcting for the direct CO2 emission from the production One of the main reactants, carbon dioxide, is assumed to be
process and divide this by the mass of HVC produced in that without contamination and thus does not require pretreat­
hour, as stated in equation 8. ment. In practice, it is unlikely that a reliable source of waste
carbon dioxide is available with that quality. In any logical
m (CO2 )intake m (CO2 )emission
CO2, consumed = (8) point source, such as a factory purge or chimney, trace
m (HVC)produced
amounts of both inert and toxic components can be
204 Chemical Engineering Research and Design 182 (2022) 194–206

Fig. 10 – Carbon Sankey diagram of route 4.

expected. Inert substances will only lower the effectiveness ethylene and propylene yields, which combined add up
of the reaction system. Toxic components such as SO2 might to 80%.
be destructive for piping, the electrolyzers or other down­ The economic analysis highlights the critical position of
stream equipment. If these substances are expected, a the renewable electricity price for systems such as this. It
scrubber installation could be installed before the inlet to was also found that a secondary controlling economic factor
match the parameters laid out in this design. For the is the price of CO2, which is used as a reactant. An optimistic
downstream process, several assumptions are made re­ pricing scenario shows that under reduced renewable elec­
garding the selectivity of the syngas-to-methanol and me­ tricity prices, in combination with a CO2 tax, this design can
thanol-to-olefins processes. Although safety factors are become competitive with other innovations in the field of
incorporated in these designs, the selectivity factors should producing ethylene. An example calculation with pricing set
be checked consistently with the electrolyzer output. at 20 $/MWh and − 120 $/t CO2, results in a positive opera­
Undersized equipment could be detrimental to the conver­ tional result and gives a payback period between 17 and 18
sion of carbon dioxide towards valuable products. Another years. While conventional ethylene plants emit over 0.7 ton
key assumption is the origin of electricity. Whereas currently CO2-eq per ton ethylene, this design has a negative emission
most industrial processes run on some form of fossil fuel of 4.8 ton CO2 per ton of ethylene produced (which corre­
based electricity, the electricity supply of the designed fa­ sponds to 2.5 ton CO2 per ton of HVC). If proven economically
cility is assumed to be fully sustainable and non-inter­ viable on an industrial scale, a great positive environmental
mittent. If the electricity does not directly come from the impact can be made.
renewable source, but from the grid, this statement may not
hold and the impact of CO2 removal will be less.
8. Outlook

This research has shown that this process certainly has po­
7. Conclusion
tential for future implementation. To reach this im­
plementation, a couple of challenges have to be addressed.
This research covers the design of an electrochemical pro­
As mentioned in the economics section, one of the draw­
duction facility to convert CO2 into ethylene with a capacity
backs at the moment is the operating cost. Therefore it is
of 100 kta. Extensive literature research was done where
important to keep up to date on the development of the
several pathways for the electrochemical production of
electricity price and the price of CO2, as these two are the
ethylene have been evaluated. The most promising path­
most influential on the economic performance of this design.
ways were selected based on Faradaic efficiency, current
More can also be gained in terms of energy efficiency.
density, overpotential, and material availability. One of these
Currently, 35% of the energy is dissipated in low-value heat
routes led to the design presented in this work. In an elec­
that cannot be used elsewhere in the process. Synergy with
trochemical device, based on experimental research by
other chemical operations is essential. Next to that, it is
Dufek et al. (2011), CO2 and water are reacted to syngas. In a
possible that operating the electrochemical setup at a higher
sequential reactor, the syngas-to-methanol reaction is car­
temperature is beneficial. Temperatures above 70 ∘ C have
ried out. A methanol-to-olefins reaction, followed by a se­
not been tested for this setup yet.
paration train, results in ethylene, propylene, and, in lower
Another major impasse is the capital cost of the electro­
quantities, higher olefins.
chemical reactor. The design in this paper is based on a
The electrochemical cells are operated at 400 A/m2 and
single cell and data from one paper. This should be experi­
2.3 V and require a total duty of 658 MW. The syngas-to-
mentally validated and scaled up as quickly as possible to
methanol reactor was based on a Lurgi-type quasi-iso­
gather more insights into the working and costs of a larger-
thermal boiling water reactor. Optimization of the reactor’s
scale system. The capital costs of the electrochemical reactor
feed regarding the H2 and CO2 content resulted in a high
are a major part of the investment costs for such a plant and
single-pass conversion of 49% and a high purity methanol
being able to make more accurate CAPEX estimates is
product. The design for the methanol-to-olefins reactor was
therefore essential. Lastly, increasing the current density of
based on the advanced UOP-Hydro process, resulting in high
such systems, while using cheap materials should be a main
 Chemical Engineering Research and Design 182 (2022) 194–206 205

research focus in the coming years to bring down the capital power-to-gas and power-to-liquids: a review. Renew. Sustain.
costs of such a system. Energy Rev. 82, 2440–2454.
Carbon Capture, Utilisation and Storage (CCUS), 2020; 〈https://
Route 5 (only producing hydrogen electrochemically) was
[Link]/onderwerpen/duurzaam-ondernemen/energie-
also deemed interesting due to the proven technology of
en-milieu-innovaties/carbon-capture-utilisation-and-
hydrogen electrolyzers. However, a lower single-pass con­ storage〉, [ Accessed 25-November-2020].
version is obtained when CO2 instead of CO is fed to the Carlson Eric, C., 1996. Aspen technology Inc., Don’t gamble with
syngas-to-methanol reactor. A more in-depth comparison physical properties for simulations. Chem. Eng. Prog. 92 (10),
between route 5 and the evaluated route (route 4) is re­ 35–46.〈[Link]
commended to identify the most viable option. [Link]〉.
Chemical Engineering - Essentials for the CPI professional, The
Finally, a proper LCA analysis could be done to evaluate
chemical engineering plant cost index.〈[Link]
the impact of this process on the environment in larger de­ [Link]/pci-home〉, [ Accessed 18-June-2020].
tail. The main impact categories recommended would be the Chen, L., Jiang, Q., Song, Z., Posarac, D., 2011. Optimization of
following: land occupation, climate change, fossil depletion, methanol yield from a Lurgi reactor. Chem. Eng. Technol. 34
metal depletion and water depletion. (5), 817–822.
In general, the intent of this paper is to show that in these CNBC, 2017. There are over 341,000 wind turbines on the planet:
changing times where increasing demand for less fossil- Hereas how much of a difference theyare actually making.
[Link]
based products is present, electrochemical approaches
[Link], [ 8-
should not be discarded. Electrification of an established September-2017].
process in an industry as large as the ethylene production COM, 2011. 112 - A Roadmap for moving to a competitive low carbon
could make an important impact on the world’s carbon economy in 2050.2011; 〈[Link]
emission. When in the coming years research activities on ments/com-2011–112-a-roadmap〉, [ Accessed 19-March-2020].
this topic increase and policies favoring renewable ap­ Douglas, J., 1988. Conceptual Design of Chemical Processes, 1st
ed. McGraw-Hill.
proaches are passed, this process can become economically
Dufek, E.J., Lister, T.E., McIlwain, M.E., 2011. Bench-scale elec­
viable and make a large step in reaching our coming climate
trochemical system for generation of CO and syn-gas. J. Appl.
goals. Electrochem. 41, 623–631.
[Link], 2020. Propylene market report and price
Declaration of Competing Interest trend. 〈[Link]
textile-market-watch/propylene-price-trends-industry-
reports/17/〉, [ Accessed 13-March-2020].
The authors declare that they have no known competing fi­
Fuller Thomas, F., Harb John, N., 2018. Chapter 1.2 –
nancial interests or personal relationships that could have Characteristics of electrochemical reactions. Electrochemical
appeared to influence the work reported in this paper. Engineering. John Wiley & Sons, Chapter 1.2 Characteristics of
Electrochemical Reactions, isbn: 978-1-119-00425-7.
Acknowledgment Ghanta, M., 2014. Environmental impacts of ethylene production
from diverse feedstocks and energy sources. Appl. Petrochem.
Res. 4.
The authors thank Matthijs Ruitenbeek, Cees Biesheuvel and
Gogate, M.R., 2019. Methanol-to-olefins process technology: cur­
Ronald Wevers of DOW Benelux BV for their guidance rent status and future prospects. Petrol. Sci. Technol. 37,
throughout the process. 559–565.
Good Energy, 2022. How do wind turbines work? 〈[Link]
Appendix A. Supporting information [Link]/how-do-wind-turbines-work/〉, [ Accessed
06-February-2022].
Green hydrogen cost reduction, scaling up electrolysers to meet
Supplementary data associated with this article can be found the 1.5∘C climate goal; IRENA, International Renewable Energy
in the online version at doi:10.1016/[Link].2022.03.034. Agency: Abu Dhabi, 2020.
Hernández, S., Amin Farkhondehfal, M., Sastre, F., Makkee, M.,
References Saracco, G., Russo, N., 2017. Syngas production from electro­
chemical reduction of CO2: current status and prospective
implementation. Green Chem. 19, 2326–2346.
Administration, U. E. I., 2020. Short term energy outlook.
Honeywell UOP A Different Game Plan, 2021.〈https://
〈[Link] [
[Link]/different-game-plan-honeywell-uop-pdf-
Accessed 23-June-2020].
[Link]〉, Accessed:2021-02-11.
Autoriteit, E., 2021. Voorlichting CO2 heffing [Link]://
Hori, Y., Kikuchi, K., Suzuki, S., 1985. Production of CO and CH4 in
[Link]/actueel/nieuws/2020/11/27/voor­
electrochemical reduction of CO2 at metal electrodes in aqu­
lichting-co2-heffing-industrie, Date visited: 11-may-2021.
eous hydrogen carbonate solution. Chem. Lett. 14, 1695–1698.
Barnicki, S.D., Fair, J.R., 1990. Separation system synthesis: a
Huang, J.E., 2021. CO2 electrolysis to multicarbon products in
knowledge-based approach. 1. Liquid mixture separations.
strong acid. Science 372, 1074–1078.
Ind. Eng. Chem. Res. 29, 421–432.
IES, 2019. Global Energy & CO2 Status Report.2019; 〈[Link]
Bhattacharyya, D., Turton, R., Shaeiwitz, J.A., Whiting, W.B.,
[Link]/reports/global-energy-co2-status-report-2019〉.
Bailie, R.C., 2012. Analysis, Synthesis, and Design of Chemical
IES, 2019. Tracking Industry.2019; 〈[Link]
Processes, Fifth Edition. Prentice Hall.
tracking-industry〉, [ Accessed 19-March-2020].
Bozzano, G., Manenti, F., 2016. Efficient methanol synthesis:
[Link], E.M. , Beurzen, Elektra.〈[Link]
perspectives, technologies and optimization strategies. Prog.
[Link]/beurzen/elektra/〉, Date visited: 04-
Energy Combust. Sci. 56, 71–105.
may-2021.
Brinkmann, T., Santonja, G.G., Schorcht, F., Roudier, S., Delgado
Intratec, Oxygen Price, 2007. 〈[Link]
Sancho, L., 2014. Best Available Techniques (BAT) Reference
chemical-markets/oxygen-price〉, [ Accessed 05-March-2020].
Document for the Production of Chlor-alkali. European
IRENA, 2020. Falling Renewable Power Costs Open Door to
Commission, (Industrial Emissions Directive 2010/75/EU).
Greater Climate Ambition.〈[Link]
Buttler, A., Spliethoff, H., 2018. Current status of water electro­
newsroom/pressreleases/2019/May/Falling-Renewable-
lysis for energy storage, grid balancing and sector coupling via
206 Chemical Engineering Research and Design 182 (2022) 194–206

Power-Costs-Open-Door-to-Greater-Climate-Ambition〉, [ 29- utilization: electrochemical conversion of CO2 to ethylene.

May-2020]. Front. Energy Res. 8, 12.
Irlam, L., 2017. Global Cost of carbon capture and storage. https:// Patel R, M., 2016. Chapter 2 - Polyethylene. Multilayer Flexible
[Link]/archive/hub/publications/ Packaging (Second Edition). William Andrew Publishing,
201688/[Link]. Chapter 2 - Polyethylene.
Jouny, M., Luc, W., Jiao, F., 2018. General techno-economic ana­ Platts Global Ethylene Price Index, 2020; 〈[Link]
lysis of CO2 electrolysis systems. Ind. Eng. Chem. Res. 57, platts/en/our-methodology/price-assessments/petrochemicals/
2165–2177. pgpi-ethylene#〉, [ Accessed 09-December-2020].
Kas, R., Kortlever, R., Yílmaz, H., Koper, M.T.M., Mul, G., 2015. [Link], 2021. World’s largest green hydrogen project
Manipulating the hydrocarbon selectivity of copper nano­ - with 100MW electrolyser - set to be built in Egypt.〈https://
particles in CO2 electroreduction by process conditions. [Link]/energy-transition/worlds-largest-
ChemElectroChem 2, 354–358. green-hydrogen-project-with-100mw-electrolyser-set-to-be-
de Klerk, A., 2013. Chapter Fischer-Tropsch Process. Kirk-Othmer built-in-egypt/2–1-1104709〉, [ 24-November-2021].
Encyclopedia of Chemical Technology. John Wiley & Sons Inc, Ripatti, D.S., Veltman, T.R., Kanan, M.W., 2019. Carbon monoxide
pp. 1–20. gas diffusion electrolysis that produces concentrated C2 pro­
Kortlever, R., Shen, J., Schouten, K.J., Calle-Vallejo, F., Koper, M., ducts with high single-pass conversion. Joule 3, 240–256.
2015. Catalysts and reaction pathways for the electrochemical Roode-Gutzmer, Q.I., Kaiser, D., Bertau, M., 2019. Renewable
reduction of carbon dioxide. J. Phys. Chem. Lett. 6, 4073–4082. methanol synthesis. ChemBioEng Rev. 6, 209–236.
Administration, U.E.I., 2020. Natural Gas.〈[Link] Salkuyeh, Y.K., Adams, T.A., 2015. Co-production of olefins, fuels,
dnav/ng/hist/[Link]〉, [ Accessed 18-June-2020]. and electricity from conventional pipeline gas and shale gas
Laboratory, 2009. N. R. E. Current state-of-the-art hydrogen pro­ with near-zero CO2 emissions. Part I: Process development
duction cost estimate using water electrolysis.〈[Link] and technical performance. Energies 8, 3739–3761.
[Link]/pdfs/[Link]〉, [ Accessed 17-June- Siemens, Hydrogen @ Siemens Siemens Hydrogen Solutions,
2020]. 2017; 〈[Link]
Linquip, 2022. A complete guide to solar panel output. https:// Siemens-FME-Industrie-en-Energie-2017_H2.pdf〉, [ Accessed
[Link]/blog/a-complete-guide-to-solar-panel- 26-Feb-2021].
output/, [ Accessed 06-February-2022]. SP, 2021. Siemens Energy targets $1.50/kg renewable hydrogen
Lu, L., Sun, X., Ma, J., Yang, D., Wu, H., Zhang, B., Zhang, J., Han, cost by 2025.〈[Link]
B., 2018. Highly efficient electroreduction of CO2 to methanol insights/latest-news/electric-power/032221-siemens-energy-
on palladium-copper bimetallic aerogels. Angew. Chem. Int. targets-150kg-renewable-hydrogen-cost-by-2025〉, [21-mar-
Ed. 57, 14149–14153. 2021].
Mar, H.S., 2009. Catalytic Coversion of Methanol/Dimethylether Statistica, 2021. Global ethylene production capacity.〈https://
to Light Olefins Over Microporous Silicoaluminousphosphates [Link]/statistics/1067372/
Catalysts (Master thesis). National University of Singapore. global-ethylene-production-capacity/〉, [ Accessed 22-March-
McKinsey&Company, 2012. Using microeconomics to guide in­ 2021].
vestmens in petrochemicals. McKinsey on Chemicals. 4, Towler, G., Sinnott, R., 2013. Chemical Engineering Design.
47–58. Elsevier.
Minoux, D., Nesterenko, N., Vermeiren, W., Donk, S.V., Grasso, G., Van Den Bussche, K., Froment, G., 1996. A steady-state kinetic
2009. Process to make olefins from ethanol. [Link] model for methanol synthesis and the water gas shift reaction
[Link]/patent/WO2009098267A1/en. on a commercial Cu/ZnO/Al2O3 catalyst. J. Catal. 161, 1–10.
Office of Energy Efficiency & Renewable Energy, 2020. DOE tech­ Verma, S., Lu, X., Ma, S., Maseld, R.I., Kenis, P.J.A., 2016. The effect
nical targets for hydrogen production from electrolysis. of electrolyte composition on the electroreduction of CO2 to
〈[Link] CO on Ag based gas diffusion electrodes. Phys. Chem. Chem.
targets-hydrogen-production-electrolysis〉, Accessed 22- Phys. 10.
April-2020. Vora, B., Marker, T., Barger, P., Nilsen, H., Kvisle, S., Fuglerud, T.,
Padurean, A., Cormos, C.-C., Agachi, P.-S., 2012. Pre-combustion 1997. In natural gas conversion IV. In: de Pontes, M., Espinoza,
carbon dioxide capture by gas-liquid absorption for integrated R., Nicolaides, C., Scholtz, J., Scurrell, M. (Eds.), Studies in
gasification combined cycle power plants. Int. J. Greenh. Gas Surface Science and Catalysis 107. Elsevier, pp. 87–98.
Control 7, 1–11. Ycharts, US Producer Price Index: Industrial Gas Manufacturing:
Palma, V., Meloni, E., Ruocco, C., Martino, M., Ricca, A., 2018. State Oxygen.2019; 〈[Link]
of the Art of Conventional Reactors for Methanol Production. price_index_industrial_gas_manufacturing_oxygen_yearly〉, [
Methanol: Science and Engineering. Elsevier, pp. 29–51. Accessed 05-March-2020].
[Link] Zhang, W., Hu, Y., Ma, L., Zhu, G., Wang, Y., Xue, X., Chen, R.,
Pappijn, C., Ruitenbeek, M., Reyniers, M.-F., Van Geem, K., 2020. Yang, S., Jin, Z., 2017. Progress and perspective of electro­
Challenges and opportunities of carbon capture and catalytic CO2 reduction for renewable carbonaceous fuels and
chemicals. Adv. Sci. 5, 1700275.