Solid Oxide Electrolysis:

A Technology Status
Assessment

Lead Author:
Gniewomir Flis, Hydrogen and Clean Energy Technologies Consultant

Contributing Author:
Ghassan Wakim, Hydrogen Technology Director

November 2023
 Executive Summary
This report provides a summary assessment of the of SOECs over other electrolyzer architectures, such
state of solid oxide electrolysis technology. In contrast as polymer electrolyte membrane (PEM), alkaline, or
to other, more widely commercialized electrolyzer anion exchange membrane (AEM), is predicated on the
options, such as polymer membrane or alkaline cells, ability to access an external source of process heat.
Without an external source of heat for steam generation,
solid oxide electrolyzer cells (SOECs) operate at much
the potential 20% efficiency advantage of solid oxide
higher temperatures. This feature confers potential
systems over future competing electrolyzer architectures
efficiency advantages, which are of interest in light of largely disappears.
strong expected future demand for clean hydrogen
4. High-temperature operation is a double edged sword:
production using clean electricity powered electrolysis.
it increases electrolyzer efficiency on the one hand but
But high-temperature operation and other features of
due to thermal stresses increases the probability of
SOEC systems also pose challenges. Several key points accelerated stack failure on the other. New materials,
emerge from this assessment: advances in manufacturing techniques, and growing
1. The maturity of solid oxide electrolysis technology is manufacturer and operator experience, including with
solid oxide fuel cells, a closely related technology, have
underestimated. But SOEC manufacturers still have
led to vastly improved durability.
a critical hurdle to clear on the path to successful
commercialization: scaling their product offerings 5. Wider SOEC deployment has been held back, not only by
from small modules, with capacity in the single-digit the perception that the technology is immature, but also
megawatts, to large systems with capacities well into because demand for electrolyzers in general has been
the hundreds of megawatts. limited until recently. Responding to growing interest in
decarbonization options, particularly for hard to electrify
2. The capability to manufacture SOECs is not a bottleneck
sectors, some solid oxide technology vendors have
to the wider deployment of this technology – in fact,
begun adding an electrolyzer product to their offerings.
gigawatt-scale manufacturing could be built up in 18 to
Industrial customers that are already using hydrogen,
36 months. As with any technology, supply chains need
including chemical plant and refinery operators, are
to be managed prudently, but there is no shortage of
actively looking at SOECs. Solid oxide electrolysis is also
raw materials.
of strong potential interest to up-and-coming companies
3. The types of industrial facilities that are well suited for that are pursuing synthetic fuel production and new
integration with SOEC include ammonia, chemical, and nuclear energy technologies, but these potential partners
steel plants, as well as refineries. The chief advantage have yet to establish a substantial commercial presence.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 2

 Table of Contents
Executive Summary..................................................................................................2

1 Introduction..............................................................................................................4

2 An Overview of High Temperature Electrolysis and SOEC Technology................ 7

2.1 Electrolysis Basics......................................................................................................... 7

2.2 Fundamentals of High Temperature Electrolysis.......................................................... 7

2.3 Basic Elements of Solid Oxide Electrolyzer Cells and Systems.................................... 9

3 Technology advantages and challenges ................................................................17

3.1 The Advantages of Operating an Electrolyzer at High Temperature........................... 17

3.2 Technology Challenges for High Temperature Electrolysis........................................ 20

3.3 Strategies for Maximizing Stack Durability................................................................ 22

3.4 Comparing SOEC with Other Electrolyzer Architectures ......................................... 25

4 Assessing SOEC Technology Readiness Based on Flagship Deployments................... 27

4.1 Overview Summary of Applications........................................................................... 33

5 Market Considerations and Commercialization Challenges................................35

5.1 An Overview of SOEC Manufacturers and Technology Developers.......................... 35

5.2 Manufacturing Cost Outlook...................................................................................... 39

5.3 Operations and Maintenance Costs........................................................................... 44

5.4 Key Input Materials and Potential Supply Chain Bottlenecks .................................... 45

5.5 Non-Technological Risks to SOEC Scaleup ............................................................... 46

6 Conclusion and Outlook......................................................................................... 47

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 3

 SECTION 1

Introduction
Hydrogen has been widely discussed as an option for to the achievement of net-zero goals, two challenges
decarbonizing sectors where direct electrification or must be overcome: hydrogen production must be scaled
other low-carbon options might not be practical, or up dramatically, and it must be decarbonized at the
even feasible at all. Examples of such sectors include same time.
the chemical and fertilizer industries, steelmaking,
synthetic fuel production for long-distance shipping Two main pathways for decarbonizing hydrogen
and aviation, and long-term energy storage. For these production are technologically available and
sectors, hydrogen offers several potential advantages: scalable today:
It can be produced with low emissions from a variety of 1. Production from fossil fuel feedstocks with carbon capture
feedstocks using a variety of methods, it emits no carbon (at CO2 capture rates greater than 90%) and strict
dioxide (CO2) at the point of use, and it is versatile and upstream methane controls.
can be used in a range of applications. 2. Production using electrolysis powered with low-carbon
electricity.
Today, hydrogen is already produced on a large scale,
primarily for use as an industrial feedstock in fertilizer The term “electrolysis” describes a process in which an
production and to refine oil products, among other electrical current is used to split molecules. Theoretically,
applications. Decarbonization efforts, however, could this can be done with many types of molecules, but water
substantially increase global demand for hydrogen. electrolysis – in which water molecules are split into their
The International Energy Agency (IEA) has estimated that constituent oxygen and hydrogen elements – is the focus
future hydrogen demand could soar more than five-fold, of this report and of most efforts to develop electrolysis
from approximately 90 million tonnes per year at present technology. Water electrolysis can deliver low-emissions
to as much as 500 million tonnes per year by 2050. hydrogen if powered entirely with low-carbon electricity.
In addition to its carbon-reduction benefits, some
Unfortunately, virtually all hydrogen produced today – more countries see this hydrogen production pathway as
than 99% – is made from fossil fuels or uses fossil fuel an opportunity to break their dependence on fossil
energy inputs, without carbon abatement. As a result, it fuel imports. Electrolysis-based, low-carbon hydrogen
is relatively carbon intensive. For hydrogen to contribute production could also provide system integration

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 4

 Figure 1: Global Demand for Hydrogen by Sector in IEA’s Net Zero Scenario, 2020–20501

500

Other
400
Power generation

H2 based fuels
300
Mt H2

Transport

Industry
200
Refining

100

0
2022 2030 2035 2040 2045 2050

Figure 2: Technologies for Hydrogen Production Including Low- or Zero-Carbon Pathways

Steam Methane Electrolysis +

Steam Methane Reforming
Reforming + CCUS Low-Carbon Electricity

OIL, COAL, NATURAL GAS OIL, COAL, NATURAL GAS RENEWABLES NUCLEAR

CO2

Carbon Intensive Low Carbon Zero Carbon, Zero Carbon,

H2 produced at oil refineries= but more expensive but currently can be economically
10.5 tons CO2 per ton of H2 more expensive competitive due to
baseload production + H2
Geographic
production efficiencies
Limitations

Oxygen(O2)
CO2 Natural
Reforming
Gas H2 Electricity
Natural with CCS Electrolysis H2
(CH4)
Gas Reforming H2 Heat
(CH4) (for SOEC)
CO2
Water(H2O)

1
IEA (2021). Net Zero by 2050: A Roadmap for the Global Energy Sector. https://www.iea.org/reports/net-zero-by-2050

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 5

 services2 for power grids that are heavily dependent on
intermittent renewable generators. The main drawback
of electrolysis is that current mainstream technologies
require a lot of electricity to break apart water. Normally,
around 30%–40% of input electricity is wasted as heat.

Among the lesser-known electrolyzer technologies,

high-temperature architectures are unique in that they
leverage external sources of heat to increase the electrical
efficiency of electrolysis. Ideally, such high temperature
electrolyzers could be paired with applications that
generate a lot of high-grade process heat, including
industrial processes such as steelmaking and refining, or
nuclear and geothermal electricity generation.

However, high temperature electrolysis does not feature

prominently in press announcements of new hydrogen
projects. The technology is widely considered to be at
the R&D stage, with significant technological challenges
to be overcome before it is ready for commercialization.
Yet, net-zero commitments and government policies,
particularly those of the European Union, require massive
scale-up of electrolytic capacity now.3

With limited time to cut carbon emissions and avoid the

worst impacts of climate change, decisionmakers face
pressure to promote technologies that are commercially
available. This report provides an update on the status
of solid oxide electrolyzer cell (SOEC) technology, which
has been the focus of development efforts for high
temperature electrolysis.

The remainder of this report is organized as follows:

Section 2 begins with a general overview of high
temperature electrolysis and SOEC. Section 3 discusses
the advantages and disadvantages of SOEC relative to
other electrolyzer technologies, highlighting current
challenges and avenues for improvement. Section 4
summarizes available information on the technology
readiness level (TRL) of SOEC systems based on
early deployments in different industrial applications.
Section 5 discusses market considerations and hurdles
to commercial deployment. Section 6 concludes with
suggestions for further policy support and investment.

2
For example, the use of excess wind and solar generation to power electrolysis facilities during periods of high wind and solar availability
could help stabilize the grid and create storable energy (in the form of hydrogen fuel) that could potentially be used during periods of low
renewables availability.
3
Odenweller, A., Ueckerdt, F., Nemet, G.F. et al. (2022). Probabilistic feasibility space of scaling up green hydrogen supply. Nat Energy 7,
854–865. https://www.nature.com/articles/s41560-022-01097-4

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 6

 SECTION 2

An Overview of High Temperature

Electrolysis and SOEC Technology

2.1 Electrolysis Basics 2.2 Fundamentals of High

As noted in the introduction, electrolysis involves using Temperature Electrolysis
an electric current to split molecules in a device called The terms “high temperature” and “solid oxide”
an electrolyzer. Electrolysis can be performed on many electrolysis are often used interchangeably for the
types of molecules, but the focus of this report is water simple reason that solid oxide electrolyzers are
electrolysis. Every cell in a water electrolyzer shares the uniquely capable of operation – indeed, they require
same three basic components: operation – at temperatures between 500 and 1000
■ A positively charged anode where oxidation takes place. degrees Celsius (°C). Electrolyzing water (steam) at a
very high temperature confers two advantages:
■ A negatively charged cathode where a reduction to
hydrogen takes place. 1. Intrinsic fast-reaction kinetics and better conductivity
■ An electrolyte that conducts ions between electrodes. lead to efficiencies approaching 100%.6
2. The option of leveraging external sources of heat (in lieu
Anodes, cathodes, and electrolytes can be built from
of using internal electrical energy) creates opportunities
a wide array of materials, but certain combinations to attain electrical efficiencies exceeding 100%.6
outperform others.4 Table 1 describes the four main
electrolyzer architectures that are on the market today.5

4
Two more electrolyzer architectures are considered emerging: membrane-less electrolysis, developed by companies like CPH2 or
Supercritical, and E-T developed by H2Pro.
5
Renewable Hydrogen: Modular Concepts from Production over Storage to the Consumer - Scientific Figure on ResearchGate. Available from:
https://www.researchgate.net/figure/Comparison-of-the-electrolysis-technologies-PEMEL-AEMEL-AEL-and-HTEL-based-on-11_fig2_349090866
6
Elaborated further in Section 3.1 of this report.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 7

 Table 1: Summary of Main Electrolyzer Chemistries

Electrolysis PEM – Polymer AEM –Anion Exchange

Alkaline HT – High Temperature
Technology Electrolyte Membrane Membrane

Electrolyte Acidic Alkaline Alkaline O2 or H+ conducting

Solid (polymer) Solid (polymer) Liquid Solid (ceramic)

Cathode - + Anode Cathode - + Anode Cathode - + Anode Cathode - + Anode

1/2O2 H2 H2 H2
H2 H+ OH- O2- 1/2O2
OH- 1/2O2 1/2O2
H2O H2O
H2O H2O
Ni/ Ni-YSZ/
Pt/C Ir/IrO2 Ni Ni/Co/Fe Ni Ni/Co/Fe Ceramics Solid O2- Cermet
Membrane Membrane Diaphragm / Conducter
Nafion® KOH
Fumapem® FFA

Operating
50°C–80°C 40°C–80°C 60°C–95°C 500°C–900°C
Temperature

Leading
Manufacturers

Figure 3: Possible Combinations of Reactants and Products in Solid Oxide Electrolyzers7

Principle Conversion SOEC

SOEC (O/H - SOEC) Reactant Single Cell Product

e- + H+ + O2 H2O CO2 O2
e+ Anode
e- + O2 O2-
H2O CO

Clean O2- H+ Electrolyte N2 H2

Electricity
CH4 NO
e- + H2O H2O + O2-
e- Cathode
e- + H+ H2 C2H5 C2H4
Stack

Application

100W 103W 104W 105W 106W 107W

7
Zheng, Y., Chen, Z. & Zhang, J. (2021). Solid Oxide Electrolysis of H2O and CO2 to Produce Hydrogen and Low-Carbon Fuels. Electrochem.
Energ. Rev. 4, 508–517. https://doi.org/10.1007/s41918-021-00097-4

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 8

 While the focus of this report is on water electrolysis, it is ■ A cathode where water is reduced to hydrogen. The
worth mentioning that SOECs can be used to electrolyze cathode is fashioned out of perovskite or perovskite-like9
other combinations of molecules so long as at least oxides to make it ionically conductive and coated with a
one of these molecules, owing to the ion-conducting nickel catalyst to make it electrically conductive.
properties of the solid oxide electrolyte, can donate an ■ The eponymous ceramic, solid oxide electrolyte. The
oxygen or a hydrogen ion. A notable combination would electrolyte provides ionic conductivity for the oxygen
be steam and carbon dioxide, which, when electrolyzed, ions to cross while blocking gas diffusion from the
produces a mixture of carbon monoxide, hydrogen, and cathode. It must be electrically insulating, mechanically
steam. This mixture, commonly referred to as syngas, can strong, and capable of withstanding high temperatures.
Doping, which is a term for adding small impurities to a
be further processed into synthetic fuels. Figure 3 shows
semiconductor, is used to mitigate cracking caused by
combinations of potential reactants and their products.
thermal expansion.
Despite its advantages, SOEC is often portrayed as a ■ An anode where the oxide will donate electrons to
much less mature technology compared to polymer produce oxygen. Anodes must be ionically conductive to
(PEM) and alkaline electrolyzers. Low stack power and receive oxide ions. They are usually made from perovskite
high operating temperature, which in turn requires or perovskite-like oxides.
more ancillary equipment to operate the electrolyzer, The typical solid oxide electrolyzer cell will have power
are widely viewed as the main drawbacks of SOEC measured in tens or hundreds of watts, due to its small
technology. SOEC systems are also considered to have a surface area. Small surface area makes it easier maintain
shorter operating life due to thermal stress. cell quality when SOECs are manufactured en masse,
and to distribute heat evenly across the stack during
The largest SOEC systems installed to date range
operation. To raise power per unit area, multiple cells
between 100 kilowatts (kW) and 1 megawatt (MW) in
are placed on top of each other and linked using
size. Most have been installed as pilot or demonstration
interconnects to form an electrolyzer stack. Glass,
projects, and thus do not represent commercial
glass-ceramics, metallic solders, or compressive gaskets
deployments.8 However, judging the commercial
are used to ensure a gas-tight seal between the cells.10
readiness of SOEC based on these projects would
not do the technology justice: SOECs are practically Stacks can come in different shapes. Figure 5 shows an
identical in design and manufacturing to solid oxide fuel example of a square solid oxide fuel cell stack. While
cells (SOFC), which have already been deployed at well SOFCs are distinct from SOECs, the visualization largely
above gigawatt (GW) scale in backup power generation applies to a solid oxide electrolyzer as well, with the key
and microgrid applications. Thus, learnings from SOFC difference that cathode and anode are inverted, and the
deployments and manufacturing experience could be chemical reactions proceed in the opposite direction.
expected to translate to SOEC. Perhaps unsurprisingly,
many companies that are active in SOFC development Other common stack designs include rectangular,
have recently expressed interest in manufacturing hexagonal, or even disc shapes, and reflect different
SOECs as demand for electrolyzers has ballooned. approaches to manufacturing. However, even in a stack
configuration, solid oxide technology will typically have
a power rating at least one order or, in some cases, two
2.3 Basic Elements of Solid Oxide
orders of magnitude less than competing technologies.
Electrolyzer Cells and Systems Thus, while a generic solid oxide stack today might be
rated at 10 kW, a typical PEM stack would be rated in
Figure 4 illustrates the basic elements of a unit solid
hundreds of kW and an alkaline stack in thousands of
oxide cell, which include:
kW. Solid oxide stacks will be connected into modules
■ A metal separator, also known as an interconnector, on to achieve cumulative power ratings in the range of
both sides of the cell. The separator conducts current. hundreds of kW.

8
More information on pilot and demonstration projects is provided in Section 4 of this report.
9
That is, with the same crystal lattice structure as perovskite, or calcium titanate (CaTiO3).
10
Ivanova, M. E., et al. (2023). Technological Pathways to produce compressed and highly pure hydrogen from solar power.
Angewandte Chemie International Edition, 62(32). https://doi.org/10.1002/anie.202218850

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 9

 Figure 4: Cross Section of a Planar SOEC Stack11

10% H2 + 90% H2O 82% H2 + 18% H2O

Anode stream

H2
Interconnect Porous cathode, nickel cermet
H2/H2O → Cathode stream → H2/H2O 4e- 2H2O + 4e → 2H2 + 2O2-
H2O
Unit
Solid structure
cell
~ 1 mm
N2/O2 → Anode stream → N2/O2
2O2- Gastight electrolyte, YSZ or ScSZ
Interconnect

O2- 2O2 → O2 + 4e-

z Cathode stream Porous anode, perovskite, e.g., LSM
x

Figure 5: (A) Square Solid Oxide Stack; (B) Upsizing the Power of a Solid Oxide System by Stacking Cells and
Combining Stacks into Modules12

Notes: Prioritize the rest of the set before starting on this one
A

Power density 10 x 10 cell 5 cells 1 stack = 50 cells 1 module = 10 stacks

= ~2 W/cm2 (~200 W) 10 cm 10 cm
10 cm
10 cm
10 cm

100W 1 kw 10 kw 100 kw

11
Idaho National Lab. 2019. Evaluation of Hydrogen Production Feasibility for a Light Water Reactor in the Midwest.
https://inldigitallibrary.inl.gov/sites/sti/sti/Sort_18785.pdf
12
Wachsman, E. D., & Lee, K. T. (2011). Lowering the Temperature of Solid Oxide Fuel Cells. Science. https://www.science.org/doi/10.1126/
science.1204090

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 10

 Other supporting components and auxiliary systems ■ A dryer to ensure the hydrogen output is completely pure.
needed to operate a working SOEC system are known ■ Compressors for hydrogen and oxygen.
as ‘balance of plant’ (BoP). For example, SOEC systems
■ Instrumentation to monitor and control the performance
require a top up heater to bring incoming steam to
of the electrolyzer.
the stack’s operating temperature and ensure that no
temperature gradients form that might damage the A simplified diagram of a solid oxide system is shown in
stack. As the electrolysis reaction proceeds at high Figure 6.
temperature, high-grade heat can be recovered and
recycled with the use of heat exchangers. All of this will Figure 7 shows an experimental setup of a SOEC module
typically take place inside a ‘hot box’ which also houses housing a single 16 kW stack by FuelCell Energy. The
the array of SOEC stacks. picture includes BoP components like the power and
controls cabinet and the vaporizer outside the stack ‘hot box.’
Other generic BoP elements include:
To put the size of a solid oxide electrolyzer in context,
■ A water purifier to ensure deionised water. a comparison to similar, containerized turnkey alkaline,
■ A water feed and recycle pumps. PEM, and AEM products is helpful. Generally, the
■ A water vaporiser to turn water into steam. rough power of the electrolyzer can be predicted from
its physical dimensions and operating current density.
■ A hydrogen/steam separator (as hydrogen gas is
Accordingly, PEM electrolyzers would be expected to be
produced at the same electrode where steam is fed,
the most compact, delivering about twice the power of
a mixture of steam and hydrogen is produced which
AEM, alkaline, or solid oxide electrolyzers for the same
requires separation).
footprint. For instance, the images in Figure 8 show that
■ Power electronics such as DC/DC converters, rectifiers,
while a 2.5 MW PEM electrolyzer would fit in a 40-ft
and transformers.
container, the same container could house only 1 MW
■ Heat exchangers/recuperators to recycle heat, including AEM or alkaline electrolyzers, or a 720-kW SOEC system.
from external sources.

Figure 6: Diagram of an Electrolyzer, Complete with Balance-of-Plant Components13

Inside the module

Gas/Water
H2 Heater
Separator

O2 Vented out

Feed liquid Pre- Pre-

water supply heater heater
Steam Air

Evaporator
Stack
Hydrogen
Oxygen
Water

Transformer Rectifier

13
International Renewable Energy Agency. (2020). Green Hydrogen Cost Reduction: Scaling Up Electrolysers to Meet the 1.5C Climate Goal.
https://www.irena.org/-/media/Files/IRENA/Agency/Publication/2020/Dec/IRENA_Green_hydrogen_cost_2020.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 11

 Figure 7: FuelCell Energy Single-Stack, 16-kW SOEC Module14

SEC Electrolyzer System SOEC Electrolyzer Module Mid Size CSA Stack Stack Module Assembled

Power and Vaporizer

Controls Cabinet

SOEC Stack Module:

Enclosure vessel is designed in accordance with ASME
125 psig (8.6 barg) Accommodates 1x150-cell stack
B&PV Code Section VIll Div. Il, with internal insulation
design pressure or 4x45-cell stacks with adapter
to allow a touch-safe vessel wall temperature

Figure 8: Turnkey Containerized Products from Electrolyzer Manufacturers

(A) Nel’s 1.25–2.5 MW PEM,15 (B) Cummins 1-MW Alkaline,16 C) A Rendering of Enapter’s AEM 1-MW Multicore,17 and (D) Sunfire’s 720 kW SOEC18

A NEL HYDROGEN PEM ELECTROLYZERS B

M series
1.25-2.5 MW solution

Larger
Installations
5-25 MW
solution

C D

14
Ghezel-Ayaagh, H. (2021). Performance Improvements for Reversible Solid Oxide Fuel Cell Systems [PowerPoint slides].
Fuelcellenergy. https://netl.doe.gov/sites/default/files/netl-file/21SOFC_Ghezel-Ayagh17.pdf
15
Ouimet, R. J., Glenn, J. R., De Porcellinis, D., Motz, A. R., Carmo, M., & Ayers, K. E. (2022). The role of electrocatalysts in the development
of gigawatt-scale PEM electrolyzers. ACS Catalysis, 12(10), 6159–6171. https://doi.org/10.1021/acscatal.2c00570
16
Cummins Inc. (2021). Hydrogen: The Next Generation. Columbus, IN; Cummins Inc. Retrieved from
https://www.cummins.com/sites/default/files/2021-08/cummins-hydrogen-generation-brochure-20210603.pdf
17
# AEM Multicore. Enapter Handbook. (n.d.). https://handbook.enapter.com/electrolyser/aem_multicore/aem_multicore.html
18
Grinhy2.0. SALCOS®. (n.d.). https://salcos.salzgitter-ag.com/en/grinhy-20.html

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 12

 Of course, not all customers will buy turnkey At the end of the day, the physical footprint of an
containerized systems. Developers of larger projects electrolyzer system will depend on the interplay of the
(100s to 1000s of MWs) will probably opt for larger cell’s power density, the stack-module size, and the
stack-module sizes. For instance, John Cockerill’s 5-MW overall plant design. Nevertheless, since PEM technology
stack occupies most of the container, leaving insufficient has intrinsically high cell density, PEM plants are
space for balance of plant, which might have to be generally likely to be the most compact. In comparison,
shipped separately. When the whole system is integrated electrolyzer systems that rely on AEM, alkaline, and solid
at its destination, it could occupy a similar space as two oxide technologies will generally be twice as big.
standalone containerized PEM products while delivering
the same amount of hydrogen. Should there be a need to upsize a solid oxide system
to gigawatt (GW) scale, modules can be connected in
Upsizing the power of a single solid oxide stack is parallel to form a larger, more powerful block that shares
difficult because a larger cell area creates thermal some system-level BoP components, such as middle- and
distribution problems. However, solid oxide stacks can low-temperature heat recuperators, compressors, water
be connected into larger arrays – all this requires is a and hydrogen purification and storage systems, and
larger hotbox (Figure 10). power electronics.

Figure 9: John Cockerill’s 5 MW-Alkaline Stack (right) Connected to its Balance of Plant

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 13

 Figure 10: FuelCell Energy’s Solid Oxide Stacks can be Bundled Together Inside a Single Hotbox Module19

SOFC Stack
Power Generation Stack Module
Only runs in power generation
mode on a wide range of fuels,
including natural gas, biofuels,
propane, and hydrogen 200kW Power Generation System

17” Electrolysis Stack Module

Produces hydrogen from
steam with power input
Electrolysis
4,000 kg/day H2 from 7.3 MW

Energy Storage Stack Module

Alternates between power
generation on hydrogen fuel and
7 kW DC Power Generation electrolysis producing hydrogen
36 kW DC / 25 kg H2/day electrolysis from water
350 cells, 17" height Energy Storage System
1MW & MWh

Figure 11: System Design of a GW-Scale Solid Oxide Electrolyzer Plant20

System Block Stack-Module

(1 GWe = 40 Blocks) (25 Modules) (1 MWe)

Purification & H2 Recycle

H2
Conditioning Compressor

Water Knockout Drum

Cooling
H2 Cooler
Water External MT Feed
LT Feed Heat Recuperator HT Feed Recuperator
Water Recycle Pump
Recuperator
Purification
H2O
& Storage
Water Combined Feed Topping 750C,
Feed Pump Feed Heater Heater 5bar
Stacks
HT Air 750C,
Air Sweep Feed Compressor Recuperator 5bar
Air
LT Air Recuperator
Air Topping Heater
O2 Enriched Air
Air Sweep Outlet Expander

19
Ghezel-Ayaagh, H. (2021). Performance Improvements for Reversible Solid Oxide Fuel Cell Systems [PowerPoint slides].
Fuelcellenergy. https://netl.doe.gov/sites/default/files/netl-file/21SOFC_Ghezel-Ayagh17.pdf
20
James, B.D., Prosser, J.H., & Das S. (2022) HTE Stack Manufacturing Cost Analysis [PowerPoint slides]. Strategic Analysis.
https://www.energy.gov/sites/default/files/2022-03/HTE%20Workshop-Strategic%20Analysis.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 14

 Block sizes may vary initially as manufacturers and While there are no technical barriers to making ever
developers figure out the optimal size, but they will larger blocks, there are trade-offs between BoP costs
likely end up in the tens of MW up to 100 MW. and redundancy. Figure 13 illustrates three possible
For instance, FuelCell Energy has proposed a GW- configurations of hot-boxes and BoP for the same
scale system composed of twenty 50-MW blocks, with nameplate capacity.
each block composed of two 25-MW stack-modules,
housing hundreds of stacks (Figure 12).

Figure 12: FuelCell Energy’s Design for a GW-Scale System for Converting Off-Peak Nuclear Power to Hydrogen21

50 MW Electrolysis Unit

148’ - 0”

Process connections Hot water, hydrogen

Process interconnection services
25 MW Electrolysis BoP
304’ - 0” 25 MW Electrolysis Module

Figure 13: Scaled-Up Designs of SOEC Hot-Boxes with Balance of Plant Elements for the Heating System:
(A) 1 Hot-Box of 1.5 MW with 16 of 94 kW Stacks and Integrated Heat BoP,
(B) 4 Hot-Boxes of 0.4 MW (i.e. 4 of 94 kW Stacks Per Hotbox) with Integrated Heat BoP, and
(C) 4 Hot-Boxes of 0.4 MW (i.e. 4 of 94 kW Stacks) with Separate Stack-Box and Heat BoP-Box22

A B C

Reactants
Reactants
steam
Reactants
steam

Key

HEX1
Stack
HEX2 Power Reactants

21
(2022) FuelCell Energy Platforms for Hydrogen Production [PowerPoint Slides]. FuelCell Energy.
https://www.energy.gov/sites/default/files/2022-03/Bulk%20Storage%20Workshop_Day1_06.pdf
22
Institute for Sustainable Process Technology. (2023). Next Level Solid Oxide Electrolysis.
https://ispt.eu/media/20230508-FINAL-SOE-public-report-ISPT.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 15

 The Science of High-Temperature Electrolysis
Solid oxide electrolyzers are also often referred to as high-temperature electrolyzers because they
are unique in operating at temperatures between 500°C and 1000°C. High-temperature operation is a
requirement of the ionic conductivity of a solid oxide electrolyte. Every electrolyte has its own activation
energy that needs to be reached for optimal ionic conduction. Since the temperature of an object is an
average of the energy of its constituent particles, an object might exhibit some conductivity even at
temperatures below the activation energy, but this conductivity will be weak. The gain in reaction kinetics
is fastest at the point where temperature matches activation energy. Adding more heat will improve
conductivity, but with diminishing returns.

Higher temperature thus leads to higher conductivity, which translates into higher current density at a
given voltage, as seen in Figure 14. Higher current density decreases equipment costs and the physical
footprint of the system as fewer stacks and balance-of-plant components are needed to produce the
same quantity of hydrogen. However, improvements in conductivity must be weighed against practical
considerations such as the durability of the electrolyzer. From a durability standpoint, high-temperature
operation results in more thermal stress, leading to faster degradation and increasing the probability of
component failure.

Solid oxide electrolyzers will usually be designed with a temperature sweet spot that optimizes for stack
lifetime and performance. Different manufacturers will target different sweet spots depending on their
choice of materials and stack design. Some systems will work optimally at 600°C, while others will be
designed to operate at 700°C or even above 800°C. Aside from choice of material, the thickness of the
solid electrolyte will influence the optimal operating temperature: thicker electrolytes are less conductive
and will require higher temperatures to match the conductivity of thinner electrolytes.

Figure 14: Cell Voltage vs. Current Density of a Solid Oxide Cell23

1.8
600ºC
1.6 550ºC
500ºC
1.4
Cell Voltage (V)

1.2
SOFC Mode
1.0
0.8
0.6
0.4
0.2
-2.5 -2.0 -1.5 -1.0 -0.5 -0.0 0.5 1.0

Current Density (Acm-2)

23
Wu, W., Ding, H., Zhang, Y., Ding, Y., Katiyar, P., Majumdar, P. K., He, T., & Ding, D. (2018). 3D self-architectured steam electrode enabled
efficient and durable hydrogen production in a proton-conducting solid oxide electrolysis cell at temperatures lower than 600 °C.
Advanced Science, 5(11). https://doi.org/10.1002/advs.201800360

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 16

 SECTION 3

Technology Advantages
and Challenges
3.1 The Advantages of Operating an low-grade heat, especially steam at 100°C–150°C,
in the condensate system.
Electrolyzer at High Temperature
A solid oxide electrolyzer that is importing steam from an
High-temperature operation enables a step change
industrial source is thus at the very least 14% more efficient
in efficiency because the electrolyzer will be fed with
than any alkaline, PEM, or AEM competitor. In practice, the
water in the form of steam. Thus, an external source of
efficiency gap today is even larger – around 25% – because
heat above 100°C can be used to evaporate the water,
of imperfect operation of PEM, alkaline, or AEM systems.
effectively relieving the electrolyzer stack from having
However, continuous improvements in alkaline and PEM
to provide the latent heat of evaporation. In terms of the
technologies will likely reduce losses from imperfect
electricity required to produce hydrogen, this results
operation by 2030s, leaving solid oxide with only the
in an efficiency gain of 0.5 kilowatt-hours per normal
efficiency advantage of importing steam.
cubic meter (kWh/Nm3) of hydrogen, or 6 kWh per
kilogram (kg) of hydrogen.24 The ability to make use SOEC operation coupled with steam imports has
of imported steam is particularly relevant to industrial already been demonstrated at a sub-megawatt scale by
processes, which typically discard large amounts of FuelCell Energy,25 as well as by Bloom Energy,26 with the

24
Note how this is equal to the difference between the higher heating value of hydrogen (39.4 kWh/kg) and the lower heating value of
hydrogen (33.3 kWh/kg).
25
Venkataraman, V. (2021). High Temperature Solid Oxide Electrolyzer [PowerPoint Slides]. Bloom Energy.
https://www.energy.gov/sites/default/files/2021-09/h2-shot-summit-panel1-hte-industrial-panel_0.pdf
26
Idaho National Lab and Bloom Energy Produce Hydrogen at Record-Setting Efficiencies. (2022, August 9). Bloom Energy. Retrieved from
https://www.bloomenergy.com/news/idaho-national-lab-and-bloom-energy-produce-hydrogen-at-record-setting-efficiencies/.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 17

 In endothermic operation the electrical efficiency of the
Table 2: Typical Stack Energy Consumption for reaction at the cell level can be understood as exceeding
Different Electrolyzer Architectures 100% (HHV basis). This is because the electrolyzer would
be producing more in hydrogen energy than it would be
Alkaline28 PEM27 AEM29 SOEC30 consuming in electrical energy.

Stack (DC) 47 – 66 47 – 51.5 – 34 Altogether, leveraging a high-grade waste heat source

energy 66 66 at 600°C to its maximum, that is to evaporate water and
consumption heat the stack, could theoretically lower the electricity
(kWh/kg H2)
required to produce hydrogen to around 2.7kWh/Nm3 or
30 kWh/kg, which would be equivalent to 131% efficiency
at the cell level for the higher heating value (HHV) of
former achieving 38.7 kWh of electricity consumption at hydrogen or 111% cell efficiency for the lower heating value
the stack level (DC) to produce a kilogram of hydrogen (LHV) of hydrogen. Stack efficiencies, excluding balance
(H2) and the latter reaching a record low of 37.7 kWh/ of plant, will be only marginally less than cell efficiencies
kg of H2, also at stack level. The heat source was a due to non-ideal fields and flow (e.g., eddy currents).
simulated27 nuclear plant.
However, endothermic operation is challenging, since
Another advantage of high-temperature operation is distributing external high-grade heat within the stack
the possibility for further integration with heat imports causes temperature gradients that in turn cause thermal
to enable operation in endothermic mode (meaning stress. Moreover, industrial processes that simply discard
that the process absorbs rather than releases heat). high-grade heat are extremely rare, since this heat can
By default, a solid oxide electrolyzer will operate in be used to do mechanical work.
‘thermoneutral voltage’ – i.e., the voltage at which the
internal resistance of the stack components creates While the stack is typically the most power-hungry
just enough heat to reach and maintain optimal kinetics element of an electrolyzer by a large margin, the
through the solid electrolyte. This voltage is around overall efficiency of the system (also known as system
1.29 Volts (V) compared with 1.48 V for alkaline, PEM, efficiency or energy demand) also depends on energy
or AEM electrolytes. Since the theoretical minimum consumption by other BoP components. Figure 16, from
potential difference for water electrolysis is 1.23 V, solid the company FuelCell Energy, shows a flow diagram
oxide electrolyzers will generally be more efficient than for an SOEC electrolyzer coupled with a nuclear power
competing architectures at the stack level. plant. In this flow diagram, BoP energy consumption is
estimated at 4 kWh per kg of hydrogen, which means
If there is a source of high-grade heat, the steam that that with steam imports, the company hopes to use
is destined for electrolysis can be heated to or near the as little as 38 kWh of electrical energy to produce a
operating temperature of the electrolyzer. This allows kilogram of H2. Given hydrogen’s higher heating value
the electrolyzer to draw in heat from its surroundings, of 39.4 kWh/kg, FuelCell Energy’s system, under these
which results in endothermic operation. Depending conditions, would achieve greater than ‘100% electrical
on the temperature of the steam and the temperature efficiency (HHV)’ even on a system level.
of the electrolyzer, endothermic operation could
theoretically reduce electricity demand by another
0.5 kWh/Nm3 of H2 (Figure 15).

27
Simulated in this case means that the input steam was heated to a temperature comparable to the output of a nuclear reactor using an
external heat source.
28
Ivanova, M. E., et al. (2023). Technological Pathways to produce compressed and highly pure hydrogen from solar power.
Angewandte Chemie International Edition, 62(32). https://doi.org/10.1002/anie.202218850
29
What is the overall efficiency of Enapter’s electrolyser? Enapter. (2023, January 9).
https://www.enapter.com/newsroom/kb_post/what-is-the-overall-efficiency-of-enapters-electrolyser
30
Institute for Sustainable Process Technology. (2023). Next Level Solid Oxide Electrolysis.
https://ispt.eu/media/20230508-FINAL-SOE-public-report-ISPT.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 18

 This example underlines how solid oxide technology is This makes solid oxide electrolyzers an ideal candidate
unique in being able to leverage external sources of heat for system integration with industrial sources of process
to minimize electricity consumption. or waste heat.

Figure 15: Energy Demand in Water Electrolysis as a Function of Temperature, Excluding Balance of Plant31

Energy Demand, kWh/Nm3 of H2 Energy Demand Volts

4.5
1.8
Current SOEC efficiency gain due to better 4.0
1.6
kinetics and conductivity Total energy
3.5 Evaporation
demand
1.4
Exclusive SOEC efficiency gain if importing steam Heat
3.0
1.2
Heat demand
2.5 1.0

2.0 0.8

Total Energy Demand, △ƒH 1.5 0.6

Electricity demand
Current alkaline, OEN & AEM technologies 1.0 0.4
Liquid

Future potential alkaline, PEM & AEM technologies 0.5 0.2

Gas

SOEC 0.0 0.0

0 200 400 600 800 1000
Temperature, °C

Figure 16: FuelCell Energy’s Flow Diagram of an SOEC Electrolyzer Coupled with a Nuclear Power Plant32

38 kWh per kg hydrogen

>100% HHV efficiency
34 kWh per kg
hydrogen

4 kWh per kg
Nuclear Electrical Power hydrogen

Oxygen

Heat BoP Auxilary Power

Hydrogen

Water Boiler Steam

Solid Oxide
Electrolysis Stacks

31
All About SOEC For PtX – TOPSOE
32
Leo, T. (2023, January 12). Low-cost hydrogen production from nuclear energy. FuellCell Energy.
https://www.fuelcellenergy.com/blog/low-cost-hydrogen-production-from-nuclear-energy

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 19

 Still, compensating with increased temperature does
Table 3: Energy Consumption of Specific Components add to overall thermal stress, and increases the risk of
of a Solid Oxide Electrolyzer System33 stack failure – most commonly as a result of electrolyte
cracking or seal breakage, which opens the possibility of
gas crossover, and which demands immediate shutdown
Component kWh/kgH2
and replacement. At higher temperature, the operator
Electric heating feed 1.2 might not observe much performance degradation
for several years, until one day, the stack fails.
Electric heating air 1.8 Manufacturers typically have enough test data to be able
to quantify and guarantee stack lifetime, provided the
H2 compressor 2.2
stack is operated within given parameters.
SOEC stack (DC) 34.6
Nevertheless, most modern solid oxide stacks will
Total LHV (with steam import) 39.8 fail before PEM or alkaline stacks. High temperature
operation is thus a double-edged sword: it increases
Steam generation (if required) 7.9 efficiency but increases the probability of accelerated
Total HHV (with internal steam generation) 47.7
stack failure. Figure 18 summarizes the main mechanisms
of SOEC degradation caused by high temperature
operation. The most consequential are nickel migration
and nickel sintering. Both lead to gradual loss of
Results from another feasibility study that estimated
active sites on the anode side and increase electrolyte
the energy consumption of BoP sub-component are
resistance, which is the main driver of the observed
summarized in Table 3.
1% degradation rate.

3.2 Technology Challenges for As already noted, interconnects and seals are also
susceptible to thermal stress. For instance, metallic
High Temperature Electrolysis interconnects and piping can require special coatings,
One of the biggest hurdles to commercializing solid oxide sometimes involving cobalt, to decrease the risk
technology is shaking its reputation for poor durability. of chromium evaporation which could poison the
A literature review that includes data up to 2020 put the electrode. The need to prevent seals from failing has
average degradation rate for SOEC systems at under also been a major obstacle to SOEC commercialization.
1% per 1,000 hours of operation. Taking performance at The problem is that seals and solid oxide cells have
80% of nameplate capacity as the industry cut-off for different thermal coefficients, which can cause
stack replacement, a 1% degradation rate would imply a mechanical failures as a temperature gradient develops
stack lifetime of around 2.5 years at full load, which, while across the stack. Temperature gradients can be
up from less than half a year a decade ago, is still four to prevented by heating the very stack slowly, or by simply
eight times shorter than the typical expected lifetimes for maintaining it at a constant temperature. Thus, solid
PEM and alkaline technologies. oxide systems will typically require several hours for
a cold start, so that all system elements heat evenly to
However, extrapolating electrolyzer lifetime purely from the requisite temperature. The problem of temperature
observed degradation rates can be misleading. Solid gradients also explains why solid oxide systems degrade
oxide stacks can compensate for loss of performance more during frequent cycling. This can make it much
by operating at a higher temperature and thereby preferable to keep the stack in a hot ‘standby’ mode.
increasing their conductivity. The increased energy Compared with PEM or alkaline systems, solid oxide
demand required to supply the extra heat will not exceed systems require a lower minimal load (3% for SOEC vs.
an extra kWh per kg H2 produced. Higher temperatures more than 10% for PEM/ALK/AEM) to be on standby.
increase viable running hours beyond what degradation
rates would suggest without sacrificing meaningful
performance. Most solid oxide manufacturers build this
functionality within their stacks.

33
Institute for Sustainable Process Technology. (2023). Next Level Solid Oxide Electrolysis.
https://ispt.eu/media/20230508-FINAL-SOE-public-report-ISPT.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 20

 Figure 17: Historical Improvements to Solid Oxide Electrolysis
(A) Cumulative Stack Test Duration, (B) Observed Degradation Rate and (C) Largest Plant Size by Year34

A B C

H2O electrolysis 10 H2 demonstration plants

CO2 electrolysis 10
CO plants (commercial)

Degradation in stacks (%V/1000 hours)

co-electrolysis

Plant nameplate capacity, MW

cumulative
Stack test duration (1000 hours)

20 8
test time 8

Cumulative test time (years)

Huntington Beach (CA, USA)

15 6 6

Salzgitter (Germany)
Leuna (Germany)

Mountain View (CA, USA)

Columbus (OH, USA)

Rotterdam (Netherlands)
Salzgitter (Germany)

Dayton (OH, USA)

10 4 4

Foulum (Denmark)
5 2 2

0 0 0
09 11 13 15 17 19 09 11 13 15 17 19 2015 2017 2020 2021 2023
Year

Figure 18: Known Mechanisms of SOEC Degradation35

1 Cathode degradation 3 GDC barrier

Ni coarsening degradation
/sintering Gd exsolution
Separator Plate / Interconnect
Ni oxidation New Gd
Ni mobility H2O / H2 → H2O / H2 Flow → H2O / H2 containing phases
90-95%/10-5% 10-20%/90-80%
H2O H2 Coarsening
Leads to reduced
triple point dispersion Porous Fuel Interfacial void
and activity and Electrode - Ni/YSZ formation
2e-
delamination from 2e- H2O + 2e- → H2 +O2-
electrolyte; Results in
decreased e- conduction Power
AC/DC O2- Gastight Electrolye YSZ
inverter
Barrier Layer GDC
2e -
Oxygen Electrode
2e- LSCF-SDC
2 Dense electrolyte 4 Anode
O2- → 1/2O2 + 2e-
degradation degradation
1/2O2
YSZ composition O2 Sweep Gas Air Interfacial void
0-100% ← Air / O2 Flow ←
changes formation
Zr reduction/ Separator Plate / Interconnect Sr reaction with
NiZr alloying Zr, SrZrO3
Results in decreased Lead to increased
O= ion conduction pO2 & delamination

34
Hauch, A., Küngas, R., Blennow, P., Hansen, A. B., Hansen, J. B., Mathiesen, B. V., & Mogensen, M. B. (2020a). Recent advances in solid
oxide cell technology for electrolysis. Science, 370(6513). https://doi.org/10.1126/science.aba6118
35
Hydrogen from Next-generation Electrolyzers of Water (H2NEW). H2NEW. (n.d.). https://h2new.energy.gov/home

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 21

 Finally, solid oxide electrolyzers, much like PEM, AEM, Interestingly, relevant experience for successful SOEC
and alkaline electrolyzers, require desalinated and manufacture and operation does not necessarily need
deionised water, as some impurities can poison the to come from solid oxide electrolyzers specifically.
catalysts. SOECs are particularly prone to poisoning by Because solid oxide fuel cells are practically identical
sulphur and silicon, as well as by amines, ammonia, and in design to electrolyzers, learnings from SOFC
oxides of nitrogen (NOx). However, quality control of manufacturing and operation can be translated to SOEC
BoP systems, and the addition of a water deioniser will manufacturing. The major differences between fuel
significantly reduce the risk of poisoning. Most cell- cells and electrolyzers occur at the balance-of-plant
level degradation mechanisms can be largely avoided configuration, rather than at the cell level. Therefore,
by assuring quality control during manufacturing or by while the current academic literature implies an average
operating the stack within design parameters. operating life of 2.5 years for solid oxide electrolyzers,
experience from the historical deployment of solid oxide
fuel cells suggests that solid oxide stacks can operate as
3.3 Strategies for Maximizing
long as 80,000 hours, or nearly 9 years at full capacity.
Stack Durability
Operating parameters other than temperature will also
Technology choices and manufacturing experience play a role in longevity. Higher operating pressure, for
can extend the longevity of SOEC stacks. Selecting example, is helpful for kinetics and mass transport in the
materials that minimize the activation energy of the electrodes. However, at pressures greater than 3 barg,
electrolyte, for example, will allow for lower operating the performance improvement is small. Meanwhile, high
temperatures, which in turn ensures that the electrolyzer pressure can cause mechanical problems, particularly to
experiences less thermal stress to start with and the thin and brittle electrolyte.36 This is why the output
enables the use of cheaper materials. Thus, minimizing mixture of hydrogen and steam produced by most solid
activation energy confers a technological advantage. oxide systems leaves the electrolyzer at a pressure of only
Since operating temperatures for most SOEC materials 0–2 barg, thus necessitating further pressurization to be
are still above 500°C, however, this approach offers, at suitable for most industrial applications. Using an external
most, incremental gains rather than a step change in compressor to pressurize the hydrogen to 30 barg will
electrolyzer efficiency and durability. increase the energy demand of the overall system by
around 2 kWh per kg of hydrogen output; compression to
Operating and manufacturing experience will
200 barg roughly doubles that requirement, to around 4
also greatly impact stack durability. More data on
kWh per kg of hydrogen (Figure 19).
electrolyzer performance under real-world conditions
will give manufacturers a better idea of when and To sum up, the operating life of a solid oxide stack
under what conditions a stack might be expected to will depend on the experience that goes into its design
fail. Manufacturers can either address these problems and manufacture, choices about technology and
through design iteration or by providing preventive materials, and the operator’s ability to stay within
maintenance. For any chemical plant, an unplanned design limits. A conservative assumption would put the
outage will be infinitely more costly than a planned average life of a solid oxide electrolyzer stack today at
stack replacement. Since SOEC manufacturers are 2.5 years; an optimistic assumption would be 5 years.
aware of their technology’s shorter stack lifespan By 2030 or before, it would be fair to expect the
compared with alkaline and PEM technologies, they technology to achieve an expected operating life of
design their stack-modules to be easily accessible in 7–8 years at full load.
case of failure. A ‘hot-box’ can be replaced in as little
as an hour after it has cooled.

36
Zhang, M., Wang, E., Mao, J., Wang, H., Ouyang, M., & Hu, H. (2022). Performance analysis of a metal-supported intermediate-temperature
solid oxide electrolysis cell. Frontiers in Energy Research, 10. https://doi.org/10.3389/fenrg.2022.888787

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 22

 Figure 19: Energy Losses for the Multi-Stage Mechanical Compression of Hydrogen37

18%

Pipeline

Underground Storage

Buses/ trucks

Passenger vehicles

Tanks
Compression losses as a fraction of hydrogen LHV (%)

16%

14%

12%

10%

8%
Max
6%
Min
4%

2%

0%

0 200 400 600 800 1000

Pressure (bar)

37
Institute for Sustainable Process Technology. (2023). Next Level Solid Oxide Electrolysis.
https://ispt.eu/media/20230508-FINAL-SOE-public-report-ISPT.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 23

 Next Generation Solid Oxide Technologies
Proton conducting solid oxide electrolytes. To this point, the discussion has focused on oxygen ion
conducting solid oxide electrolytes (O-SOEC). However, there are also solid oxides that can conduct
protons (H-SOEC). These ionic conductors exhibit higher ion conductivity at much lower temperatures
(350°C–600°C), reflecting the relatively low activation energy required for proton migration. Lower operating
temperature means less thermal stress, which addresses many of the challenges faced by O-SOEC.

However, the mechanisms behind H-SOEC are still poorly understood: many studies note significant
current leakage, and the electrical efficiencies of H-SOEC designs are still far below those of O-SOECs.38

New materials and manufacturing approaches. Some SOEC start-ups have chosen to completely reinvent
the process for manufacturing the solid oxide stack. With the help of new materials and dry power pressing
additive methods, they claim to be able to produce a monolithic, seal-free design with internal gas routing
channels. This innovation allows for a marked increase in the operating pressure range and consequently
the power density of the cell, while also removing potential failure mechanisms associated with seals
and interconnects.

Novel ionically conductive materials. An up-and-coming company called Advanced Ionics claims to
have created a hybrid design which conducts oxygen ions between electrodes but does not feature a
brittle ceramic membrane. This innovation allows the electrolyzer to operate at a lower temperature
range (200°C–600°C) than with a conventional solid oxide. Lower operating temperature translates
into less thermal stress, all the while improving the efficiency of integration with an external source of
low-temperature process heat. This is important because most industrial facilities operate at temperatures
well below the range needed for conventional SOEC operation (Figure 20).

Figure 20: Operating Temperature Ranges of Electrolyzers and the Typical Process Temperature
of Various Industrial and Power Generation Processes39

Steel

Desulfurization Cracking

Urea and Nitric Ammonia

Nuclear
Methanol
Fischer Tropsch

Alakaline & PEM Electrolyzers Advanced Ionics Solid Oxide Electrolyzers

0 100 200 300 400 500 600 700 800

Temperature (ºC)

38
Lei, L., Zhang, J., Yuan, Z., Liu, J., Ni, M., & Chen, F. (2019). Progress report on Proton conducting solid oxide electrolysis cells.
Advanced Functional Materials, 29(37). https://doi.org/10.1002/adfm.201903805
39
Source: Advanced Ionics pitch deck.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 24

 3.4 Comparing SOEC with other electrolyzer architectures
Table 4 summarizes the key advantages and drawbacks of solid oxide electrolyzers relative to other prominent electrolyzer
architectures, specifically: alkaline, polymer electrolyte membrane (PEM), and anion exchange membrane (AEM).
Table 5 summarizes operational characteristics.

Table 4: Advantages and Drawbacks of SOEC Relative to other Types of Electrolyzers

Alkaline PEM AEM Solid Oxide

Advantages

■ Mature technology with GWs ■ High-power densities ■ No expensive or ■ Lowest standby load
of historical deployment exceeding 2A/cm2, enabling critical minerals
■ Intrinsic high efficiency due
compact designs and small
■ Multi-MW stacks enable ■ Compact designs and to fast kinetics and high
physical footprint
systems with large capacities small footprint like PEM conductivity at elevated
already today ■ Fast cold start-up temperatures
■ Suitable for high pressure
time and fast load
■ Potential to use (>30 barg) operation ■ Option to leverage external
changing capabilities
earth abundant and sources of heat to attain
■ Potential to use
inexpensive materials ■ Relatively mature, with MWs stack electrical efficiencies
earth abundant and
over 100%
■ High intrinsic product inexpensive materials
gas purity ■ Some designs can operate
reversibly as fuel cells
■ Can exceed its rated power
and electrolyzers
for brief periods giving a
broader range of flexibility ■ Can electrolyse
carbon-based molecules
and co-electrolyse them
with water
■ Potential to use
earth abundant and
inexpensive materials

Disadvantages

■ High material effort on ■ Use of expensive materials ■ Available stack sizes are ■ Highest physical footprint,
system level by using highly as titanium and critical in the low kWs, driving up 1/3rd more than alkaline and
alkaline liquid as electrolyte platinum group metals balance of plant costs. AEM and 3.5x more than
(PGM) on cell level PEM
■ Low power densities and ■ Shorter stack lifetimes than
large footprint ■ Long-term stability needs to for alkaline and PEM due to ■ High operating
be proven at MW scale membrane degradation temperatures cause
■ Additional effort for gas
thermally accelerated aging
purity required ■ Current generation reliant
on ‘forever chemicals’ such ■ Produced hydrogen needs
■ Slow cold start-up time
as PFAS which don’t occur in to be separated from steam
■ High minimal load due to nature, and hardly degrade and purified
gas permeability of the in the natural environment
■ Low output pressure
membrane
requiring external
compression
■ Cold starts take several
hours
■ Historical reputation for
poor durability

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 25

 Table 5: Operational Characteristics of Different Electrolyzer Architectures40

Technology Alkaline PEM AEM SOEC

Development stage >1 GW deployed >100 MW deployed <10 MW deployed >1 GW deployed
(fuel cell)

<10 MW deployed
(electrolyser)

Electrolyte Liquid: 25 – 40% KOH Solid: proton exchange Liquid-solid hybrid: Solid: ceramic –
membrane (Nafion) 1% KOH/anion zirconia or ceria based
exchange membrane

Operating 70 – 90 50 – 80 40 – 80 500 – 900

temperature (°C)

Operating pressure (bar) Conventional: Up to 80 Up to 35 with 0 – 2 bar

atmospheric (350 among startups) potential for much
higher in the future
Modern: up to 30 bar
(50 among startups)

Typical current densities 0.4 – 1.0 0.2 – 4.0 0.2 – 2.0 0.5 – 1.5
today (A/cm2)

System energy 50 – 78 50 – 83 57 – 69 38 (with steam import)

consumption (kWh/kg H2) 48 (without steam import)

Stack cost (2020 $/kW) 270 – 450 400 – 870 200 250 – 2,000

Electrolyzer system cost 800 – 1,500 1,400 – 2,100 3,333 917 – 4,000
with BoP (2020 $/kW)

Stack lifetime 60,000 – 100,000 50,000 – 90,000 5,000 – 40,000 20,000 – 50,000
(full load hours)

Degradation 0.13 0.25 0.4 0.55 – 1%

(% / 1000 hours)

Ramp up time hot idle 60s 10s 30 minutes 10 minutes

to nominal power

Cold ramp up time 30 – 60 minutes 5 minutes 20 minutes >600 minutes

Minimum load 10 – 40% 5 – 10% 10 – 20 >3%

40
Zheng, Y., Chen, Z. & Zhang, J. (2021). Solid Oxide Electrolysis of H2O and CO2 to Produce Hydrogen and Low-Carbon Fuels.
Electrochem. Energ. Rev. 4, 508–517. https://doi.org/10.1007/s41918-021-00097-4 ; Wang, J. (2020, October 20). AEM water electrolysis:
How it works. Enapter. https://www.enapter.com/newsroom/aem-water-electrolysis-how-it-works ; European Commission, Joint
Research Centre, Davies, J., Dolci, F., Weidner, E. (2021). Historical analysis of FCH 2 JU electrolyser projects : evaluation of contributions
towards advancing the state of the art, Publications Office. https://data.europa.eu/doi/10.2760/951902 ; Pem Electrolyser. Nel Hydrogen.
(2022, December 16). https://nelhydrogen.com/product/m-series-3/ ; The Oxford Institute for Energy Studies. (2022). Cost-competitive
green hydrogen: how to lower the cost of electrolysers?. https://www.oxfordenergy.org/wpcms/wp-content/uploads/2022/01/Cost-
competitive-green-hydrogen-how-to-lower-the-cost-of-electrolysers-EL47.pdf ; Department of Energy. (2016). Hydrogen Production Cost
from Solid Oxide Electrolysis. https://www.hydrogen.energy.gov/docs/hydrogenprogramlibraries/pdfs/16014_h2_production_cost_solid_
oxide_electrolysis.pdf ; Wang, L., Chen, M., Küngas, R., Lin, T.-E., Diethelm, S., Maréchal, F., &amp; Van herle, J. (2019). Power-to-fuels
via solid-oxide electrolyzer: Operating window and Techno-Economics. Renewable and Sustainable Energy Reviews, 110, 174–187. https://
doi.org/10.1016/j.rser.2019.04.071 ; Institute for Sustainable Process Technology. (2023). Next Level Solid Oxide Electrolysis. https://ispt.
eu/media/20230508-FINAL-SOE-public-report-ISPT.pdf ; Fraunhofer ISE. (2021). Cost Forecast for Low-Temperature Electrolysis. https://
www.ise.fraunhofer.de/content/dam/ise/de/documents/presseinformationen/2022/2021-11-17_CATF_Report_Electrolysis_final.pdf ;
Sunfire. (n.d.). Renewable Hydrogen for Industrial Applications Sunfire-Hylink SOEC. Dresden, Germany; Sunfire GmbH. Retrieved from
https://www.sunfire.de/files/sunfire/images/content/Sunfire.de%20(neu)/Sunfire-Factsheet-HyLink-SOEC-20210303.pdf.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 26

 SECTION 4

Assessing SOEC Technology Readiness

Based on Flagship deployments
This section reviews recent deployments of SOEC maturity of a new technology. Originally developed by the
technology in various industrial applications. Experience National Aeronautics and Space Administration (NASA)
from these deployments is helpful in assessing the and now widely used by the U.S. Department of Energy
technology readiness of solid oxide electrolysis as (DOE) and other government agencies, the TRL system
an option for hydrogen production. Throughout, the assigns a number, on a scale of 1 to 9, based on the
discussion references “technology readiness level” development milestones a new technology has reached.41
(TRL), which is a metric commonly used to describe the

41
For a description of how different U.S. government agencies assign TRLs, see: https://www.gao.gov/assets/gao-20-48g.pdf.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 27

 Chemical Industry – Figure 21: Topsoe eCOs™ Electrolyzers at a
Carbon Monoxide Production DeLille Oxygen Co. Site42,43
TRL: 8
Since 2020, U.S.-based DeLille Oxygen Company
has used two commercially available, 750-kW
Topsoe eCOs™ units for the on-site production of
‘clean’ carbon monoxide via solid oxide electrolysis.

Chemical Industry – Figure 22: One of Sunfire’s SOEC Modules

Hydrogen Production at the MultiPLHY Site45
TRL: 7
Sunfire began testing a 1-MW electrolyzer for use in
methanol synthesis at the Hydrogen Lab Leuna, in
Germany, in 2021. Various operational studies were
carried out at the lab, in close cooperation with Sunfire,
to evaluate likely system performance when connected
to an intermittent renewable electricity supply.44

Another MW-scale Sunfire project, MultiPLHY, involves

the first commercial-scale use of high temperature
solid oxide electrolyzers at Neste’s biofuel refinery in
Rotterdam. A 2.6-MW electrolyzer was installed in
April 2023 and was scheduled to start operations by the
end of 2023. If this project meets performance targets
by the time it concludes (expected by end of 2024),
the Sunfire technology should reach TRL 8.

42
Ravn, S. (2019, May 29). Delille Oxygen Co. leases two ECOSTM Units for cost-competitive onsite co production. Topsoe.
https://www.topsoe.com/blog/delille-oxygen-co.-leases-two-ecos-units-for-cost-competitive-onsite-co-production
43
Wix Christian (2022) Solid Oxide Electrolysis Explained [PowerPoint Slides]. TOPSOE.
https://fortesmedia.com/files/files/Doc_Pack/Hydrogen_%26_P2X/Christian_Wix_Topsoe.pdf
44
Junghans, U., & Rasch, C. (2023, August 8). E-CO2Met – Electricity & CO2 to methanol. Fraunhofer Center for Chemical-Biotechnological
Processes CBP. https://www.igb.fraunhofer.de/content/cbp/en/reference-projects/e-co2met.html
45
European Commission. (2021). Multimegawatt high-temperature electrolyser to generate green hydrogen for production of high-quality
biofuels. https://doi.org/10.3030/875123

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 28

 Steelmaking Figure 23: The GrinHy 2.0 720-kW Sunfire Electrolyzer
TRL: 7 at Salzgitter’s Site47

The benefits of coupling SOEC with a steel plant are

similar to those of coupling SOEC with a nuclear plant,
in that industrial process heat can be reused in the SOEC
to increase the overall electrical efficiency of hydrogen
production. In this configuration, a SOEC system will
require less low-carbon electricity than PEM or alkaline
electrolyzers to produce the same quantity of hydrogen.

In the first large-scale project to test this concept, called

GrinHy 1.0, the German company Salzgitter AG installed
but has been further integrated into Salzgitter’s
a 140-kW Sunfire SOEC system at one of its steelmaking
steelmaking operations in that it supplies hydrogen
plants. The project proved that heat from a steel
for the annealing process. By the end of 2022, stack
plant could be used with a solid oxide electrolyzer to
performance had been proved for 20,000 hours of
produce hydrogen at the purity required for low-carbon
operation, producing a total of around 100 tons of high
steelmaking. In the first phase of the GrinHy project,
purity ‘green’ hydrogen at an electrical efficiency of
however, hydrogen was turned back to electricity using
minimum 84% (measured at the lower heating value
the reversibility of solid oxide technology, rather than
(LHV) of hydrogen), and proving the technology at
being used in steelmaking.46
TRL level 7. In the next phase, the electrolyzer will supply
GrinHy 1.0 was succeeded by GrinHy 2.0, which involves hydrogen for the reduction of iron ore, which will reduce
a substantially larger, 720-kW SOEC system. This CO2 emissions from this step of the steelmaking process
prototype also recycles process heat from steelmaking by 95% and prove the technology at TRL 8.

46
European Commission. (2016). Green Industrial Hydrogen via Reversible High-Temperature Electrolysis. https://doi.org/10.3030/700300
47
Grinhy2.0. SALCOS®. (n.d.). https://salcos.salzgitter-ag.com/en/grinhy-20.html

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 29

 Synthetic Fuels Figure 24: Sunfire Co-Electrolyzer Module as Part of
TRL: 5 – 6 the Kopernikus PtX Project

Because a solid oxide electrolyte can conduct oxygen

ions, SOEC technology can be used to electrolyze
molecules such as CO2. The resulting product in this case
would be carbon monoxide (CO).

When carbonated water is fed to a solid oxide

electrolyzer, both water and CO2 can be electrolyzed
simultaneously – this process is called co-electrolysis.
Co-electrolysis produces a mixture of carbon monoxide
and hydrogen, also known as synthetic gas or ‘syngas’,
alongside steam and CO2. Syngas can be used to
produce a range of synthetic hydrocarbons, including
fuels and materials.

Sunfire has demonstrated a 150-kW co-electrolyzer for applications. The project, called C2 Fuels is deploying a
methanol production through its SynLink project, putting small solid oxide system, with electrical capacity in the
the technology at TLR 6.48 In parallel, Sunfire has also single-digit kWs, in Dunkirk, France.49
validated a 10-kW module to produce synthetic fuels
through its Kopernikus PtX project. The next challenge for co-electrolysis technology is
demonstration on a MW scale. To that end, the European
Another solid oxide cell manufacturer, Estonia-based Union is co-funding an ongoing project, called MegaSyn,
Elcogen, has partnered with French energy company to demonstrate syngas production by co-electrolysis
Engie SA on a pilot project to produce dimethyl ether in an industrial environment. The aim is to lift the solid
(DME), a synthetic alternative to diesel fuel for transport oxide co-electrolysis technology to TRL 7 by 2025.50

48
SynLink – Synthetic E-fuels as key enabler for sector linking. Fraunhofer Center for Chemical-Biotechnological Processes CBP.
(2023, August 9). https://www.cbp.fraunhofer.de/en/reference-projects/synlink.html
49
Elcogen takes a new step for c2fuel! | C2FUEL. (n.d.). https://c2fuel-project.eu/elcogen-takes-a-new-step-for-c2fuel/ ; Lehtinen, T., &
Noponen, M. (2021). Solid oxide electrolyser demonstrator development at Elcogen. ECS Meeting Abstracts, MA2021-03(1), 285–285.
https://doi.org/10.1149/ma2021-031285mtgabs
50
Hansen, K. V. (n.d.). MEGASYN – Power-to-X Project in MEGAWATT-SCALE. https://www.megasyn.eu.
https://www.megasyn.eu/highlights/megasyn-start

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 30

 Energy Storage Figure 25: FCE rSOFC Targets and Current Capability51
TRL: 5 – 6
Stack Durability Performance
Another practical application of reversible SOEC 40k h 0.22 W/cm2
technology would be clean energy storage: water could >-1.0 W/cm2
be electrolyzed at times when clean sources of power 10k h

are abundant, the resulting clean hydrogen could then Size 200 k W Round Trip
be stored and turned back to electricity using the same >1 MW Efficiency
device at times when clean power supply is low. This 70% AC-AC
High Volume
approach would benefit from lower capital expenditures Projects to
<200 $/kW ~400
and reduced space requirements because it avoids stack
Stack Cost Number of Cycles
the need to purchase and install electrolyzers and fuel 200 $/kW >1,500
cells separately.

Reversible SOEC/SOFC technology has already been

RSOFC is at low TRL building Commercialization Target
demonstrated in large-area stacks. For instance, FuelCell off significant prior SOFC &
Current Capability Range
Energy (FCE) has a reversible SOFC (rSOFC) technology SOEC core technology and
system demonstrations.
that is currently at TRL 5 and expected to reach TRL 6 due
to concurrent work supported by DOE’s National Energy
Technology Laboratory (NETL) (project award number
DE-FE0031974). The aim of this work is to validate and
verify engineering/pilot-scale rSOFC technology in a
Figure 26: Sunfire’s GrinHy 1.0 Reversible SOEC55
relevant environment. FCE has been open about the
outstanding challenges that need to be solved before its
rSOEC technology is commercial, including challenges
with respect to stack-module size, durability, and number
of cycles. In fact, improvements in these three areas are
recurring themes for most SOEC manufacturers.

European firms are also conducting trials of reversible

SOEC: for example, Elcogen is currently testing its
rSOEC technology at Finland’s VTT Technical Research
Centre. This project is due to wrap up in the summer
of 2023, with the aim of reaching TRL 6.52 Sunfire has
already validated a small, 140-kW rSOEC system through
its GrinHy project at the Salzgitter steel plant, which
puts its technology at TRL level 6.53,54

51
Ghezel-Ayagh, Hossein. Reversible Solid Oxide Fuel Cell Systems for Energy Storage and Hydrogen Production. United States.
https://doi.org/10.2172/1874500
52
European Commission. (2023). Reversible solid oxide Electrolyzer and Fuel cell for optimized Local Energy miX.
https://doi.org/10.3030/779577
53
Di Molfetta, R. (2022, April 5). Strategies for regional deployment&nbsp; of hydrogen infrastructure: The case of North Rhine-Westphalia,
Germany. http://kth.diva-portal.org/smash/get/diva2:1652006/FULLTEXT01.pdf
54
Wulf, C., Linßen, J., & Zapp, P. (2018). Review of Power-to-Gas Projects in Europe. Energy Procedia, 155, 367-378.
https://doi.org/10.1016/j.egypro.2018.11.041
55
European Commission. (2019). Green industrial hydrogen production powering Europe along the road to a decarbonised future.
https://doi.org/10.3030/700300

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 31

 Nuclear Actually harnessing nuclear process heat from an
TRL: 3 – 4 existing reactor for electrolysis is a lot more complicated
than conducting a simulation. For instance, tapping into
The benefits of pairing SOEC with a nuclear plant go a boiling water reactor’s steam system would require
both ways: the electrolyzer gets access to high-grade adding a radiation-shielded heat exchanger loop.
process heat (300°C–500°C) while the nuclear plant This could quickly turn into a permitting nightmare,
gains the option to switch between feeding electricity given the regulatory and safety requirements placed
to the grid or using it to produce hydrogen. In a system on nuclear plants, particularly in the United States.
with high penetration of wind and solar generators, However in other regions like the UK regulators are
this optionality would allow the nuclear plant to keep willing to allow small modifications to plants without re-
a steady electricity generation profile, without having licensing. For pressurised water or gas cooled reactors
to ramp up and down in response to fluctuating output which have a shielded loop, coupling with SOEC
from low-cost renewable generation. This could be becomes a little easier, and indeed EDF has embarked
particularly important for integrating nuclear with on such a project. However, EDF has described three
high-renewable-energy systems. big limitations facing similar retrofits:

Researchers at Idaho National Lab (INL) in the United 1. Steam harnessing potential is limited to low-pressure
States have been conducting a variety of tests on Bloom tap-off, significantly limiting the size of the electrolyser to
Energy’s solid oxide electrolyzer at the Dynamic Energy double digit MWs
Testing and Integration Laboratory. These tests include 2. This site had an onsite alkaline electrolyser as part of the
physical steam and load simulations that replicate original design and license
nuclear power station conditions, an important step 3. The electrolyser had to be sited at least several hundred
in validating full compatibility with a nuclear facility. meters from the reactor for safety reasons, limiting the
Early tests at high temperatures and high nuclear plant quality of heat that can be transferred to the SOEC to
availability revealed that Bloom Energy’s electrolyzer about 200C.
could produce hydrogen at 37.7 kWh per kilogram of
hydrogen and with 88.5% electrical efficiency (with Thus, it is unlikely that SOEC will be deployed in 100s
respect to LHV of hydrogen and conversion to direct of MW scale at existing nuclear plants. If we are to see
current (DC)).56 FuelCell Energy has also deployed a any big rollout of SOEC at greenfield nuclear plants, it’s
250- kW electrolysis module at INL (Figure 27). imperative that these plants are designed, engineered,
and licensed with solid oxide electrolyser optionality
from the get-go.

Figure 27: FCE

Electrolyzer
Module at the INL
Testing Facility

56
Idaho National Lab and Bloom Energy Produce Hydrogen at Record-Setting Efficiencies. (2022, August 9). Retrieved from
https://investor.bloomenergy.com/press-releases/press-release-details/2022/Idaho-National-Lab-and-Bloom-Energy-Produce-Hydrogen-at-
Record-Setting-Efficiencies/default.aspx.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 32

 4.1 Overall Summary of Applications TRL status notwithstanding, all SOEC applications still
share at least one hurdle to clear: Even in applications
At first glance, flagship deployments of SOEC where the technology has been demonstrated at a small
technology across different applications show mixed commercial scale, the challenge of engineering systems
results: For applications in the chemical industry, SOEC at the hundreds-of-megawatts scale.
has recently been demonstrated at a commercially
relevant scale. In steelmaking, the technology is nearly
there. Further work is needed to prove the commercial
viability of co-electrolysis for energy storage and
synthetic fuel production, especially with regard to
issues of durability and scale.

The Final Frontier: Upscaling SOEC to Meet the Requirements

of Industrial Users
The path to upscaling SOEC is relatively straightforward: cell area and current density need to increase,
and more cells need to be packed into a stack. These improvements would uprate the hotbox to single-digit
megawatts; multiple hotboxes could then be packed into double-digit megawatt blocks (also referred to as
modules in Figure 28). Large blocks would avoid over duplicating balance-of-plant requirements, allowing
the capital costs of SOEC systems to become competitive with those of large PEM and alkaline systems.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 33

 Figure 28: SOEC Scale Up Towards a Multi-MW Plant57

SOE cell SOE stack Hot-box HT BoS

1 - 10 kW → 20 - 100 kW 100 - 500 kW → 1 - 2 MW

SOE cells SOF stacks with 30-60 Stacks and highT balance of stacks
100 → 400 → 800 cm2 → 120 - 150 cells insulated and forming a Hot-box
Current density Amount and stack ratings: → 15-60 stacks
→ 0.6 → 085 → 1.5 A⁄cm2 of few kW→ several stacks with sub-100 kW

Repeating module Complete SO plant

1.8-3 MW → 5-20 MW

Purification & compression

(LP Steam; Storage/ Pipeline;)

Power electronics

Large Plant INFRA

Multiple hotboxes integrated with Multi-MW SO plant
common lowT BoP and power electronics 2-20 MW → >100 MW
forming a repeating module

57
Institute for Sustainable Process Technology. (2023). Next Level Solid Oxide Electrolysis.
https://ispt.eu/media/20230508-FINAL-SOE-public-report-ISPT.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 34

 SECTION 5

Market Considerations and

Commercialization Challenges

5.1 An Overview of SOEC Manufacturers

Figure 29: Global Nameplate Electrolyzer
and Technology Developers
Manufacturing Capacity58
Table 6 on pages 35-37 summarizes information about
current SOEC manufacturers and technology developers, SOEC 2 GW
based on primary research into the state of the industry PEM
and interviews with industry leaders. The table provides 11.1 GW
information on key technology characteristics, R&D
efforts, manufacturing capacities, and upcoming projects.

Total in 2023
32.8 GW

Alkaline
19.7 GW

58
A breakneck growth pivot nears for green hydrogen (2022) BloombergNEF. Available at: https://about.bnef.com/blog/a-breakneck-growth-
pivot-nears-for-green-hydrogen/ (Accessed: 02 November 2023); Clean Air Task Force

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 35

 Table 6: Summary of Key Solid Oxide Technology Developers and Manufacturers

Electrolyzer mfg Upcoming

Brief description Technology
capacity per year59 electrolyzer projects

US based manufacturer of solid Today: 2GW ■ Electrode-supported ■ 10 MW electrolyzer with

oxide systems. Spun-out in 2001 manufacturing. LSB Industries.
from NASA tech to create fuel
■ Built a 2 GW electrolyser/1 GW ■ Westinghouse and Bloom
and oxygen using Martian water.
fuel cell factory in less than Energy signed a letter of
Bloom has deployed over two years. intent to accelerate large-
1 GW of solid oxide fuel cells scale hydrogen production
■ Operational temperature
worldwide and has recently in the nuclear industry.
above 800°C.
returned to providing an ■ Bloom announced plans to
electrolyzer product. They had ■ Current modules of 120 kW
install a 240 kW electrolyzer
recently built, installed, and can be combined to form
at Xcel Energy’s Prairie Island
operationalised the largest 2 MW blocks.
nuclear plant in Minnesota.
solid oxide electrolyzer in the ■ Above-industry average stack Power-on expected in
world – 4 MW – all in a span of lifetimes exceeding 5 years early 2024.60
two months and even reaching 8 years.
■ Bloom is also expected to
■ R&D focussed on increasing provide stacks for a nuclear-
power density and further powered hydrogen micro
automating manufacturing hub in South Korea.61

Electrolyzer mfg Upcoming

Brief description Technology
capacity per year electrolyzer projects

Ceres is a UK based solid Ceres is a technology ■ Gadolinium-doped ceria ■ Ceres had signed an
oxide technology developer. developer and licenses out (GDC) electrolyte allows agreement with Shell to
The company traces its roots its tech for other companies lower temperature operation deliver a 1 MW solid oxide
to research at Imperial College to manufacture. (500°C– 600°C) which electrolyzer demonstrator in
London and has been spun out improves durability and 2023 in Bangalore, India.
in 2001. allows the use of standard
■ Another 1 MW demonstrator
automotive gaskets and
is scheduled for
stainless steel, as the
deployment and testing
temperature is below
by 2024 with Bosch and
the evaporation point of
Linde Engineering.
chromium, a strong poison
to SOEC.
■ The trade-off with using
GDC is lower current density
at a given temperature.
■ Metal supported
manufacturing.

59
Note: solid oxide manufacturing lines for fuel cells and electrolyzers are interchangeable, but electrolytic capacity is roughly twice that of
fuel cell capacity adjusted for efficiency.
60
Xcel Energy and Bloom Energy to Produce Zero-Carbon Hydrogen at Nuclear Facility. (2022, September 19). Retrieved from
https://www.bloomenergy.com/news/xcel-energy-and-bloom-energy-to-produce-zero-carbon-hydrogen-at-nuclear-facility/.
61
Patel, S. (2023, April 27). South Korean companies snap up opportunities to advance floating nuclear, nuclear hydrogen, smrs. POWER Magazine.
https://www.powermag.com/south-korean-companies-snap-up-opportunities-to-advance-floating-nuclear-nuclear-hydrogen-smrs/

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 36

 Table 6: Summary of Key Solid Oxide Technology Developers and Manufacturers – continued

Electrolyzer mfg Upcoming

Brief description Technology
capacity per year electrolyzer projects

Elcogen is an Estonia based Today: 1 to 10 MW ■ Anode supported ■ Elcogen’s system integrator

solid-oxide technology manufacturing, enabling a partner – Convion – is
developer and manufacturer. Soon: 100 to 200 MW thin electrolyte that operates collaborating with Shell to
The company has been at 650°C which improves deploy a 1 MW demonstrator
operating since 2001. conductivity. system out of four 250kW
stacks at Shell’s Energy
■ Current R&D focussed
Transition campus
on automating stack
in Amsterdam.
assembly process – cell
printing already takes
single digit seconds.
■ Focussed on scaling
manufacturing.

Electrolyzer mfg Upcoming

Brief description Technology
capacity per year electrolyzer projects

Originally founded in 1969 Today: 10 to 100 MW ■ Stacks are 95% recyclable. ■ FuelCell Energy has an
Connecticut and called Energy agreement with Malaysia and
GW scale by 2030 ■ Current R&D efforts to
Research Corporation. In 1992 Marine Heavy Engineering
increase stack lifetime to
renamed to FuelCell Energy and Holdings (MHB) to
five years, and beyond.
demonstrated first product – the collaborate on development
molten carbonate fuel cell. As of ■ Opted for disc shape to of large-scale electrolyzers
2011, FuelCell Energy is also in reduce manufacturing in Asia, Australia, and
the business of developing solid costs by repurposing DVD New Zealand.
oxide technology. production machinery.
■ FCE is also partnering
■ Have recently announced with Idaho National Lab to
they are taking commercial explore coupling of SOEC
orders for electrolyzers. with nuclear energy to
leverage process heat for
■ Also working on a reversible
steam imports, and to extend
solid oxide stack product.
the life of nuclear plants.
The 250-kW module is meant
to enter operation in 2023.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 37

 Table 6: Summary of Key Solid Oxide Technology Developers and Manufacturers – continued

Electrolyzer mfg Upcoming

Brief description Technology
capacity per year electrolyzer projects

Sunfire was founded in Today: 1 – 10 MW ■ Electrolyte supported ■ Norsk e-fuels plans

Germany in 2010 and originally manufacturing. to commission a
focussed exclusively on PtL (power to liquids)
■ Operational temperature
solid oxide technologies. plant running on a mixture
of 850°C.
As of 2021, Sunfire had also of alkaline and SOEC
expanded into the alkaline ■ Co-electrolysis capability electrolyzers from
electrolysis market by acquiring is a unique selling point, Sunfire in 2024.62
Industrie Haute Technologie. positioning the company well
for the emerging PtL market.
■ The company is planning for
a pre-commercial 10 MW
pilot to reach TRL level 8 in
the next five years, focussing
on proving longevity of over
10 years.

Electrolyzer mfg Upcoming

Brief description Technology
capacity per year electrolyzer projects

Topsoe is a Danish company 500 MW by 2025 with ■ Anode supported ■ First Ammonia signed
founded in 1940 which optionality to expand to 5 GW manufacturing. a capacity reservation
specialises in catalysis and agreement with Topsoe for
■ Operates at 750°C.
process technology in chemical 500MW with the option
and hydro processing. ■ Current 2nd Generation to increase to 5GW to
lifetime of 4 years. produce green ammonia
Topsoe has been active in as fuel for transportation,
solid oxide development since ■ 3rd Generation coming in
power storage & generation,
2004. From 2014 the company second half of this decade
and fertilizer.63
has focussed on developing will feature improvements
electrolysis systems for the to lifetime, cost, and a
production of hydrogen, new geometry.
synthesis gas, and pure carbon ■ Focused on coupling their
monoxide. Topsoe has 80+ years electrolyzer product to
of experience in catalysis ammonia and methanol
utilized in SOEC development. production, where they
have complementary
expertise through their
catalyst business.

62
Sunfire. (2022, February 28). GREEN KEROSINE FROM 2024: NORSK E-FUEL PRESENTS ITS PLANS. Retrieved from
https://www.sunfire.de/en/news/detail/green-kerosine-from-2024-norsk-e-fuel-presents-its-plans.
63
TOPSOE and first ammonia to launch green ammonia production. F&L Asia. (2022, September 16). https://www.fuelsandlubes.com/
flo-article/topsoe-and-first-ammonia-to-launch-green-ammonia-production/#:~:text=Topsoe%20and%20First%20Ammonia%20
have,storage%20and%20generation%2C%20and%20fertilizer

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 38

 5.2 Manufacturing Cost Outlook Building the Electrolyzer ‘Sandwich’

Before discussing details, it is important to note that Once the substrate has been selected, the layers of the
solid oxide manufacturing will look largely the same electrolyzer ‘sandwich’ can be assembled. This layering
whether the end product is electrolytic cells or fuel cells. of cell elements can be done in several ways, but the
In fact, the manufacturing process, and the machinery it three most common industrial methods are tape casting,
requires, will be shared and switched between the two screen printing, and vapor deposition.
products depending on the order book. Switching over
Vapor deposition is the slowest of these three options
from fuel cell production to electrolytic cell production
because it is often done manually. This method is widely
and vice versa will take less than several hours.
used in early-stage factories and labs because it does not
cause mechanical stress and the capital requirements
First Step: the Substrate are relatively light. In contrast, screen printing and
tape casting are more capital intensive but allow for far
Solid oxide cell manufacturing starts with a substrate,
greater throughput and automation. Still, because screen
on which other cell elements will be deposited in
printing is a batch process, it is typically slower than tape
layers. The substrate can be thought of as the bread
casting, which is a continuous process.
to a sandwich. The choice of substrate will vary by
electrolyzer design and manufacturing method, but Nonetheless, most solid oxide manufacturers currently
the substrate is an integral element of the cell, like the prefer screen printing as it is easier to control and
electrolyte, the electrode, or the interconnect. Older perfect than tape casting. This leads to more consistent
solid oxide designs rely on electrolyte-supported quality which is paramount during high-temperature
manufacturing, while newer designs are electrode- operation when even the slightest imperfections
or even metal-supported. compound. Screen printing is also cheaper and easier
to set up at smaller manufacturing scales (megawatts to
The choice of substrate has implications for the
single-digit gigawatts).
operational characteristics of the cell, and for the
cost of stack manufacturing. Electrolyte-supported Factories might employ a mix of these methods at
manufacturing arose because of the difficulty of different stages of manufacturing. For instance, a factory
depositing thin layers of yttria-stabilized zirconium (YSZ) might tape cast the solid oxide electrolyte substrate,
electrolyte, which typically needs to be laid in thicker and then apply electrodes using screen printing. The
layers (>10 micrometers or μm) to prevent cracking. optimization problem in this respect has different
Newer ceria-based electrolytes can be deposited in solutions depending on the manufacturer’s experience,
thinner layers, which opens the possibility of electrode- throughput, choice of materials, and choice of substrate.
or metal-supported manufacturing.
Figure 31 and Figure 32 show how the process might
Thinner electrolytes mean better conductivity at change between electrolyte- and electrode-supported
lower temperatures, but they also have manufacturing manufacturing.
advantages such as lower material costs and lower heat
input requirements for sintering. Heat treatment steps
are usually the costliest element of the manufacturing
process for SOEC cells and will account for approximately
40% of the stack’s manufacturing cost.64

64
James, B.D., Prosser, J.H., & Das S. (2022) HTE Stack Manufacturing Cost Analysis [PowerPoint slides]. Strategic Analysis.
https://www.energy.gov/sites/default/files/2022-03/HTE%20Workshop-Strategic%20Analysis.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 39

 Figure 30: Substrate Types for Electrolyzer Manufacturing65

Cathode Cathode Cathode Cathode

Electrolyte Cathode Electrolyte Electrolyte
Electrolyte
Electrolyte Anode Anode
Anode
Anode Anode Porous Substrate Interconnect

Electrolyte Supprted Anode Supprted Cathode Supprted Substrate Supprted Metal Supprted
(ES) (AS) (CS) (SS) (MS)

Figure 31: Process Flow Diagram of a ‘Generic’ SOEC Stack Fabrication Using Electrolyte Supported Cells66

Ball Mill Ball Mill Hydrogen Ball Mill Barrier Ball Mill Air INTERONNECT ETCHING
Electrolyte Slurry Electrolyte Slurry Layer Slurry Electrode Slurry
Shear Interconnect Metal
CELL FABRICATION
Apply Photoresist
Tape Cast Electrolyte Screen Print Hydrogen Screen Print Air Electrode
& Illuminate
(in-line drying) Electrode (in-line drying) (in-line drying)

Dissolve in Etching Solution

Singulate Cells Screen Print Air Barrier Fire Air
(stamping) Layer (in-line drying) Electrode
Photo-Electrochemical
Etching

Sinter Electrolyte Fire Hydrogen Electrode Laser Cut

& Barrier Layer Cells Alkaline Wash

Laser Cut Interconnects

Dry Glass Seals Glass Seal Bead Glass Seal Ball Mill Hydrogen Electrode Ball Mill Air Electrode
(batch) Dispensing Slurry Contact Paste Slurry Contact Paste Slurry

Apply MCO to Interconnects (PVD) Synthesize MCO Coating (glycine)

Place Cell

Screen Print A.E. Paste Screen Print H.E. Paste Machine End Plates

Place Interconnect After n repeats

Stack Braxing & QC Final Stack QC Install Stack Hardware

65
Kuterbekov, K. A., Nikonov, A. V., Bekmyrza, K. Zh., Pavzderin, N. B., Kabyshev, A. M., Kubenova, M. M., Kabdrakhimova, G. D., et al. (2022).
Classification of Solid Oxide Fuel Cells. Nanomaterials, 12(7), 1059. MDPI AG. Retrieved from http://dx.doi.org/10.3390/nano12071059
66
James, B.D., Prosser, J.H., & Das S. (2022) HTE Stack Manufacturing Cost Analysis [PowerPoint slides]. Strategic Analysis.
https://www.energy.gov/sites/default/files/2022-03/HTE%20Workshop-Strategic%20Analysis.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 40

 Figure 32: Process Flow Diagram for a ‘Generic’ SOEC Stack Fabrication Using Electrode-Supported Cells

Ball Mill Ball Mill Ball Mill Ball Mill Ball Mill INTERONNECT
Hydrogen Electrode Hydrogen Electrode Electrolyte Barrier Layer Air Electrode ETCHING
Support Slurry Functional Slurry Slurry Slurry Slurry

Shear Interconnect
CELL FABRICATION Metal

Tape Cast Hydrogen Tape Cast Screen Print

Electrode Support Apply Photoresist
Electrolyte Air Electrode
Layer 1 (in-line drying) & Illuminate
(in-line drying) (in-line drying)

Dissolve in
Sinter Hydrogen Etching Solution
Tape Cast Hydrogen Fire Air
Electrode Layers
Electrode Support Electrode
& Electrode
Layer 2 (in-line drying) Photo-
Electrochemical
Etching
Screen Print
Tape Cast Hydrogen Barrier Layer Laser Cut
Electrode Functional (in-line drying) Cells
Layer (in-line drying) Alkaline
Wash

Fire Barrier
Layer Laser Cut
Interconnects

Dry Glass Glass Glass Ball Mill Hydrogen Ball Mill Air
Seals Seal Bead Seal Electrode Contact Electrode Contact
(batch) Dispensing Slurry Paste Slurry Paste Slurry

Apply MCO to Synthesize

Interconnects MCO Coating
(PVD) (glycine)

Place
Cell

Machine
Screen Print Screen Print
End
A.E. Paste H.E. Paste
Plates

Place
After n repeats
Interconnect

Stack Final Install

Braxing Stack Stack
& QC QC Hardware

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 41

 Figure 33: Solid Oxide Cell Fabrication Processes and Robots67

Tape Casting Automated Screen Printing Conditioning & Test Infrastructure

Sintering Automated QC / Stacking

3000 m2 facility with pilot cell / stack production, R&D and extensive 33 test stand capability (Calgary, AB)

Once the electrode or electrolyte layers are printed, Manufacturers will probably deviate slightly from flow
they are sintered in ovens to make a solid oxide cell sheets by, for instance, outsourcing some steps such as
‘sandwich.’ Cells are then joined to make a stack, which interconnect etching. Flow sheets also don’t list all the
involves first coating the cells with a sealing material quality control steps that are taken at factories. These
to prevent gas leakage and then encasing them with a quality controls are critical to ensure that the stack will
current-conducting interconnect to join the cells. not fail before its intended lifetime. Most of these steps
The stack is then pressed together, dried, fitted with in the manufacturing process can be automated, using
any extra hardware like piping, if needed, and subjected analytical methods involving lasers and x-rays
to a final quality control process. and robotics (Figure 33).

67
Ghezel-Ayaagh, H. (2021). Protonic Ceramics for Energy Storage and Electricity Generation Using Ammonia [PowerPoint slides].
Fuelcellenergy. https://www.energy.gov/sites/default/files/2021-08/12-proton-ceramics-using-ammonia.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 42

 Cost Reductions From Scaling Manufacturing inside an insulated hotbox and includes mechanical
equipment (pumps, compressors) as well as electrical
When goods are mass manufactured, their costs balance-of-plant components.
tend to come down, owing to learning-by-doing, and
economies of scale. In the case of solid oxide stacks, A 2016 study by the Technical University of Denmark
scaling production from megawatts to gigawatts drives (DTU) outlined how BoP costs decline with module
drastic reductions in manufacturing and materials costs. size, based on component costs from existing chemical
Resulting savings in total stack cost range from 67% (from plants.69 Today, solid oxide electrolyzers are typically
$355/kW to $115/kW) to 77% (from $335/kW to $78/ packaged and sold in modules of hundreds of kilowatts.
kW) for electrolyte- and electrode-supported processes
respectively (Figure 34). Furthermore, at gigawatt scale, Thus, according to the DTU study, typical BoP costs
electrode-supported stacks are estimated to be one- would be around $688/kW. Other sources put current
third cheaper to manufacture than electrolyte-supported BoP costs at $700/kW70 or even $1,000/kW.71
stacks. Once manufacturing reaches the gigawatt scale,
Nevertheless, BoP costs can fall, driven by economies
the rate of cost savings tapers off.
of scale as BoP is shared by a larger block of SOEC
Costs for stack manufacture, however, are only part of the modules. Estimates of future BoP costs predict a decline
final cost of the electrolyzer. A typical stand-alone solid of 60%–75%.
oxide electrolyzer system also integrates many stacks

Figure 34: SOEC Stack Production Costs by Manufacturing Method68

Electrolyte-supported Cells (EsC) H2-supported Cells (HEsC)

$450 $450 Contingency
Tooling Costs
$400 $400
Labor Costs
$350 $350 Manufacturing Costs (w/o Labor)
Material Costs
$300 $300 Total Cost
Production Cost ($/kW)

Production Cost ($/kW)

$250 $250
$115/kW $78/kW
$200 $200

$150 $150

$100 $100

$50 $50

$0 $0
25

50
75
0
0
0
0
0

1,0 5
00

1,3 0
00

1,6 0
00

2, 0
0

25

50
75
0
0
0
0
0

1,0 5
00

1,3 0
00

1,6 0
00

2, 0
0
82

82
10
20
35
50
65

0
00

10
20
35
50
65

0
00
1,1

1,4

1,1

1,4
1,8

1,8

Yearly Manufacturing Rate (MW/year) Yearly Manufacturing Rate (MW/year)

68
https://www.energy.gov/sites/default/files/2022-03/HTE%20Workshop-Strategic%20Analysis.pdf
69
James, B.D., Prosser, J.H., & Das S. (2022) HTE Stack Manufacturing Cost Analysis [PowerPoint slides]. Strategic Analysis.
https://www.energy.gov/sites/default/files/2022-03/HTE%20Workshop-Strategic%20Analysis.pdf
70
James, B.D., Prosser, J.H., & Das S. (2022) HTE Stack Manufacturing Cost Analysis [PowerPoint slides]. Strategic Analysis.
https://www.energy.gov/sites/default/files/2022-03/HTE%20Workshop-Strategic%20Analysis.pdf
71
FuelCell Energy. (2022). Reversible Solid Oxide Fuel Cell Systems for Energy Storage and Hydrogen Production.
https://www.osti.gov/servlets/purl/1874500

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 43

 Estimates of total direct cost for a solid oxide electrolyzer 5.3 Operations and Maintenance Costs
are scant and vary widely. Table 8 presents estimates
from a leading SOEC manufacturer, Bloom Energy, and Due to their higher operating temperatures and shorter
from the Dutch independent research organization stack lifetimes, SOEC systems tend to have higher
TNO (Netherlands Organisation for Applied Scientific maintenance costs than PEM and alkaline electroyzers.
Research), which recently concluded a cost study with Bloom Energy has developed cost estimates for its own
leading SOEC manufacturers. Of course, developers of SOEC technology and competitor products. Another
SOEC projects would also need to account for additional, study, by Idaho National Laboratory, estimated annual
indirect costs such as engineering, procurement, operations and maintenance (O&M) cost for an SOEC
and construction (EPC) costs and contingency costs. coupled directly to a nuclear plant at $39.5 per kW.
These indirect costs could double the final, installed cost
for an SOEC plant.

Table 7: SOEC Balance of Plant Cost Estimates

2022 – Cost of SOEC BoP (2022 $/kW) 2030 – Cost of SOEC BoP (2022 $/kW)

Source DTU67 Strategic Analysis68 FuelCell Energy69 DTU67 Strategic Analysis68 FuelCell Energy69

Heat Exchanger 66 – – 16 – –

Rectifier 456 – – 115 – –

Compressor 122 – – 30 – –

Other BoP 44 – – 11 – –

Total BoP 68872 700 $1,00073 17274 – 40075

Table 8: Estimates of Electrolyzer System Direct Costs Table 9: O&M Cost Estimates for an SOEC System

Direct System Cost – Electrolyzer Plus Installation ($/kW) O&M Annual Cost Excl. Stack Replacement ($/kW/yr)

Bloom Energy76 ISPT77 Bloom Energy78 INL79

2020 1,302 3,006 Technology ALK PEM SOEC SOEC

2030 359 738 2020 36 61 69 39.5

2030 15 17 17 39.5

72
Based on plant size of 0.1 MW.
73
Based on a manufacturing capacity of 20 MW/year.
74
Based on plant size of 10 MW.
75
Based on a manufacturing capacity of 500 MW/year.
76
Bloom Energy. (2022). The Role of Solid Oxide Technology in the Hydrogen Economy: A Primer.
https://f.hubspotusercontent30.net/hubfs/5242085/BE21_22%20Hydrogen-white-paper_D.pdf
77
Institute for Sustainable Process Technology. (2023). Next Level Solid Oxide Electrolysis.
https://ispt.eu/media/20230508-FINAL-SOE-public-report-ISPT.pdf
78
Bloom Energy. (2022). The Role of Solid Oxide Technology in the Hydrogen Economy: A Primer.
https://f.hubspotusercontent30.net/hubfs/5242085/BE21_22%20Hydrogen-white-paper_D.pdf
79
Idaho National Lab. 2019. Evaluation of Hydrogen Production Feasibility for a Light Water Reactor in the Midwest.
https://inldigitallibrary.inl.gov/sites/sti/sti/Sort_18785.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 44

 5.4 Key Input Materials and Potential only $4.4/kW of the cost to manufacture a SOEC. Since
the mining of nickel is geographically dispersed, overall
Supply Chain Bottlenecks supply chain risk from this element is low.
Input materials are an important driver of manufacturing
An important advantage of SOEC technology is that it
cost for SOECs. Per megawatt, a ‘generic’ SOEC will
avoids the use of platinum and iridium – rare earth metals
require 150–200 kg of nickel, approximately 40 kg of
for which the supply chain is concentrated and known
zirconium, approximately 20 kg of lanthanum, and up to
reserves are relatively low. Nonetheless, SOEC supply
5 kg of yttrium.
chains are not risk free, because these electrolyzers
Some SOEC designs may also require tens of kilograms do require other rare earth metals such as zirconium,
of ceria and smaller quantities (single-digit kilograms) lanthanum, yttrium, and scandium. Until relatively
of scandium and gadolinium. Overall, the IEA expects recently, supply chains for some of these elements were
that these materials requirements could be halved very concentrated in China. But since 2016, substantial
through design improvements over the next decade, with new production has surfaced – chiefly in the United
technical potential to reduce nickel content to below States, Myanmar, and Australia. While the mining of rare
10 kg per MW. earth metals is still relatively concentrated globally, the
small quantities of these elements required for SOEC
Compared with alkaline electrolyzers, SOECs require an manufacture mean that alternative supplies can likely
order of magnitude less nickel. With nickel demand set to be found if needed. Some SOEC manufacturers have
soar due to rapid growth in global battery manufacture, already had to adjust as a result of supply chain problems
solid oxide electrolyzers are less exposed to nickel price with scandium but managed to find alternative sources
volatility than alkaline electrolyzers. At the current nickel without affecting their ability to deliver.
price of $22,000 per tonne,80 this metal accounts for

Figure 35: Estimated Levelized Demand for the Main Minerals in Electrolyzers and Fuel Cells Today81

10

1
kg per GWh output

0.1

0.01

0.001

0.0001
Nickel Zirconium Nickel Zirconium Lanthanum Yttrium Platinum Palladium Iridium Platinum

Alkaline electrolyser SOEC electrolyser (SOEC fuel cell) PEM electrolyser Fuel cell

80
Lme Nickel | London Metal Exchange. (n.d.). https://www.lme.com/en/metals/non-ferrous/lme-nickel
81
IEA (2022). The Role of Critical Minerals in Clean Energy Transitions Sector. https://iea.blob.core.windows.net/assets/ffd2a83b-8c30-4e9d-
980a-52b6d9a86fdc/TheRoleofCriticalMineralsinCleanEnergyTransitions.pdf

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 45

 Figure 36: Global Production of Rare-Earth Oxides82

(Metric tons - rare earth oxide equivalent )

300,000

240,000

180,000

120,000

60,000

0
1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022
Others United States Russia Burma/Myanmar China Australia

To illustrate why it should be relatively easy to risks, and obsolescence risk. Examples of financial risk
accommodate rare-earth demand from SOEC include the possibility that banks would be unwilling
manufacture, it is instructive to consider the market for to finance an unfamiliar technology or that project
lanthanum. In 2019, around 50,000 tonnes83 of lanthanum developers would prioritize cheaper upfront costs and
oxide, containing 42,500 tonnes of elemental lanthanum, select alkaline and PEM technologies. An example
were mined globally. Producing 1,000 MWh of hydrogen of operational risk is the possibility that an SOEC
using SOECs requires 0.2 kg of lanthanum. Thus, just system might be run outside the manufacturer’s design
one year's global supply of lanthanum would be enough parameters (e.g., with frequent cold starts or without
to produce 212.5 billion MWh of hydrogen (assuming maintaining optimal temperature conditions). This could
50,000 GW of SOEC running at 50% capacity factor). cause equipment failures and entrench the perception
that the technology remains underdeveloped. Other
external developments, meanwhile, can create
5.5 Non-Technological Risks to
obsolescence risk – for example if the electrification
SOEC Scaleup of industrial processes vastly reduces sources of
process heat for SOEC systems to tap into.
To successfully scale SOEC technology, several
additional, non-technological risks will need to be
overcome. These include financial risks, operational

82
King, H. M. (n.d.). REE - Rare Earth Elements and their Uses. Geology. https://geology.com/articles/rare-earth-elements/
83
Davis, S. (2020, July 29). Rare Earth Elements Supply Uncertain for IC Fabs. Semiconductor Digest.
https://www.semiconductor-digest.com/rare-earth-elements-supply-uncertain-for-ic-fabs/

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 46

 SECTION 6

Conclusion and Outlook

Solid oxide electrolyzer technology is at a higher The main advantage of SOEC technology is the step
readiness level than many give it credit for. Most of the change improvement in efficiency it offers when
companies that are pursuing SOEC development have integrated with an external process that can provide
already completed pilot projects, and even put their high-purity steam. This makes SOEC a perfect fit for
technology into commercial operation with industrial industrial sites with enough unused low-grade process
partners. One manufacturer – Bloom Energy – leads the heat to evaporate water. Ammonia, chemical, and steel
pack as it has already deployed more than 1 GW of solid plants, as well as refineries, would be particularly well
oxide systems worldwide. suited for SOEC integration. Geothermal plants are
another option. Some have proposed that SOEC systems
Scaling SOEC manufacturing to meet the world’s growing could tap into other sources of process heat, such as
appetite for electrolyzers will not be a bottleneck. from a nuclear power plant, but this approach is only
Experienced manufacturers with a track record can build scalable for new nuclear plants rather than for retrofits
a large, gigawatt-scale factory in less than two years and of existing assets or those currently under construction.
navigate supply chain risks for raw materials. The industry This will also require new regulatory approaches to
leader in SOEC manufacturing leader is again Bloom nuclear licensing and permitting.
Energy: its 2 GW/year stack manufacturing capacity
already exceeds that of most alkaline, PEM, and AEM Looking forward, SOEC manufacturers still have one
electrolyzer companies. Other solid oxide developers are critical technology hurdle to clear: they must successfully
planning to catch up: the runner up is Topsoe, which has scale their product from current modules, which – though
started building a 500-MW electrolyzer factory that is they qualify as commercial – are still relatively small, to
due to come online in 2025. Elcogen and FuelCell Energy large systems (in the hundreds of megawatts) that match
have also indicated they will be expanding production the size and the output of a typical industrial plant. If they
into the hundreds of MWs. succeed, SOEC is well positioned to be the preferred
hydrogen generation technology at industrial sites with
available process steam.

CATF – Solid Oxide Electrolysis: A Technology Status Assessment 47