Applied Energy 309 (2022) 118398

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Country-specific cost projections for renewable hydrogen production

through off-grid electricity systems
Jacob L.L.C.C. Janssen a, *, Marcel Weeda a, Remko J. Detz a, Bob van der Zwaan a, b, c
a
TNO, Energy Transition Department (ETS), Amsterdam, the Netherlands
b
University of Amsterdam, Faculty of Science (HIMS and IAS), Amsterdam, the Netherlands
c
Johns Hopkins University, School of Advanced International Studies (SAIS), Bologna, Italy

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• In Europe renewable H2 production

costs may reduce to <2 €/kg in 2050.
• Hybrid solar-wind electricity systems
result in the lowest H2 production costs.
• Individual off-grid electricity systems
yield the lowest H2 cost with onshore
wind.
• Electrolyser cost reductions are more
critical for H2 produced with PV than
wind.

A R T I C L E I N F O A B S T R A C T

Keywords: Renewable hydrogen is increasingly recognized as one of the key decarbonisation options compatible with the
H2 EU’s climate neutrality goal. We quantify possible cost reductions for renewable hydrogen production until 2050
Electrolysis through electrolysis with off-grid renewable electricity generation systems. We focus on the use of solar PV and
Off-grid electricity
on- and offshore wind energy in 30 European countries. We project that towards 2050 hydrogen production costs
Renewable energy
Cost reductions
can fall below 2 €/kg in several countries in Europe. Hybrid configurations, consisting of both onshore wind and
solar PV electricity generation, generally result in lower renewable hydrogen production costs. Systems with a
relatively high level of full load hours benefit from a reduced share of investment costs for the electrolyser
component. The levelized cost of hydrogen produced via solar PV systems can only compete with wind-based
systems when significant electrolyser cost reductions are realized, despite the ultimately low expected lev­
elized costs of solar PV-based electricity generation. The novelty of this analysis is that it proffers an overview of
the dependencies of the costs of green hydrogen production, and how these costs could decrease over the
forthcoming decades across a large set of European countries. Specifically, we show how the dynamics behind
the projected renewable hydrogen production costs per country highlight the role that technological learning
could have in identifying the most suitable locations for hydrogen production.

* Corresponding author.
E-mail address: [email protected] (J.L.L.C.C. Janssen).

https://doi.org/10.1016/j.apenergy.2021.118398
Received 9 July 2021; Received in revised form 28 November 2021; Accepted 14 December 2021
Available online 31 December 2021
0306-2619/© 2022 Elsevier Ltd. All rights reserved.
J.L.L.C.C. Janssen et al. Applied Energy 309 (2022) 118398

1. Introduction €/kg from 2030 onwards [26]. In all hydrogen studies it becomes evident
that substantial investments are necessary to scale up the required capacities
If countries around the world comply with the goals of the Paris agree­ and to benefit from cost reductions by further riding down the learning curve
ment, greenhouse gas (GHG) emissions need to be reduced to net-zero this of hydrogen production technologies [19]. The impact of learning-by-doing
century and fossil fuel usage has to decline significantly during the coming on renewable fuel production costs has been explored in several recent
decades [1]. This requires a fundamental change in the global energy system publications [20,21,7]. Analysis of cost reductions and learning phenomena
[2,3]. The large-scale production and use of hydrogen can be an important for especially renewable hydrogen production deserves increased attention,
contributor to the large adjustments necessary to transform the global energy in particular so as to determine changes in the competitivity of hydrogen as a
system [4]. Aside from high levels of electrification, sustainable energy result of the construction of new local hydrogen production facilities and the
supply will also require large amounts of renewable liquid and gaseous en­ implementation of updated regional and international hydrogen strategies.
ergy carriers [5,6,7,8]. Eventually, most processes employed to produce In this paper we focus on renewable hydrogen produced in 30 Eu­
materials and chemicals with fossil fuels as input will probably need to be ropean countries with either solar PV, onshore or offshore wind energy,
replaced by sustainable non-fossil alternatives (e.g. [9]). The most abundant or a combination of these options. We make cost projections up to 2050
renewable energy resources, solar and wind energy, are likely candidates to based on estimates retrieved from literature. We also study the dynamics
produce liquid or gaseous energy carriers in big volumes. The use of of the cost developments for specific components of off-grid renewable
renewable energy to produce hydrogen through electrolysis of water might hydrogen production systems. Potential differences across countries in
enable this. deployment rates and associated cost reductions of both renewable
Today, hydrogen is mainly produced from fossil resources with a domi­ electricity generation and electrolyser technology may have conse­
nant role for steam methane reforming (SMR). Hydrogen production costs by quences for the uptake of specific hydrogen production chains over time
SMR range roughly between 1.0 and 1.8 €/kg and depend largely on the price and space. We analyze the dependence on relative investment costs of
of natural gas [10,4]. With CO2 capture and storage (CCS), costs may increase different renewable hydrogen production systems in order to determine
to around 2.4 €/kg, although some analysts project that these costs may go which production chains may be deployed first, and in which countries.
down to below 2.0 €/kg during the coming decade as a result of declining The novelty of this research is that we provide an overview of the de­
costs for CCS [11]. Although hydrogen production cost might feasible at pendencies of the cost developments of green hydrogen production across a
2–2.5 €/kg, there are doubts as to whether CCS is viable to produce at scale large set of European countries, and show how these could inform its optimal
[12–14]. Dedicated hydrogen production through electrolysis of water with location. We quantify that a low cost of electricity does not necessarily imply
renewable electricity as input accounts currently for only 0.1% of global low renewable hydrogen production costs in a future with high shares of
hydrogen supply [4]. The cost of this renewable hydrogen depends largely on intermittent renewable energy production. In order to effectively reduce the
the price of electricity and the investment costs of the electrolyser. Even CAPEX component of the electrolyser, a high number of full-load hours is
under the unrealistically optimistic assumptions that electrolysis is 100% required. As the number of full-load hours for renewable energy resources
efficient and the electrolyser is for free, electricity prices still may not exceed varies by location, the optimal location for green hydrogen production de­
45 €/MWh in order to produce renewable hydrogen for 1.5 €/kg. Hence, the pends on the way learning rates for electrolysis develop, as well as on the
price of electricity needs to be really low in order for renewable hydrogen to dynamics in the levelized cost of electricity over time. To analyse both the
be able to compete with hydrogen produced with natural gas. Yet because the optimal timing and location of green hydrogen production, as well as their
costs of renewable electricity continue to decline, particularly from solar and interaction, this data needs to be reviewed synchronously. The graphical
wind energy, renewable hydrogen production may become competitive in representation we develop has both the depth and breadth to provide an
the future. Nearly every recent outlook on hydrogen published during the adequate overview that helps perform such analysis.
past few years (see, for instance, [15,11,4,16,17]) has proffered questions In Section 2 of this paper we describe the methodology and as­
like: which countries will adopt renewable hydrogen production first, when sumptions used for calculating the costs of renewable electricity gen­
will they start using renewable hydrogen, and what will the overall pro­ eration and associated renewable hydrogen production. The results of
duction chain look like? These are the questions that we attempt to answer in our analysis are presented and discussed in Section 3. In Section 4 we
this paper. formulate some conclusions and proffer several recommendations.
We here focus on the cost of electricity rather than its price. The use of
costs for electricity from off-grid renewable energy systems, instead of its 2. Methodology
prices as observed in electricity markets, allows us to best compare the
hydrogen production costs from systems employing different renewable For a combination of country i, technology j, and year t, we calculate
energy options, using different time scales, and involving different countries. for 30 European countries the levelized cost of electricity generation
This approach – applied for the research underlying this paper – is adopted in (LCOE) for several off-grid renewable electricity generation systems, as
several well-publicized studies, as it serves well the purpose of presenting the well as the levelized cost of hydrogen production (LCOH) for electro­
average global long-term costs of renewable hydrogen production via solar lysers connected to these off-grid renewable energy systems. The for­
and wind energy (see, for example, [4,11]). Recently, also hydrogen pro­ mulas for LCOE and LCOH are given by:
duction costs based on a combination of solar and onshore wind energy have [ ] [ ]
been analyzed in detail (see [26]). Optimal use of such hybrid systems allows [ € ] ρi,j *CAPEXj,t kW€
+ OPEXj,t kW €

for running the electrolyser during more hours per year (hence increasing the LCOEi,j,t = , (1)
kWh FLHi.j
amount of so-called full load hours, or FLH), which reduces the relative cost
contribution from the electrolyser to the overall costs of the renewable
hydrogen production system. Fasihi and Breyer project that at the best lo­
cations in the world, renewable hydrogen production costs may drop below 2

⎛ [ ] [ ]⎞
[
€
] [ € ] ρi,j *CAPEXj,t €
kW
+ OPEXj,t €
kW
LCOHi,j,t = ⎝LCOEi,j,t + ⎠*44.4, (2)
kg kWh FLHi,j

2
J.L.L.C.C. Janssen et al. Applied Energy 309 (2022) 118398

Table A1 Table A3
Outlook cost of electrolyser components. Cost projections (CAPEX for each year and OPEX in percentage) for renewable
Description Unit 2020 2030 2040 2050
energy generation technology. Based on the JRC Pro-Res Scenario [23] and
Agora [34].
CAPEX - electrolyser €/kW 700€ 600€ 520€ 450€
OPEX electrolyser €/kW 30 27 24 21 Unit 2020 2030 2040 2050 OPEX
Depreciation of loan Years 10 10 10 10 Utility-scale photovoltaics €/kW 690 450 370 320 2%
Rated system electrical efficiency % 75 75 75 75 without tracking
(LHV) Onshore wind turbines (high €/kW 1040 960 920 890 3%
Plant lifetime Years 30 30 30 30 specific capacity, low hub
height)
Offshore wind turbines €/kW 2890 2310 2150 2100 2%
WACCi,j Nj (monopole, medium
ρi,j = ( ) * . (3)
1− 1 + WACCi,j − Nj L
j
distance to shore)

In these equations, FLHi,j represents the amount of full load hours of

the combined off-grid renewable electricity plus electrolysis facility, consumption of the electrolyser determines the variable OPEX. The
44.4 is the assumed electricity consumption for producing a unit of LCOH varies for each combination of country, year, and technology. A
hydrogen in kWh/kg, and Ni and Li are the loan period and the lifetime full description of all data used can be found in the Appendix.
of the complete system, respectively. We thus assume that the electricity
generation system and electrolyser have the same lifetime, rather than 3. Results
each a separate lifetime (see Table A1 in the Appendix for the adopted
values of the system lifetime). We present the results of our LCOH and LCOE calculations for renewable
We first determine the FLH of the different electricity generation off-grid hydrogen production systems in Fig. 1. These systems consist of an
systems, projections of their capital expenditures (CAPEX) and opera­ off-grid renewable electricity generator connected exclusively to an elec­
tional expenditures (OPEX), and country- and technology-specific esti­ trolysis unit that produces hydrogen. The y-axis of Fig. 1 reports the
mates of the weighted average cost of capital (WACC, see Tables A2–A4, hydrogen production costs as described in the methodology section, that is,
respectively, in the Appendix; data are obtained from [22,23,24,25]. For without inclusion of the associated hydrogen storage and transportation
each of the 30 European countries we calculate, if applicable, the LCOE for costs since these fall outside the scope of our analysis. We recognize that the
solar, onshore and offshore wind electricity (see Eq. (1)). We include a intermittent nature of renewable hydrogen production systems may in
differentiation in FLH across generation types and countries, but neglect practice require some storage capacity to enable a baseload hydrogen supply.
changes in FLH over time. Within Europe, CAPEX and OPEX values for the Although these costs are excluded from our analysis, they can in principle
hydrogen production systems may differ between countries. In our model, quite easily be added: hydrogen storage in e.g. salt caverns may add around
however, for reasons of simplicity we assume that these costs for 2020, 10–20% to total hydrogen production costs [26].
2030, 2040 and 2050 are the same for each country (see Table A3). The The cost developments in Fig. 1 read from the upper right to the
CAPEX component only varies between countries due to influence of the lower left corner as time progresses from 2020 to 2050. The lines in this
country-specific (and technology-specific) WACC, which is kept constant Figure represent individual renewable hydrogen production systems
over time. The LCOE is the sum of operational and annuitized capital deployed in different countries across Europe. As can be seen, the lines
expenditures and is expressed in €/kWh. Each combination of country, are grouped in three distinct clusters, respectively referring to solar
year, and technology, thus, has a unique LCOE because the technology electricity systems (orange, top), onshore wind energy systems (light
costs develop over time, the FLH vary per country and technology, and the blue, middle) and offshore wind energy systems (yellow, bottom) that
WACC varies per country and technology. feed the units producing hydrogen through electrolysis. Some electricity
Next, the LCOE is used as input to determine the LCOH. We apply generating technologies are not shown for some countries, either
cost projections for the electrolyser as shown in Table A1 and use the because we did not find the required data or since our calculated elec­
same value in all countries and for each electricity generator irrespective tricity and hydrogen production costs fell outside the area displayed in
of the typical system. WACC and FLH data for the electrolyser match Fig. 1. The latter may be the result of relatively unfavorable local con­
those adopted for the electricity generation system in the country under ditions assumed in our analysis for some countries (as is the case for e.g.
consideration (see Tables A2 and A4; [25]). The LCOH is assumed to be solar electricity in Norway and offshore wind energy in Cyprus, amongst
the sum of electricity costs, and operational and annuitized capital ex­ others). To make the figure easier to follow, we have included the same
penditures for the electrolyser and is expressed in €/kg (see Eq. (2)). graph for each individual technology in the appendix.
Multiplication of the LCOE with the total level of electricity Solar electricity generation is cheapest, of course, in places with high

Table A2
Amount of full load hours per year for offshore wind, onshore wind and solar energy. Based on data from JRC [22] and JRC [23].
Country Code Wind onshore Wind offshore Solar Country Code Wind onshore Wind offshore Solar

Austria AU 28% 13.04% Latvia LA 28% 46% 9.29%

Belgium BE 24% 47% 11.78% Lithuania LT 29% 46% 9.70%
Bulgaria BU 23% 34% 14.30% Luxembourg LU 11.36%
Croatia HR 24% 24% 13.08% Malta MT 27% 38% 16.92%
Cyprus CY 21% 17.15% Netherlands NL 32% 47% 11.24%
Czechia CZ 24% 12.12% Norway NO 32% 45% 5.16%
Denmark DK 37% 47% 10.43% Poland PO 26% 45% 10.85%
Estonia EE 27% 46% Portugal PT 23% 34% 21.25%
Finland FI 29% 46% 6.77% Romania RO 23% 36% 14.25%
France FR 27% 42% 13.37% Slovak Republic SK 24% 10.70%
Germany DE 24% 44% 10.21% Slovenia SV 14.25%
Greece GR 28% 29% 16.89% Spain ES 26% 31% 17.51%
Hungary HU 23% 14.07% Sweden SE 30% 44% 7.50%
Ireland IE 45% 53% 9.02% Switzerland CH 28% 13.69%
Italy IT 25% 25% 13.28% United Kingdom UK 34% 49% 9.48%

3
J.L.L.C.C. Janssen et al. Applied Energy 309 (2022) 118398

Table A4
Country level data for the WACC for renewable energy generation. Based on data from Eurobserv’er (Eurobserv’er, 2017). *For these countries we assume a WACC of
8% in the absence of data.
Country Wind onshore Wind offshore Solar Country Wind onshore Wind offshore Solar

Austria 5% 7% 5% Latvia 18% 20% 18%

Belgium 8% 10% 8% Lithuania 1% 3% 1%
Bulgaria 7% 8% 7% Luxembourg 1% 3% 1%
Croatia 7% 8% 7% Malta 5% 6% 5%
Cyprus 5% 6% 5% Netherlands 1% 3% 1%
Czechia 4% 6% 4% Norway* 8% 8% 8%
Denmark 1% 2% 1% Poland 4% 6% 4%
Estonia 5% 6% 5% Portugal 6% 8% 6%
Finland 5% 6% 5% Romania 6% 8% 6%
France 5% 7% 5% Slovak Republic 7% 8% 7%
Germany 5% 7% 5% Slovenia 5% 6% 5%
Greece* 8% 8% 8% Spain 4% 6% 4%
Hungary 3% 4% 3% Sweden 1% 3% 1%
Ireland 9% 11% 9% Switzerland* 8% 8% 8%
Italy 4% 5% 4% United Kingdom* 8% 8% 8%

Fig. 1. Cost projections for off-grid electricity generation and renewable hydrogen production in various European countries.

levels of irradiation, typically located in more southern countries of Europe The reason for this is the relatively high amount of FLH for the latter. With
like Italy, Portugal and Spain. The LCOH for solar-based hydrogen produc­ more FLH, fewer electrolysers and thus lower investments are required per
tion systems in 2020 ranges from 3.5 €/kg in Portugal to almost 15 €/kg in kilogram of hydrogen output. Economies-of-scale, although not specifically
Norway. Our calculations show that this large bandwidth narrows down to a analyzed here, may likely improve the case for offshore wind-based systems
range between 2.0 and 8.4 €/kg in 2050. Onshore wind electricity generation because of their relatively large size in comparison to most onshore wind-
benefits from lower investment costs per kW of installed capacity, but and solar-based energy parks. In Fig. 1 we see that in all countries the cost
generally possesses fewer FLH, in comparison to offshore wind energy. With developments follow similar patterns: in all cases they are largely driven by
onshore wind-based systems, in the majority of countries an LCOH of 3–4 reductions in technology investment costs.
€/kg can be obtained in 2020. Towards 2050, this range is expected to reduce Fig. 2 again plots the results of our LCOH and LCOE calculations and
to 2–3 €/kg, while in some countries even slightly lower hydrogen produc­ shows the same lines depicted in Fig. 1, but now without the colored shapes
tion costs might be achievable, for instance in Denmark and Ireland. In that indicate the clusters (i.e. types of electricity generation) to which indi­
Ireland, an LCOH below 2 €/kg might already be realized in 2030. Offshore vidual renewable hydrogen production systems belong. Instead, we indicate
wind electricity generation has most FLH per year on average for the three with a dashed line the contribution of electricity generation to the overall
types of systems considered in this paper. The LCOH in 2020 for offshore renewable hydrogen production costs. The dashed line is the threshold below
wind-based systems ranges between 3.2 and 8.0 €/kg and is projected to fall which the LCOH can never fall, as it represents the (unrealistic) case that
to 2.3–5.7 €/kg in 2050. In countries with a combination of low WACC and electricity generation constitutes the sole contribution to the costs of
high value for FLH thanks to the availability of high offshore wind resources, renewable hydrogen production. The graph shows that towards 2050 the
like Denmark and the Netherlands, the lowest offshore renewable hydrogen cost share of the electrolyser will be relatively large in the best-case solar-
production costs can be achieved (2.3 €/kg in 2050). based hydrogen production systems, in comparison to that in the best-case
Although electricity generated with solar PV can in some countries onshore and offshore wind-based systems. This means that reducing the
probably be generated at costs lower than those for wind energy, especially in costs of electrolysers will more positively impact the business case for solar-
the longer term, the hydrogen production costs associated with PV are pro­ based hydrogen generation in sunny areas in comparison to that for wind-
jected to remain similar or even higher than those for wind-based systems. based hydrogen production in windy areas. Fig. 2 also visualizes that the

4
J.L.L.C.C. Janssen et al. Applied Energy 309 (2022) 118398

Fig. 2. Electricity and electrolysis cost components in off-grid renewable hydrogen production: the arrows above the diagonal dashed red line indicate the elec­
trolyser cost component, the arrows below this line indicate the electricity cost component. Arrows are shown by way of example for three different cases
(distinguished by different countries, years, and technologies). (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)

lowest levelized cost of renewable electricity will not automatically result in difficult to achieve (700 €/kW). Only when electrolyser costs decline
the lowest renewable hydrogen production costs, because the business case down to 240 €/kW in 2050, renewable hydrogen production costs in
also depends on the amount of FLH. If the amount of FLH increases (for these two countries can reach a break-even point. We assume that the
example when going from solar- to wind-based systems), the relative elec­ peak capacity of electricity generation is matched by the capacity of the
trolyser cost component of LCOH decreases. We thus find that hydrogen electrolyser. Therefore, the system in Ireland, with more FLH per year
production systems with more FLH (e.g. based on onshore or offshore wind but a lower peak capacity, will have a lower electrolyser capacity and,
power) in comparison to those with fewer FLH (e.g. based on solar energy) thus, costs per unit of output are lower than for the solar-based hydrogen
provide better business cases, even while the former possess a higher LCOE, production system in Portugal. As a consequence, hydrogen production
until the electrolyser cost component drops below a certain level. costs in Portugal will be more sensitive to the CAPEX of the electrolyser
To further illustrate the effect of electrolyser cost reductions, we than those based on onshore wind energy in Ireland. Determining an
compare the cost of renewable hydrogen production in the case of solar optimal location for renewable hydrogen production systems for the
electricity generation in Portugal with that of onshore wind power long-term is not trivial, due to the prevailing uncertainty in projected
generation in Ireland (see Fig. 3). As can be seen, electricity generation electrolyser cost reductions. It is unlikely though that the total invest­
by solar PV in Portugal is projected to be cheaper than electricity from ment costs of electrolyser plants will readily drop below 240 €/kW, in
wind energy in Ireland in 2050 (Figs. 1 and 2 show that this could which the competitive edge between these two cases is rather clear.
already be the case in 2030). Renewable hydrogen production costs from Based solely on production costs, we find that the market-uptake of large
onshore wind in Ireland, however, remain lower according to our pro­ off-grid renewable hydrogen production facilities will most probably
jections due to the relatively high cost of the electrolyser (450 €/kW). initiate in countries in which the electrolyser cost component of the
The difference in hydrogen production costs between these two cases overall hydrogen production system is lowest.
becomes even larger if reductions in the CAPEX of the electrolyser prove Deeper reductions in the costs of producing renewable hydrogen can be

Fig. 3. Example of the impact on LCOH of electrolyser cost reductions: solar electricity in Portugal versus onshore wind energy in Ireland, both in 2050.

5
J.L.L.C.C. Janssen et al. Applied Energy 309 (2022) 118398

LCOH vs LCOE for hybrid onshore wind/solar systems

4

3.5

LCOH (€/kg)
BE
3
RO
FI BG
2.5 UK SK
SE NO DE
MT CZ
PT HR
2 IE CH PL
IT HU
LT
DK AT FR
NL ES GR
1.5
15 20 25 30 35 40 45 50
LCOE (€/MWh)
Fig. 4. Cost developments for off-grid renewable hydrogen production systems based on a combination of solar and wind electricity generation.

made by combining solar- and wind-based energy systems, since the hours the LCOH declines to below 2.5 €/kg in 2050 in 90% of the countries that
during which electricity is generated in these two types of systems are largely we inspected. In several countries in Europe, renewable hydrogen pro­
complementary [18]. We assume that the capacities of the wind and solar duction costs may eventually drop below 2 €/kg and in a few cases, like in
electricity generation systems are the same. Fasihi and Breyer [26] showed in Ireland and Denmark, by even as early as 2030. If projected technology
that a hybrid PV-wind baseload hydrogen production plant with a salt cavern cost reductions materialize, renewable hydrogen production costs may
storage facility requires, under cost-optimal conditions, nearly identical thereby become competitive with conventional reforming techniques to
installed capacities for the PV and wind energy systems, respectively [26]. produce hydrogen, especially when carbon pricing increases.
Thanks to the combination of two different types of renewable electricity The top five countries with the lowest LCOH from individual
generation technologies with distinct load profiles, the hydrogen production renewable electricity generation plants are almost exclusively repre­
system provides a more stable power pattern in comparison to plants with sented by onshore wind-based systems: Ireland, Denmark, Netherlands,
only a single technology. The combined system therefore requires the lowest UK, and Sweden. Despite the low future LCOE of solar-based energy
electrolyser capacity, even slightly lower than the nominal capacity of either systems in several countries, only one country with a solar-based system
the wind or solar electricity generation system alone. The electrolyser ca­ enters the top five in 2040: Portugal. The low amount of FLH of solar-
pacity in our case equals that of the peak output capacity of either the solar or based systems causes the electrolyser costs to dominate the LCOH,
wind energy system. This means that only the electricity generation above which yields a higher sensitivity of the LCOH to projected costs and cost
the nominal output has to be curtailed. In combined systems like these, reductions for the electrolysis unit.
Gerlach et al. [27] estimate that in most of Europe only roughly 3–4% of the It is particularly important to assess renewable hydrogen production
generated electricity is left unused. The energetic loss is compensated for by costs in those countries in which renewable hydrogen production is likely
lower electrolyser investment costs per unit of hydrogen output because of to take-off at a substantial scale. We show that in countries in the northern
the higher usage time. The resulting cost of renewable hydrogen production coastal areas of Europe with a high potential for onshore and offshore
is shown in Fig. 4. The costs of hydrogen production for these hybrid systems wind-based power, LCOH values are the lowest. These countries therefore
are generally lower, typically 2.4–4.4 €/kg in 2020, than these costs for in­ are appealing locations for initiating and demonstrating renewable
dividual systems. This can be ascribed to an increase in FLH. For some hydrogen production plants at a large scale. The additional “learning in­
countries, however, such as Ireland and Norway, the inclusion of solar en­ vestments” (usually from public sources) needed to reach competitive
ergy supply results in higher hydrogen production costs in comparison to break-even (under which industry can enter the market) would be the
systems solely based on wind energy. A reduction in hydrogen production lowest for countries in these regions. From a hydrogen value chain
costs by as much as 0.6–0.7 €/kg for the hybrid system, instead of the perspective, production costs only do not capture the entire picture. Other
cheapest individual system, can be observed in countries such as Bulgaria, aspects, such as hydrogen storage and distribution, hydrogen demand
Croatia, Hungary, and Romania, since they benefit from a higher amount of potential, renewable electricity demand and supply potential, and local
FLH for the combination of wind and solar energy. In 2050, the LCOH for policy measures, can also be determinant in final decision making with
hybrid hydrogen production systems drops to values between 1.6 and 3.0 regard to the realization of hydrogen production projects. Although these
€/kg. The lowest renewable hydrogen production costs are achieved in aspects are not part of our techno-economic analysis, we here discuss
Denmark, Ireland, and the Netherlands (1.6, 1.7, and 1.8 €/kg, respectively). several of them. The data only reflects a limited set of countries, and there
Hence, we confirm the values projected by others (for instance, [18,26]) and is reason to believe other regions could produce low-cost hydrogen as
show that several European countries belong to the best locations in the well, both for intended use in European countries and elsewhere. We
world for renewable hydrogen production. chose to focus on production costs within Europe and hope future work
will expand our analysis to more regions (see, for example, [28] ).
4. Conclusion and policy implications Renewable hydrogen production through off-grid renewable energy
systems probably requires a substantial storage facility in order to
In this paper we project significant reductions in renewable hydrogen enable non-intermittent supply. Especially for seasonal hydrogen stor­
production costs for 30 European countries. Our analysis demonstrates age, underground formations like salt caverns, depleted oil or gas res­
that values for the LCOH of renewable hydrogen production systems cover ervoirs, and aquifers seem promising. From these options, salt caverns
a broad range, starting from 2.1 to 15 €/kg in 2020 and declining down to likely offer the best features, like low cushion gas requirement, large
1.6–8.4 €/kg in 2050. The costs of renewable hydrogen production differ sealing capacity and inert nature of rock salt, and existing experience for
as a result of varying technologies, locations, and points in time. In nearly natural gas storage purposes. Countries with high technical potentials
all countries, hybrid systems based on a combination of solar PV and for on- and offshore hydrogen storage in salt caverns, exceeding 10 EJ of
onshore wind energy systems lead to the lowest LCOH. For these systems, H2, are notably: Germany, the Netherlands, United Kingdom, Denmark,

6
J.L.L.C.C. Janssen et al. Applied Energy 309 (2022) 118398

Norway, Poland, and Portugal [29]. Except perhaps for Ireland and the moment remain well below those of the former. To achieve com­
Sweden, this shows that several countries with low LCOH projections petitivity, substantial cost reductions are deemed necessary. It is important
also possess the ability to store hydrogen in salt caverns. to gain experience and accelerate learning processes through an increased
If hydrogen storage is not available in a country, a distribution network deployment of renewable hydrogen production technology at commercial
with pipelines may allow the transportation of hydrogen throughout Europe. scale as soon as possible [42,43]. Competitivity of these projects should
Such a network is currently not widely available and its absence may hamper temporarily be guaranteed by well-designed and long-term policy. To limit
the uptake of hydrogen production at some locations. Distribution and the necessary learning investments, the first projects are preferably realized
transport over long distances might be unnecessary if hydrogen demand in countries in which a low LCOH can be achieved, like Ireland and Denmark.
sectors are nearby the production site. Various applications exist for which We here only assessed solar- and wind-based renewable electricity supply
hydrogen can be used, but its future role remains uncertain. Its current use in options. Other low-carbon electricity supply options, based on for instance
mainly the chemical industry makes it attractive to connect pioneering off- hydropower and nuclear or geothermal energy, benefit from high amounts of
grid renewable hydrogen production facilities with industrial clusters, so FLH and might, already during the 2020s, result in a relatively low LCOH
as to benefit from existing hydrogen infrastructure. Historically, industrial because of a small electrolyser cost component. The experience gained in
clusters have grown in areas well connected to energy supply, for example in early projects based on these other low-carbon energy options may trigger
regions with easy access to crude oil via rivers and coastal harbors, and with cost reductions toward 2030, and beyond, and open up markets in an
access to natural gas resources via pipelines. In the longer term, it can be increasing number of countries for solar and wind energy based hydrogen
envisaged that novel industrial sites will appear around locations with high production.
availabilities of renewable energy supply. We have shown that our type of LCOH analysis can provide good
The deployment of off-grid renewable hydrogen production systems first-order approximations for answering the question in which coun­
requires substantial investments. Countries in which the costs of capital tries renewable hydrogen production chains can best be deployed at
are relatively low have an advantage over other countries. In the present large scale first. Our calculations can also be used for assessing which
study we have used country-specific WACC values to express this differ­ type of renewable electricity systems should be employed for feeding the
ence in our LCOH analysis. To enhance the competitivity of renewable electrolysers that produce hydrogen. We have pointed out that topics
hydrogen production during the initial phase, local policies are key: they other than LCOH calculations should also be contemplated for
will likely play a decisive role in furthering their maturity and in deter­ enhancing our understanding of renewable hydrogen production chains
mining the optimal location. Countries can adopt various types of policies, and for increasing our knowledge of where to optimally locate them. In
such as subsidy schemes, and CO2 taxation, which can enhance the this paper we have not studied renewable hydrogen production systems
attractiveness of hydrogen production projects for investors. Clear and through the use of learning curves, as done in e.g. Detz et al. [7]. Rather,
long-term policy can reduce the risk and uncertainty for these investors we have here adopted a supplementary approach. We recommend to
and possibly be as important as a low value of the LCOH for them. complement and extrapolate the analysis proffered in this paper with
In simulations of grids with near-100% renewable generators, we studies that explore the possible learning effects of renewable hydrogen
observe a relatively high degree of spillage (e.g. for Australia, see production technologies based on off-grid electricity systems. This
[30,31]). One can ask what the potential is of this spilled energy to be would fit well in the context of the new European hydrogen strategy
either used or at least complemented to off-grid facilities, for the pro­ [33]. Of course, Europe cannot be studied entirely in isolation for the
duction of hydrogen. We think that a future with a high share of purpose of renewable hydrogen production systems. We therefore
renewable energy on the grid will have most spillage when there is more recommend that also the global trends in hydrogen and renewable
wind or solar power generated than can be used or stored. These mo­ electricity generation are taken into account: these could significantly
ments will typically occur when it is very windy or sunny, respectively. influence the technology learning rate and thereby affect the deploy­
It is at these times that off-grid systems are also operating at high ca­ ability of these systems in Europe. We intend to include such effects in
pacity. Therefore, the amount of spilled energy that can be used in off- future follow-up assessments.
grid systems could well be limited.
Another challenging aspect in the renewable hydrogen production chain Declaration of Competing Interest
might be the availability of renewable electricity. The renewable share of
total electricity consumption in Europe on average is currently approxi­ The authors declare that they have no known competing financial
mately 30%. To meet the GHG emission reduction and renewable energy interests or personal relationships that could have appeared to influence
targets of the European Union, this share should at least be doubled by 2030 the work reported in this paper.
[32]. Ongoing electrification of many processes and applications steadily
increases overall electricity demand. It is still uncertain if this growing de­ Acknowledgements
mand can be fulfilled by renewable energy so quickly, and if so, whether
there would be enough renewable electricity left to produce hydrogen, and to This work has been undertaken as part of the H2Future project,
what extent that may depend on the location. Some countries with a low which received funding from the Fuel Cells and Hydrogen Joint Un­
share of renewable electricity supply might preferably dedicate this gener­ dertaking of the European Commission under grant agreement No.
ated power primarily to decarbonize electricity for the principal grid and 735503.
disfavor electricity use for renewable hydrogen production. In other coun­
tries electricity demand can be low in comparison to their renewable supply Appendix
potential, which may favor its use for renewable hydrogen production earlier
in time. At some locations hydrogen supply could surpass local demand. This section provides details of the different datasets used in the
From these locations hydrogen may be exported to other regions, which analysis.
could result in the establishment of an international hydrogen market, within
and/or outside Europe. Electrolysis assumptions
All these aspects taken together likely determine which countries will
adopt renewable hydrogen production through off-grid electricity systems The investment costs for electrolysis are based on the lower side of
first. We find that onshore wind-based systems are probably the early win­ the range as reported by Glenk et al. [15] for 2020 and 2030, while
ners, while hybrid systems might take over in the 2030s. It will take at least projections up to 2050 refer to the average long term cost estimate for
another decade before renewable hydrogen generation can compete with alkaline electrolysis as described by the IEA [4]. We utilize an electrical
conventional fossil-based hydrogen production, as the costs of the latter for efficiency of 75% (LHV) for all years (see Fig. A1).

7
J.L.L.C.C. Janssen et al. Applied Energy 309 (2022) 118398

Fig. A1. Cost projections for off-grid solar electricity generation and renewable hydrogen production in various European countries.

Renewable electricity generation assumptions solar energy refers to ‘utility-scale photovoltaics with one-axis tracking’.

For the CAPEX and OPEX of the renewable energy generators, data WACC Assumptions
from the JRC was used as well [23]. We use the cost data as specified in
the Pro-RES scenario in JRC [23]. For the full load hours, data was used For electrolysis, we use an average WACC of 8% throughout the
from JRC [22] and JRC [23]. Available data is shown in Table A2. The analysis. For the remaining renewable energy generation activities, we
onshore wind energy refers to turbines with relatively high specific ca­ follow [25], with country- and technology-level WACC data below in
pacity and relatively low hub height of JRC [24], the offshore wind Table A4 (see Figs. A2 and A3).
energy refers to ‘monopoles and medium distance to shore’, and the

Fig. A2. Cost projections for off-grid offshore wind electricity generation and renewable hydrogen production in various European countries.

8
J.L.L.C.C. Janssen et al. Applied Energy 309 (2022) 118398

Fig. A3. Cost projections for off-grid onshore wind electricity generation and renewable hydrogen production in various European countries.

Appendix A. Supplementary material [18] Shi X, Qian Y, Yang S. Fluctuation analysis of a complementary wind− solar energy
system and integration for large scale hydrogen production. ACS Sustain Chem Eng
2020;8(18):7097–110.
Supplementary data to this article can be found online at https://doi. [19] Schoots K, Ferioli F, Kramer G, van der Zwaan B. Learning curves for hydrogen
org/10.1016/j.apenergy.2021.118398. production technology: an assessment of observed cost reductions. Int J Hydrogen
Energy 2008;33(11):2630–45.
[20] Schmidt O, Hawkes A, Gambhir A, Staffell I. The future cost of electrical energy
References storage based on experience rates. Nat Energy 2017;2017(2):17110. https://doi.
org/10.1038/nenergy.2017.110.
[1] UNFCCC. Adoption of the Paris agreement. In: United nations framework [21] Schmidt O, Gambhir A, Staffell I, Hawkes A, Nelson J, Few S. Future cost and
convention on climate change, 21st conference of the parties, Paris, United performance of water electrolysis: an expert elicitation study. Int J Hydrogen
Nations; 2015. Energy 2017;42(52):30470–92. https://doi.org/10.1016/j.ijhydene.2017.10.045.
[2] IPCC. Climate change 2014: synthesis report. contribution of working groups I, II [22] JRC. In: Gonzalez-Aparacio I et al. (Eds.), JRC Science for policy report. EMHIRES
and III to the fifth assessment report of the intergovernmental panel on climate dataset Part II: Solar power generation; 2017.
change; 2014. [23] JRC. In: Dalla Longa F et al. (Eds.), JRC technical reports. Wind potentials for EU
[3] IPCC. Global warming of 1.5◦ C. An IPCC Special Report on the impacts of global and neighbouring countries; 2018a.
warming of 1.5◦ C above pre-industrial levels and related global greenhouse gas [24] JRC. Tsiropoulos I, Tarvydas, D, Zucker A. Cost development of low carbon energy
emission pathways, in the context of strengthening the global response to the technologies – Scenario-based cost trajectories to 2050, 2017 Edition, EUR 29034
threat of climate change, sustainable development, and efforts to eradicate EN, Publications Office of the European Union, Luxembourg; 2018b. Accessed
poverty. In: Masson-Delmotte V, Zhai P, Pörtner HO, Roberts D, Skea J, Shukla PR, through: http://publications.jrc.ec.europa.eu/repository/bitstream/JRC
et al., (Eds.); 2018. 109894/cost_development_of_low_carbon_energy_technologies_v2.2_final_online.
[4] IEA. The Future of Hydrogen. Report prepared by the IEA for the G20 in Japan; pdf.
2019. [25] EurObserv’er. The state of renewable energies in Europe. 13th EurObserv’ER
[5] Hydrogen Council. Hydrogen scaling up. A sustainable pathway for the global report; 2018. p. 4–9.
energy transition; 2017. [26] Fasihi M, Breyer C. Baseload electricity and hydrogen supply based on hybrid PV-
[6] Shell. Sky shell scenarios. Meeting the goals of the paris agreement; 2018. wind power plants. J Cleaner Prod 2020;243:118466. https://doi.org/10.1016/j.
[7] Detz RJ, Reek JN, van der Zwaan BCC. The future of solar fuels: when could they jclepro.2019.118466.
become competitive? Energy Environ Sci 2018;11(7):1653–69. https://doi.org/ [27] Gerlach AK, Stetter D, Schmid J, Breyer C. PV and wind power-complementary
10.1039/C8EE00111A. technologies. In: Proc 26th European photovoltaic solar energy conference, Vol. 2;
[8] FCHJU. Hydrogen roadmap Europe. A sustainable pathway for the European 2011.
energy transition; 2019. [28] van der Zwaan BCC, Lamboo S, Dalla F, Longa. Timmermans’ dream: an electricity
[9] Detz RJ, van der Zwaan B. Transitioning towards negative CO2 emissions. Energy and hydrogen partnership between Europe and North Africa. Energy Policy 2021;
Policy 2019;133:110938. https://doi.org/10.1016/j.enpol.2019.110938. 159:112613.
[10] IEA. Techno-economic evaluation of SMR based standalone (merchant) hydrogen [29] Caglayan DG, Weber N, Heinrichs HU, Linßen J, Robinius M, Kukla PA, et al.
plant with CCS; 2017. Accessed through https://ieaghg.org/exco_docs/2017-02. Technical potential of salt caverns for hydrogen storage in Europe. Int J Hydrogen
pdf. Energy 2020;45(11):6793–805. https://doi.org/10.1016/j.ijhydene.2019.12.161.
[11] Hydrogen Council. Path to hydrogen competitiveness. A cost perspective; 2020. [30] Elliston B, Diesendorf M, MacGill I. Simulations of scenarios with 100% renewable
Available from: https://hydrogencouncil.com/wp-content/uploads/2020/01/Path electricity in the Australian National Electricity Market. Energy Policy 2012;45:
-to-Hydrogen-Competitiveness_Full-Study-1.pdf. 606–13.
[12] Keyßer LT, Lenzen M. 1.5 ◦ C degrowth scenarios suggest the need for new [31] Lenzen M, McBain B, Trainer T, Jütte S, Rey-Lescure O, Huang J. Simulating low-
mitigation pathways. Nature Commun 2021;12:2676. carbon electricity supply for Australia. Appl Energy 2016;179:553–64.
[13] Budinis S, Krevor S, Dowell NM, Brandon N, Hawkes A. An assessment of CCS costs, [32] Banja M, Jégard M. Renewable technologies in the EU electricity sector: trends and
barriers and potential. Energy Strategy Rev 2018;22:61–81. projections – analysis in the framework of the EU 2030 climate and energy
[14] Fajardy M, Mac Dowell N. The energy return on investment of BECCS: is BECCS a strategy. JRC109254; 2017. https://doi.org/10.2760/733769.
threat to energy security? Energy Environ Sci 2018;11(6):1581–94. [33] European Commission (EC). A hydrogen strategy for a climate-neutral Europe.
[15] Glenk G, Reichelstein S. Economics of converting renewable power to hydrogen. COM(2020) 301 final; 2020. Available from: https://ec.europa.eu/energy/sites/e
Nature Energy 2019;4(3):216–22. ner/files/hydrogen_strategy.pdf (accessed July 2020).
[16] IRENA. Hydrogen from renewable power: technology outlook for the energy [34] Agora. The future cost of electricity-based; 2018. https://www.agora-energiewend
transition; 2018. e.de/en/publications/the-future-cost-of-electricity-based-synthetic-fuels-1/.
[17] NOW. Studie IndWEDe. Industrialisierung der Wasserelektrolyse in Deutschland: [42] Rivera-Tinoco R, Schoots K, van der Zwaan B. Learning curves for solid oxide fuel
Chancen und Herausforderungen für nachhaltigen Wasserstoff für Verkehr, Strom cells. Energy Convers Manage 2012;57:86–96.
und Wärme; 2018. [43] Schoots K, Kramer GJ, van der Zwaan BCC. Technology learning for fuel cells: an
assessment of past and potential cost reductions. Energy Policy 2010;38(6):
2887–97.