i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/he

Techno-economic calculations of small-scale

hydrogen supply systems for zero emission
transport in Norway

Øystein Ulleberg, Ragnhild Hancke*

Institute for Energy Technology, P.O. Box 40, NO-2027 Kjeller, Norway

article info abstract

Article history: In Norway, where nearly 100% of the power is hydroelectric, it is natural to consider water
Received 18 November 2018 electrolysis as the main production method of hydrogen for zero-emission transport. In a
Received in revised form startup market with low demand for hydrogen, one may find that small-scale WE-based
19 May 2019 hydrogen production is more cost-efficient than large-scale production because of the
Accepted 20 May 2019 potential to reach a high number of operating hours at rated capacity and high overall
Available online 13 June 2019 system utilization rate. Two case studies addressing the levelized costs of hydrogen in local
supply systems have been evaluated in the present work: (1) Hydrogen production at a
Keywords: small-scale hydroelectric power plant (with and without on-site refueling) and (2) Small
Water electrolysis hydrogen refueling station for trucks (with and without on-site hydrogen production). The
Zero emission transport techno-economic calculations of the two case studies show that the levelized hydrogen
Hydrogen refueling system refueling cost at the small-scale hydroelectric power plant (with a local station) will be 141
Techno-economic calculations NOK/kg, while a fleet of 5 fuel cell trucks will be able to refuel hydrogen at a cost of 58 NOK/
kg at a station with on-site production or 71 NOK/kg at a station based on delivered
hydrogen. The study shows that there is a relatively good business case for local water
electrolysis and supply of hydrogen to captive fleets of trucks in Norway, particularly if the
size of the fleet is sufficiently large to justify the installation of a relatively large water
electrolyzer system (economies of scale). The ideal concept would be a large fleet of heavy-
duty vehicles (with a high total hydrogen demand) and a refueling station with nearly 100%
utilization of the installed hydrogen production capacity.
© 2019 The Authors. Published by Elsevier Ltd on behalf of Hydrogen Energy Publications
LLC. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).

avoid CO2-emissions from the production and distribution of

Introduction hydrogen. As pointed out in a previous study [1] there are
several factors making Norway a suitable country for intro-
The supply of hydrogen as a transportation fuel is challenging, duction of hydrogen for zero emission transport, including the
especially if the hydrogen is to be produced locally based on high untapped potential for wind power and the very high
renewable energy sources. To claim zero emission in the full fraction (nearly 100%) of hydroelectric power. These condi-
value chain (from production to utilization), it is necessary to tions also make it natural to consider water electrolysis (WE)

* Corresponding author.
E-mail address: ragnhild.hancke@ife.no (R. Hancke).
https://doi.org/10.1016/j.ijhydene.2019.05.170
0360-3199/© 2019 The Authors. Published by Elsevier Ltd on behalf of Hydrogen Energy Publications LLC. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1202 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1

as the main production method of renewable hydrogen. Water envisaged hydrogen station is owned and operated by a
electrolysis is the key technology for Power-to-Gas concepts trucking company and offers refueling at 350 bar only. The
and is thus expected to play a major role in future energy station is dimensioned according to the fleet size to ensure full
systems which need to store excess energy coming from capacity utilization, and by designing the dispenser system for
renewable and intermittent power sources [2,3]. time fills [19], unattended overnight refueling can be intro-
In a startup market with low demand for hydrogen, one duced and very small hydrogen buffer storage volumes
may find that local small-scale WE-based hydrogen produc- installed. For this station configuration we investigate the
tion is more cost-efficient than regional large-scale produc- levelized cost of refueling hydrogen for various fleet sizes and
tion because of the potential to reach a high number of compare a refueling facility supplied by on-site hydrogen
operating hours at rated capacity and high overall system production to one which is supplied by trailers from a regional
utilization rate. This hypothesis is part of the motivation production plant.
behind the feasibility studies carried out the last few years The most important output variable of these case studies is
addressing hydrogen production at small-scale hydroelectric the levelized cost of dispensing 1 kg hydrogen into a FCEV.
power stations in Norway [4e7]. The rationale, from the plant This cost is derived from the average total capital and oper-
owner's point of view, is that it may be more profitable to ating cost (CAPEX and OPEX) of producing, compressing and
convert the electricity to hydrogen rather than to sell it refueling hydrogen over the project's lifetime, divided by the
directly to the grid, particularly during periods with low spot total hydrogen dispensed over that lifetime. The future cost
prices. Local hydroelectric power can avoid being subject to savings based on learning curves and economies of scale have
electricity tax, grid fees, or electricity certificates, if it is not been considered in this study, as the intention has been to
operated in stand-alone power mode. If such a power plant is inform today's stakeholders and decision makers of the ex-
dedicated for hydrogen production, the cost of connecting to pected near-term costs of hydrogen refueling for different
the local grid can be saved altogether. supply chains based on the existing and state-of-the art
The decision on whether a hydrogen refueling station technology.
(HRS) is to be established on the production site, or if the
hydrogen should be compressed and transported to a nearby
refueling station is influenced by the strategy on how to build Methodology and assumptions
up a nationwide hydrogen infrastructure in Norway. One
strategy is to establish a national network of 700 bar HRSs as A techno-economic modelling tool which can be used to es-
soon as possible so that the market for fuel cell electric ve- timate the costs of different designs of water electrolysis-
hicles (FCEVs) can be allowed to grow faster, while another based hydrogen refueling stations on a case-by-case basis
approach is to focus on establishing stations for captive has been developed in the program Engineering Equation
fleets, such as trucks and buses, as this will ensure high Solver (EES). The simulation tool can be adjusted and used to
utilization of refueling capacity and predictable returns on access the techno-economics of a wide range of RE-based
investment in the startup market [8]. In most cases the water electrolyzer systems, including hydrogen compres-
answer will depend on geographical location and several sion, storage and dispensing systems. The main technical
other factors and is therefore normally a decision made on a performance parameters (e.g. WE efficiency, lifetime, auxil-
case-by-case basis. Yet, several general models and strategies iary power needs) and cost functions (e.g. specific costs as a
for the deployment of hydrogen infrastructures in countries function of rated power or flow rates) for the main systems
and regions around the world have been proposed in and key pieces of equipment are entered into the model, in
numerous studies over the years [1,9e15], and a few high- addition to a set of economic parameters (e.g. electricity price,
level techno-economic studies have attempted to evaluate interest rate, project lifetime). From this the model calculates
the cost efficiency of the various supply schemes and tech- the total CAPEX and OPEX, and eventually the overall
nologies [16e18]. hydrogen cost (NOK/kg) for different operating scenarios (e.g.
In this paper we present a more detailed techno-economic different load profiles or electricity prices).
system analyses of two different small-scale hydrogen supply The supply chains and main system components which
systems in Norway based on water electrolysis. The first Case have been evaluated are schematically shown in Fig. 1 below.
Study evaluates the possibility of producing hydrogen at the The HRS system configuration shown in the upper part of
hydroelectric power plant Rotnes in Nittedal, Norway and Fig. 1 is designed for a dual-use hydrogen station serving both
builds on a report published by Institute for Energy Technol- passenger cars with on-board tanks at 700 bar and heavy duty
ogy in 2017 [4]. The economic feasibility of establishing and vehicles with tanks at 350 bar. This refueling system is rele-
operating a local hydrogen production facility by a small-scale vant for Case Study 1 addressed in Chapter 3.1 (hydrogen
hydroelectric plant with a maximum power output of 200 kW production from a small-scale hydroelectric power plant, with
and a 60% utilization rate is addressed. The cost efficiency of or without local HRS) and includes a dry running piston-type
installing a 700 bar HRS at the plant is furthermore compared compressor which feeds hydrogen from the medium pres-
to selling compressed hydrogen to a nearby station, and sure storage bank to a high-pressure storage at about 900 bar.
hydrogen cost sensitivity analyses of spot price variations and The HRS system configuration shown in the bottom part of
public support on CAPEX are carried out. Fig. 1 shows a refueling system designed for heavy duty ve-
The second Case Study focuses more on the end users’ hicles refueling at 350 bar and is the one considered in Case
perspective as it addresses the techno-economics of a HRS Study 2 (HRS for trucks, with or without on-site hydrogen
dedicated to a fleet of heavy duty vehicles. In this case the production). When the gas in this case is produced by an on-
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1 1203

Fig. 1 e The hydrogen supply chains and key system components evaluated in the Case Studies.

site WE-unit, only one compressor (directing the gas to a hence the intrinsic cost-inefficiency of small-scale systems
450 bar storage) is needed. such as those studied herein. Note that for an intermittent
The standards for hydrogen refueling of passenger vehicles energy source (such as the hydroelectric power plant consid-
(SAE J2601) prescribes the precooling of hydrogen to - 40
C to ered in Case Study 1), water electrolysis technology based on
allow for fast fueling into a 700 bar storage without over- polymer electrolyte membranes (PEM) is preferable over the
heating the vehicle's tank and thus ensure a more complete alkaline technology due to the superior ability to handle var-
fill. This pre-cooling, which requires additional equipment iable power production. In general, the PEM-technology has a
and energy [20], is not necessary for the slow refueling (flow higher specific investment cost than alkaline water electro-
rate  1.8 kg/min [19]) of hydrogen at 350 bar. Hence, the lyzer technology [24], but the collection of quotes from key
complexity and energy requirement for the considered 350 bar water electrolyzer providers interestingly revealed that for the
systems are lower than that for the 700 bar systems. smallest systems, the cost difference between the two tech-
When the hydrogen refueling stations are supplied by nologies is marginal (c.f. orange vs. blue symbols in Fig. 2).
trailers from off-site hydrogen production plants (a supply For the hydrogen pressure vessels, we assume storage
scheme which is evaluated in both Case Studies), the gas is costs of 6300, 8100 and 19800 NOK/kg for tanks rated for 250,
assumed to be transported according to the X-store concept of 450 and 900 bar, respectively. For the dispenser we assume a
Hexagon xperion [21]. This comprises a module of composite cost of 1.17 million NOK for the 350 bar-system and 1.62
tanks with 345 kg of hydrogen at 250 bar which is unloaded at million NOK for the 700 bar-system.
the station in exchange of an empty module. The advantage of The economic assumptions made in the system simulations
composite tanks is that they weigh much less than steel tanks are summarized in Table 1. Project costs such as planning, civil
and are therefore significantly easier to handle and transport. works, connecting to water and electricity etc. are included in
The transport of hydrogen from the production site to the the installation costs, and assumed to be 10% of the CAPEX. The
refueling station is covered by the regulations of overland project lifetime is set to 10 years, which corresponds to the
transport of dangerous cargo, and the hourly rate for transport estimated lifetime of a water electrolyzer stack [27,28]. Sensi-
of compressed hydrogen has been estimated by Oslo Trans- tivity analyses shows that if a payback time of only 7 years had
portsentral to be around 1200 NOK [22]. By assuming that one been assumed, the hydrogen refueling costs (NOK/kg) would
assignment (including loading, transport and unloading) takes increase by 20e30% compared to those presented herein. An
5 h, the total cost of transport will amount to 6000 NOK (excl.
VAT and road toll) and the specific cost for a trailer with a
payload of 345 kg hydrogen will thus be 17 NOK/kg H2.
The techno-economic models employed in this study are
based on up-to-date technical performance data obtained
from leading water electrolyzer companies and other
hydrogen technology suppliers around the world, as well as
cost data collected in various projects conducted at IFE over
the past few years. All figures have been controlled against
data known from public studies on water electrolysis [23,24]
and hydrogen stations [17,18]. All costs are furthermore pre-
sented in Norwegian kroner (NOK) using exchange rates of 8.5
NOK/USD and 9 NOK/EUR.
Figs. 2 and 3 below shows the specific capital costs of water
electrolyzers and hydrogen compressors, respectively, as a Fig. 2 e Specific capital costs of alkaline water electrolyzer
function of capacity. These data, based on quotations from (blue symbols) and PEM water electrolyzer systems (orange
suppliers, were collected in connection with Task 33 of IEA's symbols). The dashed line represents the best fit to the
Hydrogen Technology Program [25] and first presented at ICE alkaline WE cost data. (For interpretation of the references
2017 [26]. The figures highlight the importance of economies to colour in this figure legend, the reader is referred to the
of scales by moving from small to medium capacities, and Web version of this article.)
1204 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1

Fig. 3 e Specific capital costs of compressors for various input and output pressures.

interest rate variation of ±2% would on the other hand cause divided into four different power ranges from 0 to 200 kW. The
the hydrogen refueling cost to vary by ± 5%. number of operating hours for each effect range is listed in
parenthesis. The hydro plant produces power about 8000 h per
year, and by installing a PEM-based water electrolyzer system,
Results which can be operated continuously between 50 and 200 kW
(i.e. minimum part-load operation of 20e25% of nominal ca-
CASE STUDY 1: hydrogen production from a small-scale pacity assumed), about 7500 h of hydrogen production can be
hydroelectric power plant achieved annually.
The maximum power available from the hydroelectric
Power available for hydrogen production plant is about 180 kW and the installed production capacity is
The power produced at the hydroelectric power plant Rotnes therefore assumed to be 200 kW nominal power. The potential
in Nittedal in the period 2009e2015 is shown in Table 2 and for hydrogen production in the various power ranges is
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1 1205

to a nearby refueling station (Concept 1.1) and one in which an

Table 1 e Economic assumptions made in the system
HRS is installed on-site (Concept 1.2):
simulations.
Project lifetime (Years) 10
 Concept 1.1: Local hydrogen production and compression,
Interest rate 5%
without HRS
Stack lifetime (h) 90 000
Installation costs (of CAPEX) 10%  Concept 1.2: Local hydrogen production and compression,
Operation & Maintenance costs (of CAPEX) 4% with HRS
Redundancy on compressors 100%
Electricity represents the most important variable cost
when considering hydrogen production from water electrol-
depicted in Table 3 and has been calculated based on the ysis, and there are potential cost savings associated with
available power and the number of operating hours, assuming installing the hydrogen generator at the local power station
the average specific energy consumption for production and (ref. discussion in Section Methodology and assumptions).
compression of hydrogen is 87 kWh/kg H2 (from 82 kWh/kg at The cost of using local power for hydrogen production can in
200 kW to 108 kWh/kg at 50 kW). The expected average annual this case be considered the equivalent to the loss of income by
hydrogen production amounts to about 12 000 kg, corre- refraining from selling the electricity to the grid. The effect of
sponding to a water electrolyzer capacity utilization of 60% variations in spot price on the hydrogen production costs has
(1035 MWh/1750 MWh). A station with this capacity would be therefore been investigated: a spot price of 0.25 NOK/kWh
able to serve about 100e200 FCEVs [29]. (based on historical spot prices in this region of Norway [30]) is
compared to a hypothetical high tariff of 0.45 NOK/kWh that
Concepts and scenarios one may see in a few years’ time when the new transmission
The Rotnes hydroelectric power station has the advantage of lines from Norway to the European continent and the United
being located relatively close to existing hydrogen stations in Kingdom are completed.
the Oslo region, and distribution of pressurized hydrogen to In an early market phase, public support is considered
one of these stations may thus be a good option. Another necessary for the establishment of new hydrogen infrastruc-
possibility is to install a local HRS at the power plant. This will ture [12]. Projects such as the ones described in this study may
eliminate the need for distribution and transport of hydrogen, be eligible for support from various public organizations in
but at the same time increase the investments and operating Norway, for example Enova SF, which is a public enterprise
costs. A dual-use hydrogen station, serving both heavy-duty responsible for the promotion of environmentally friendly
trucks (350 bar) and passenger cars (700 bar), will give the production and consumption of energy. In 2017 Enova issued a
highest flexibility with respect to utilization. On this basis, two program to support the establishment of hydrogen infra-
business concepts for the small-scale hydroelectric plant is structure [31], and in the present study this is used to calculate
evaluated, namely one in which pressurized hydrogen is sold the levelized cost of hydrogen including public support on
CAPEX. The following sensitivity analyses wrt. electricity

Table 2 e The energy (MWh) available for hydrogen production at Rotnes between 2009 and 2015. The number of operating
hours in the respective effect ranges is given in parenthesis.
Year 0e49 kW 50e99 kW 100e149 kW 150e200 kW Total MWh and no. of operating hours
2009 23 (679) 106 (1512) 411 (3135) 504 (3101) 1044 (8427)
2010 18 (546) 178 (2435) 269 (2072) 572 (3462) 1037 (8515)
2011 22 (1003) 61 (832) 315 (2299) 611 (3656) 1008 (7790)
2012 6 (213) 44 (534) 544 (4112) 562 (3520) 1156 (8379)
2013 20 (855) 111 (1410) 338 (2551) 484 (2960) 954 (7776)
2014 14 (451) 80 (1016) 556 (4103) 456 (2821) 1107 (8391)
2015 23 (105) 23 (291) 382 (2805) 640 (3927) 1068 (7128)
Average 18 (550) 86 (1147) 402 (3011) 547 (3350) 1053 (8058)

Table 3 e Estimated hydrogen production (kg) for a plant with a maximum capacity of 60 kg/day (200 kW), based on
historical power production (2009e2015).
Year 0e49 kW 50e99 kW 100e149 kW 150e200 kW Total kg
2009 0 1071 4725 6072 11 868
2010 0 1799 3094 6886 11 779
2011 0 612 3621 7356 11 589
2012 0 441 6252 6773 13 466
2013 0 1125 3888 5835 10 848
2014 0 810 6388 5498 12 696
2015 0 234 4388 7715 12 337
Average 0 870 4622 6591 12 083
1206 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1

Fig. 4 e The rated capacity and energy consumption of supply chain key components in Case Study 1 (with on-site HRS).

prices and public support for the two different concepts in Concept 1.2 (right). The total CAPEX amounts to 8600 kNOK
Case Study 1 is therefore performed: and 11 800 kNOK, respectively. It should be noted here that the
dual pressures refueling system in Concept 1.2 includes the
A High electricity costs (0.45 NOK/kWh) with 50% public high-pressure storage tanks and the second compressor.
support on CAPEX The calculated levelized costs of hydrogen, assuming a
B High electricity costs (0.45 NOK/kWh) without public sup- utilization rate of 60%, are summarized for the different cases
port on CAPEX and scenarios in Table 4. For Concept 1.1 (the option without a
C Low electricity costs (0.25 NOK/kWh) with 50% public local HRS) the hydrogen cost varies between 83 and 125 NOK/
support on CAPEX kg, while when adding an on-site HRS, the cost ranges from
D Low electricity costs (0.25 NOK/kWh) without public sup- 105 to 157 NOK/kg. By comparison, today's retail price for
port on CAPEX hydrogen in Norway is fixed at 72 NOK/kg, excl. VAT (pump
price: 90 NOK/kg) to be gasoline-equivalent [32].
For Concept 1.1 one must keep in mind that to evaluate the
System description cost of actually refueling hydrogen, the transportation costs
Fig. 4 shows the graphical user interface of the modelling tool and the operating expenses of the refueling station must be
for the hydrogen supply system addressed in Case Study 1. It covered as well. Assuming weekly deliveries of 230 kg H2
depicts the rated capacities and energy consumption of the (corresponding to the average weekly production), this gives a
various components of a delivery system with an installed ca- transportation cost of 25 NOK/kg (ref. Section Methodology
pacity of 205 kW (corresponds to 60 kg/day at full utilization). and assumptions).
Note that the calculated system efficiency of 82 kWh/kg H2 in- According to the results shown in Table 4, it will be
cludes the energy consumption of a small PEMWE in a 15 feet impossible to sell hydrogen with a profit in Case Study 1 (and
container with all auxiliaries, as well as the energy required for similar cases with small-scale distributed renewable energy-
H2 compression by the diaphragm compressor with output based hydrogen production), particularly in an early market
pressure of 250 bar. The hydrogen flow is only 2.5 kg/h, signifi- when there are few vehicles available. Hence, one way to
cantly driving up the costs of the compressors as well (ref. Fig. 3). evaluate the potential profit is to look at the alternative
operating costs. In a commercial market with many fuel cell
Hydrogen costs vehicles and several HRS in operation we estimate that HRS
In Fig. 5 the contribution of the various system key compo- operators will need to charge at least 20 NOK/kg H2 to cover
nents to the CAPEX are shown for Concept 1.1 (left) and basic operating expenses (excluding cost of hydrogen). When

Fig. 5 e Cost breakdown of CAPEX (in kNOK) for the hydrogen production system considered in Case Study 1. The production
plant without HRS (left) is compared to a plant with HRS installed on-site (right).
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1 1207

Table 4 e Levelized cost of producing, compressing and refueling hydrogen at Rotnes power station.
Scenario A Scenario B Scenario C Scenario D
0.45 NOK/kWh, 0.45 NOK/kWh, no 0.25 NOK/kWh, 50% 0.25 NOK/kWh, no
50% public support public support public support public support
Concept 1.1 83 NOK/kg 125 NOK/kg 66 NOK/kg 109 NOK/kg
H2 for distribution
Concept 1.2 99 NOK/kg 157 NOK/kg 83 NOK/kg 141 NOK/kg
With local HRS

the hydrogen transport costs (25 NOK/kg) and HRS operating Finally, it should be noted that in the case of centralized
costs (20 NOK/kg) are included, the calculated levelized costs hydrogen production and bulk delivery of hydrogen (Concept
for the most favorable Scenario in Concept 1.1 (Scenario C at 2.2), it is assumed that all of the hydrogen comes from an off-
66 NOK/kg) reaches 111 NOK/kg, which is almost 40 NOK/kg site alkaline water electrolyzer plant with a hydrogen pro-
more than today's hydrogen retail price. duction cost of 30 NOK/kg. This cost is calculated based on a
standard industrial water electrolyzer plant with a hydrogen
CASE STUDY 2: HRS for small fleet of heavy-duty vehicles production rate of 1000 kg H2/day (rated capacity of 2.6 MW)
and full capacity utilization. In addition, the costs of trans-
System description porting hydrogen from the plant to the HRS is assumed to be
The feasibility of investing in an HRS for a small fleet of heavy- 17 NOK/kg (somewhat lower than 20 NOK/kg in Case Study 1
duty trucks is evaluated in this case study where two different due to more efficient hydrogen distribution systems). Hence,
HRS concepts are assessed and compared: Hydrogen from an the total cost of supplying hydrogen from a centralized
on-site water electrolyzer production unit (Concept 2.1) and hydrogen production plant to the HRS is 47 NOK/kg, which
hydrogen delivered by trailers (Concept 2.2). Fig. 6 illustrates adds to the OPEX of the hydrogen refueling station.
the system considered in Concept 2.1 and includes the rated
capacities and typical energy consumption of the key com- Hydrogen costs
ponents. In this example the hydrogen production and refu- Fig. 7 shows the breakdown of the capital costs for the key
eling facility has a daily turnover of 90 kg H2 (corresponding to components and sub-systems required in Concept 2.1 with
a fleet of 3 trucks, assuming each truck carries 30 kg H2 [13]). It on-site production (left) and Concept 2.2 with truck delivery
is furthermore assumed that unattended night fills take place (right). The total CAPEX for the two concepts is 9600 kNOK and
so that for Concept 2.1 the size of the hydrogen buffer storage 8500 kNOK, respectively. It is interesting to observe that the
can be minimized to only 50% of the daily consumption. In cost savings of the system without an on-site water electro-
Concept 2.2, on the other hand, the hydrogen buffer storage lyzer are nearly canceled out by the large and costly hydrogen
capacity is assumed to correspond to the payload of the de- storage based on composite pressure vessels. A comparison of
livery truck, i.e. 345 kg. Fig. 7 with Fig. 5 (right) shows that significant cost savings can
In Case Study 2, the hydrogen production capacity of the be made by switching from 700 bar to 350 bar refueling sys-
water electrolyzer is 100% utilized, as opposed to 60% in Case tems (the CAPEX of the 700 bar refueling system includes both
Study 1. This is based on the assumption that small-scale on- the high pressure storage and the extra compressor).
site water electrolyzer systems and corresponding hydrogen The bar chart in Fig. 8 shows the total annual CAPEX and
systems can be designed and dimensioned to meet the exact OPEX (values on the left-hand y-axis) as a function of daily
hydrogen demand for a given fleet of vehicles. Furthermore, turnover for the two concepts of on-site production and
the cost of electricity in this case includes both taxes and grid hydrogen delivery. The OPEX is divided into the three most
fees, albeit the power required for the water electrolysis is costly elements: hydrogen procurement, hydrogen transport
exempted from electricity tax (amounting to 0.166 NOK/kWh and electricity (see figure legend). The connected single points
in 2018) [33]. From this it can be assumed that the cost of (circles and triangles) plotted in the figure (values on the right-
power to run the auxiliary systems is 0.45 NOK/kWh, while the hand y-axis) show the levelized cost per kg dispensed
cost of power for water electrolysis is 0.30 NOK/kWh. hydrogen as a function of daily turnover for the two concepts.

Fig. 6 e The rated capacity and energy consumption of supply chain key components in Case Study 2.
1208 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1

Fig. 7 e Cost breakdown of CAPEX (in kNOK) for 350 bar HRS with on-site hydrogen production (left) and hydrogen delivered
by truck (right). The hydrogen turnover for both system concepts is 90 kg/day.

The total annual costs increase with increasing fleet size WE capacity (100% of rated power) and the power available at
(larger installed capacities and increased operating expenses), the local hydroelectric plant.
while the levelized costs of hydrogen decreases due to lower A similar point can be made for Case Study 2: If the daily
specific investment costs (Figs. 2 and 3) and lower relative hydrogen demand at a 300 kW (i.e. 90 kg/day) heavy duty fa-
energy consumption (energy per kilogram of hydrogen) for the cility is only 60 kg/day, the hydrogen refueling cost will in-
auxiliary systems. The results show that despite the relatively crease from 63 NOK/kg at full utilization to 85 NOK/kg at 66 %
low HRS investment costs for delivered hydrogen (Fig. 7, right), utilization. In this case, a better alternative would be to
the increased operating expenses associated with the pro- dimension the WE for the needed capacity (200 kW), as this
curement and transport of hydrogen (30 NOK/kg and 17 NOK/ would yield a hydrogen cost of only 68 NOK/kg.
kg, respectively) makes this option a less favorable business These results highlight the importance of achieving the
case overall compared to on-site hydrogen production. The highest possible utilization of the installed capacity in
results also show that in a fleet of 5 fuel cell trucks (a daily hydrogen stations. The results also show that although there
hydrogen turnover of 150 kg), the levelized hydrogen fueling are significant economic advantages associated with large
cost can reach 71 NOK/kg at an HRS based on delivered centralized hydrogen production plants, these benefits may
hydrogen and 58 NOK/kg at an HRS with on-site production. be drastically reduced if the plants are not fully utilized. In a
startup market it is therefore advisable to install smaller
refueling stations with on-site production units for customers
Summarizing discussion with a predictable and steady hydrogen demand (such as
fleets of heavy duty vehicles). Water electrolyzer systems are
Capacity utilization modular and the station capacity can be gradually increased
as the number of users grow.
The techno-economic analyses performed in the two case
studies described above show that the cost-efficiency of Comparison with fossil fuel costs
small-scale hydrogen supply systems is significantly influ-
enced by the scale of the hydrogen production plant and the It is important to keep in mind that the calculated cost of
refueling option (350 or 700 bar). Another important factor refueling hydrogen depends on the underlying economic and
that strongly affects the hydrogen production costs is the technical assumptions such as the interest rate, the project
utilization of installed capacity. The capacity utilization in lifetime and specific costs of key components, and the results
Case Study 1 and 2 is assumed to be 60% and 100%, respec- reported in this study must therefore be considered to be es-
tively. The consequences of this is discussed in some more timates. Nevertheless, it is interesting to compare hydrogen
detail below. cost estimates to fossil fuel options.
In Fig. 9 the levelized cost of refueling hydrogen is plotted As shown in Case Study 1 (local HRS for cars) the cost of
as a function of utilization rate for 350 and 700 bar refueling refueling hydrogen produced at the small-scale hydroelectric
systems. The figure shows that the cost of refueling hydrogen power plant (Rotnes) far exceeds the retail price for hydrogen
at the hydroelectric power station Rotnes with an installed (which is fixed at 72 NOK/kg, excl. VAT in order to be gasoline-
capacity of 200 kW drops from 141 NOK/kg at 60% utilization equivalent [32]) and will not be able to compete in today's
(Concept 1.2, Scenario D in Table 4)e93 NOK/kg in an ideal case market. In Case Study 2 (local HRS for fleet of trucks), on the
where the utilization rate is brought up to the theoretical other hand, there may be a larger chance for hydrogen to
maximum (100%). This can only be realized by buying elec- compete with conventional fuels, (in this case diesel). If it is
tricity from the grid to cover the gap between the full installed assumed that a trucking company gets a volume discounted
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1 1209

Fig. 8 e Annual CAPEX and OPEX (bar chart) and levelized cost of hydrogen (connected single points) as a function of station
turnover for the two HRS concepts considered in Case Study 2. Circle symbols represent on-site production and triangle
symbols truck-delivery of hydrogen.

diesel price of 11.13 NOK/liter [34] and that the average fuel
consumption is 3.3 L diesel/10 km [35] and 0.7 kg H2/10 km (39
and 55% energy efficiency, respectively), it can be estimated
that the cost of refueling hydrogen needs to be 52 NOK/kg or
less in order to be cost-competitive with diesel. This implies
that the system configurations assumed in Case Study 2 also
need to be scaled up to be cost-efficient (i.e., increase fleet of
trucks), and that public support probably will be needed in a
startup market, until economies of scale is reached.

Conclusions

The techno-economic calculations performed in this study

Fig. 9 e The levelized cost of hydrogen for different show that the hydrogen refueling cost at a small-scale hy-
refueling systems and capacities as a function of droelectric power plant in Norway (e.g., Rotnes) will signifi-
utilization rate. cantly exceed today's retail price for hydrogen. This is mainly
1210 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1

due to the small scale (200 kW) and low average utilization [6] Sundseth K, Møller-Holst S, Midthun K. Hydrogenproduksjon
rate (60%) of the load-following water electrolyzer. Even if the ved småkraftverk. Delprosjekt 2: flerbruk av hydrogen,
utilization rate is brought up to the theoretical maximum oksygen og varme ved Smolten settefiskanlegg. 2017. Report:
ISBN 978-82-410-1626-4. Retrieved from, http://
(100%), it will be difficult to develop economically viable
publikasjoner.nve.no/rapport/2017/rapport2017_73.pdf
hydrogen refueling systems without public support schemes. [Accessed 9 November 2018].
In comparison, a heavy duty HRS with an on-site water elec- [7] Møller-Holst S, Midthun K, Nørstebø V. Hydrogenproduksjon
trolysis unit serving a fleet of trucks seems to be a better ved småkraftverk. Delprosjekt 3: potensial for lønnsom
business case. This is because a predictable and steady utbygging av vassdrag i Rullestad. 2019. Report: ISBN 978-82-
hydrogen demand is ensured and because the investment and 410-1833-6. Retrieved from, http://publikasjoner.nve.no/
operational costs of a 350 bar refueling systems are lower than eksternrapport/2019/eksternrapport2019_10.pdf [Accessed 9
May 2019].
for 700 bar systems.
[8] Pütz K, Nørbech T. The way ahead for hydrogen in transport
The results in this study clearly show that although there in Norway. Which lessons can be learned from the
are significant economic advantages associated with large and successful implementation of battery electric vehicles?. 2012.
cost-efficient centralized hydrogen production plants, these Retrieved from, http://www.oecd.org/futures/ [Accessed 9
benefits may be drastically reduced if the plants are not fully May 2019].
utilized. The costs for transporting and distributing hydrogen [9] Melaina MW. Initiating hydrogen infrastructures:
preliminary analysis of a sufficient number of initial
from the centralized plant to the hydrogen station are
hydrogen stations in the US. Int J Hydrogen Energy
furthermore relatively high. The study shows that there is a
2003;28:743e55.
relatively good business case for local water electrolysis and [10] Tzimas E, Castello P, Peteves S. The evolution of size and cost
supply of hydrogen to captive fleets of trucks in Norway, of a hydrogen delivery infrastructure in Europe in the
particularly if the size of the fleet is sufficiently large to justify medium and long term. Int J Hydrogen Energy
the installation of a relatively large water electrolyzer system 2007;32:1369e80.
(economies of scale). The ideal concept would be a large fleet [11] Brey JJ, Brey R, Carazo AF. Eliciting preferences on the design
of hydrogen refueling infrastructure. Int J Hydrogen Energy
of heavy-duty vehicles (with a high total hydrogen demand)
2017;42:13382e8.
and a refueling station with nearly 100% utilization of the [12] Kang JE, Brown T, Recker WW, Samuelsen GS. Refueling
installed hydrogen production capacity. hydrogen fuel cell vehicles with 68 proposed refueling
stations in California: measuring deviations from daily travel
patterns. Int J Hydrogen Energy 2014;39:3444e9.
[13] Kim J-G, Kuby M. The deviation-flow refueling location
Acknowledgements model for optimizing a network of refueling stations. Int J
Hydrogen Energy 2012;37:5406e20.
The authors acknowledge The Norwegian Water Resources [14] Li Y, Cui F, Li L. An integrated optimization model for the
and Energy Directorate (NVE) for funding of the research location of hydrogen refueling stations. Int J Hydrogen
related to Case Study 1, and Jan Carsten Gjerløw for valuable Energy 2018;43:19636e49.
[15] Stephens-Romero SD, Brown TM, Kang JE, Recker WW,
input to the NVE-study. Acknowledgements are also given to
Samuelsen GS. Systematic planning to optimize investments
MoZEES, a Norwegian Centre for Environment-friendly Energy
in hydrogen infrastructure deployment. Int J Hydrogen
Research (FME), co-sponsored by the Research Council of Energy 2010;35:4652e67.
Norway (project number 257653) and 38 partners from [16] Sandia National Laboratories and NREL (National Renewable
research, industry and public sector. Energy Laboratory). Comparison of conventional vs. modular
hydrogen refueling stations, and on-site production vs.
delivery. 2017. Technical Report: SAND2017-2832. Retrieved
references from, https://www.energy.gov/sites/prod/files/2017/03/f34/
fcto-h2first-reference-station-phase2-2017.pdf [Accessed 9
May 2019].
[17] NREL (National Renewable Energy Laboratory). Hydrogen
[1] Stiller C, Bünger U, Møller-Holst S, Svensson AM,
station compression, storage, and dispensing technical
Espegren KA, Nowak M. Pathways to a hydrogen fuel
status and costs. 2014. Technical Report: NREL/BK-
infrastructure in Norway. Int J Hydrogen Energy
6A10e58564. Retrieved from, https://www.hydrogen.energy.
2010;35:2597e601.
gov/pdfs/58564.pdf [Accessed 9 May 2019].
[2] Robinius M, Otto A, Heuser P, Welder L, Syranidis K,
[18] Reddi K, Elgowainy A, Rustagi N, Gupta E. Impact of hydrogen
Ryberg DS, et al. Linking the power and transport
refueling configurations and market parameters on the
sectorsdPart 1: the principle of sector coupling. Energies
refueling cost of hydrogen. Int J Hydrogen Energy
2017;10:956.
2017;42:21855e65.
[3] Schiebahn S, Grube T, Robinius M, Tietze V, Kumar B,
[19] SAE (Society of Automotive Engineers). Fueling protocol for
Stolten D. Power to gas: technological overview, systems
gaseous hydrogen powered heavy duty vehicles. Surface
analysis and economic assessment for a case study in
vehicle standard. J2601/2. September 2014.
Germany. Int J Hydrogen Energy 2015;40:4285e94.
[20] Elgowainy A, Reddi K, Lee D-Y, Rustagi N, Gupta E. Techno-
[4] Ulleberg Ø, Gjerløw JC. Hydrogenproduksjon ved
economic and thermodynamic analysis of pre-cooling
småkraftverk. Delprosjekt 1: case studie Rotnes bruk. 2017.
systems at gaseous hydrogen refueling stations. Int J
Report: ISBN 978-82-410-1625-7. Retrieved from, http://
Hydrogen Energy 2017;42:29067e79.
publikasjoner.nve.no/rapport/2017/rapport2017_72.pdf
[21] Hexagon xperion. X-STORE gas container modules, version
[Accessed 9 November 2018].
ADR V2. 2017. Data Sheet. Retrieved from: https://www.
[5] Seljom P. Hydrogenproduksjon fra Rotnes Bruk. 2016. Report:
hexagonxperion.com/fileadmin/user_upload/xperion_
ISBN 978-82-7017-894-0.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 5 ( 2 0 2 0 ) 1 2 0 1 e1 2 1 1 1211

Energy___Environment/Datenblaetter_-_xperion_EE/ [29] Thomas CE. Fuel cell and battery electric vehicles compared.
XPERION-ENERGY-ENVIRONMENT-X-STORE-Gas_Container- Int J Hydrogen Energy 2009;34:6005e20.
Modules-Version-ADR-V2-Full-Carbon-Design-250Bar-H2. [30] Nordpool. Historical Market Data. Retrieved from: https://
pdf [Accessed 16 May 2019]. www.nordpoolgroup.com/historical-market-data/[Accessed
[22] Personal communication with jan carsten Gjerløw. 2017. 30 April 2019].
[23] Sa rl E4tech, Element Energy Ltd. Study on development of [31] ENOVA. Hydrogeninfrastruktur. Retrieved from https://www.
water electrolysis in the EU. 2014. Report retrieved from, enova.no/bedrift/landtransport/stotte-til-infrastruktur/
https://www.fch.europa.eu/sites/default/files/ hydrogeninfrastruktur/[Accessed 30 April 2019].
FCHJUElectrolysisStudy_FullReport%20(ID%20199214).pdf [32] Uno-X. Hydrogen: Spørsmål og svar. Retrieved from https://
[Accessed 5 May 2019]. unox.no/hydrogen/sporsmal-og-svar#4 [Accessed 30 April
[24] Saba SM, Müller M, Robinius M, Stolten D. The investment 2019].
costs of electrolysis e a comparison of cost studies from the [33] Lovdata. x 3-12-13. Avgiftsfritak for kraft som leveres til
past 30 years. Int J Hydrogen Energy 2018;43:1209e23. bestemte formål. Retrieved from: https://lovdata.no/
[25] IEA (International Energy Agency). Task 33: Local H2 supply dokument/SF/forskrift/2001-12-11-1451#KAPITTEL_3
For Energy Applications (2013-2016). Retrieved from: http:// [Accessed 12 November 2018].
ieahydrogen.org/Activities/Task-33.aspx [Accessed 30 April [34] Greensight. Hydrogen i tungtransporten. 2017. Report
2019]. retrieved from, https://www.greensight.no/wp-content/
[26] Ulleberg Ø. Techno-economic modeling of renewable energy uploads/sites/6/2017/07/Rapport-offentlig-versjon2-min-4.
hydrogen supply systems based on water electrolysis. 2017. pdf [Accessed 15 October 2017].
Conference presentation. Retrieved from, http://www. [35] The International Council on Clean Transportation. Fuel
ice2017.net/conference/presentations [Accessed 30 April efficiency technology in european heavy-duty vehicles:
2019]. baseline and potential for the 2020e2030 time frame. 2017.
[27] Schmidt O, Gambhir A, Staffell I, Hawkes A, Nelson J, Few S. White paper. Retrieved from, https://www.theicct.org/
Future cost and performance of water electrolysis: an expert publications/fuel-efficiency-technology-european-heavy-
elicitation study. Int J Hydrogen Energy 2017;42:30470e92. duty-vehicles-baseline-and-potential-2020 [Accessed 30
[28] Carmo M, Fritz DL, Mergel J, Stolten D. A comprehensive April 2019].
review on PEM water electrolysis. Int J Hydrogen Energy
2013;38:4901e34.