energies

Review
Review of the Liquid Hydrogen Storage Tank and Insulation
System for the High-Power Locomotive
Daehoon Kang 1, * , Sungho Yun 2 and Bo-kyong Kim 1

1 Smart Electrical & Signaling Division, Korea Railroad Research Institute, Uiwang 16105, Korea;
[email protected]
2 Railroad Safety Division, Korea Railroad Research Institute, Uiwang 16105, Korea; [email protected]
* Correspondence: [email protected]

Abstract: Hydrogen has been attracting attention as a fuel in the transportation sector to achieve
carbon neutrality. Hydrogen storage in liquid form is preferred in locomotives, ships, drones, and
aircraft, because these require high power but have limited space. However, liquid hydrogen must
be in a cryogenic state, wherein thermal insulation is a core problem. Inner materials, including
glass bubbles, multi-layer insulation (MLI), high vacuum, and vapor-cooled shields, are used for
thermal insulation. An analytic study is preferred and proceeds liquid hydrogen tanks due to safety
regulations in each country. This study reviewed the relevant literature for thermodynamic modeling.
The literature was divided into static, dynamic, and systematic studies. In summary, the authors
summarized the following future research needs: The optimal design of the structure, including
suspension, baffle, and insulation system, can be studied to minimize the boil-off gas (BOG). A
dynamic study of the pressure, mass flow, and vaporizer can be completed. The change of the
components arrangement from the conventional diesel–electric locomotive is necessary.

Keywords: liquid hydrogen; hydrogen tank; insulation; cryogenics

Citation: Kang, D.; Yun, S.; Kim, B.-k.

Review of the Liquid Hydrogen
Storage Tank and Insulation System 1. Introduction
for the High-Power Locomotive.
Mobile vehicles require independent energy carriers, for which alternatives to fossil
Energies 2022, 15, 4357. https://
fuels, such as batteries and hydrogen, are currently under research and development, as
doi.org/10.3390/en15124357
shown in Table 1. Electric propulsion systems using motors have been gradually introduced
Academic Editors: Hyun-Goo Kim into the large transportation sector, such as trains, ships, and aircraft. Therefore, the vehicles
and Youngkuk Kim can be arranged according to the required power. Conventional internal combustion
Received: 9 May 2022
engines (ICEs) or fuel cells can be viewed as power generators.
Accepted: 13 June 2022
Hydrogen can be stored in gaseous, liquid, or chemical forms [1]. Metal hydrate, which
Published: 15 June 2022
can store and take out hydrogen through a thermal reaction, has also been researched [2].
Liquid organic hydrogen carriers (LOHCs) and ammonia, which are chemical forms, have
Publisher’s Note: MDPI stays neutral
the advantages of safe use but are not suitable for vehicles, because they require large-
with regard to jurisdictional claims in
volume equipment to input energy for conversion. Gaseous hydrogen storage provides
published maps and institutional affil-
a fast response, but the energy content per weight and volume remains low, even if the
iations.
tank pressure is high (350–700 bar). The liquid hydrogen (LH2 ) form has the highest energy
density and can be easily converted to hydrogen gas through a vaporizer. Hence, LH2 ,
rather than gaseous hydrogen, could be preferred as a fuel for vehicles requiring a high-
Copyright: © 2022 by the authors.
speed rail or enormous power. The use of LH2 saves space and reduces the locomotive’s
Licensee MDPI, Basel, Switzerland. weight, allowing more cargo to be loaded.
This article is an open access article The battery’s energy density is 300–500 Wh/kg and 700–1000 Wh/L, and the energy
distributed under the terms and density of LH2 is 2000–2500 Wh/kg and 110–140 Wh/L. Therefore, for the same energy
conditions of the Creative Commons storage system (ESS), hydrogen is lighter in weight than the battery, and the battery is
Attribution (CC BY) license (https:// smaller in volume than hydrogen. Due to these characteristics, batteries are considered for
creativecommons.org/licenses/by/ low-powered vehicles with a relatively short mileage, whereas hydrogen is considered for
4.0/). high-powered vehicles with a long mileage.

Energies 2022, 15, 4357. https://doi.org/10.3390/en15124357 https://www.mdpi.com/journal/energies

Energies 2022, 15, 4357 2 of 13

Table 1. Applications and required power and energy storage amounts.

Application
Required Form of ESS
Transportation
Power Stationary
Vehicle Ship Aircraft Space Battery H2
1–3 kW House Bike Yacht Drone Shuttle ~100 kg -
~10 kg
10–100 kW Building Car Fishboat Cargo drone Station ~500 kg
(~100 km/kg)
Self-driving
River taxi Air mobility Base ~250 kg
200–400 kW Bus, Truck ~1 ton
Submarine Light-craft (Moon, Mars) (~5 km/kg)
Railcar
Ferry ~1 ton
1–5 MW Complex Locomotive Helicopter ~10 ton
Cruise (~1 km/kg)
5–20 MW Small city Cargo Airplane - ~10 ton

For example, hydrogen or battery trains can replace diesel trains in the train sector.
Using catenary lines is energetic and economically efficient, but the electrification rate of
the railway is not high worldwide. The electrification rate of Northeast Asia and Europe
is over 60%, but the rest of the Earth, including the Americas, Africa, and Australia, is
under 10%. As an alternative to the diesel trains operating on routes without catenary lines,
battery trains can be used as short-distance trams in cities, and hydrogen trains can be used
as multiple-unit or long-distance, high-powered locomotives.
The replacement of diesel trains can be helpful for international environmental reg-
ulations. In Korea, 265 diesel locomotives existed as of 2018. They emit air pollutants
equivalent to 300 commercial diesel vehicles. By converting diesel locomotives to hydrogen
locomotives, direct air pollutants (particulate matter (PM)) and indirect air pollutants (CO
and NOx ) are expected to be reduced.
Hydrogen trains have been actively studied since the commercialization by Alstom in
France in 2014, and their model ‘Coradia iLint’ was unveiled in 2016. Beyond the railcar
and tram, hydrogen locomotive and shunter have been researched by CP in Canada, CRRC
in China, CZ in the Czech Republic, PESA in Poland, Wabtec in the United States, and
Korean Railroad Research Institute (KRRI), as shown in Table 2. Especially, KRRI is working
on the world’s first LH2 locomotive.
Table 2. Recent cases of hydrogen trains.

Type Year Institution Power (Fuel Cell) Hydrogen Mileage Note

Railcar 2016 Alstom (France) 390 kW (250 kW) 250 kg 1000 km Commercial
2017 CRRC (China) 200 kW (200 kW) 12 kg 40 km Commercial
2021 KRRI (Korea) 200 kW (200 kW) 166 kg 600 km
2022 Siemens (Germany) 400 kW (400 kW) - 800 km
2022 (target) JR East (Japan) 240 kW (240 kW) 25 kg 140 km
2023 (target) Hyundai (Korea) 400 kW (400 kW) 40 kg 150 km Tram
Locomotive 2021 PESA (Poland) 600 kW (180 kW) 175 kg - Shunter
2021 CRRC (China) 700 kW (400 kW) - 627 km
2022 (target) CP (Canada) 1200 kW (1200 kW) - -
2023 (target) CZ LOKO (Czech) 800 kW (800 kW) - - Shunter
2024 (target) KRRI (Korea) 1800 kW (1200 kW) 70 kg - Liquid hydrogen
- Wabtec (USA) - - -

The study of LH2 storage came under the spotlight in earnest during the space race in
1957. The National Aeronautics and Space Administration (NASA) built the Saturn rocket
series using LH2 as fuel at the Kennedy Space Center (KSC) in Florida, and preliminary
tests were conducted at the Marshall Space Flight Center (MSFC) in Alabama. The LH2
 In addition to heat leakage, flashing, sloshing (especially after thermal stratificatio
ortho-to-parahydrogen conversion can cause boil-off [3]. This study focuses on hea
age in the static state.
Energies 2022, 15, 4357
Following the space race, MSFC researched long-term LH2 storage in the 1990s
3 of 13
than temporary storage for a rocket launch, with the NASA Glenn Research Center
and Lockheed Martin Space Systems [4,5]. The multi-purpose hydrogen test bed (M
included
tank requires an 18.09
a highmlevel
3 tank, insulated
of thermal with ato3.53-cm-thick
insulation avoid thermal spray-on foam insulation
loss. The significant
temperature
and differencemulti-layer
a 3.75-cm-thick between LH2insulation
(as low as 20 K) and
(MLI) the surrounding
with 45 aluminum atmosphere
blankets and s
induces boil-off gas (BOG), which leads to high pressure and the loss of stored hydrogen.
as shown in Figure 1. The MLI was simulated using layer-by-layer, Lockheed, and
In addition to heat leakage, flashing, sloshing (especially after thermal stratification), and
fied Lockheed models,
ortho-to-parahydrogen and the
conversion candifferences
cause boil-offin
[3].the methods
This study focusesare on
described
heat leakagein the fol
Section 2.3 state.
in the static [6]. A vacuum gap separated the outer and inner vessels, and the two
wereFollowing
fixed throughthe spacetherace,
ground-hold
MSFC researchedor orbit-hold
long-term LH method.
2 storage The results
in the 1990s,showed
rather a he
than temporary storage for a rocket launch, with the NASA
age of 0.22 W/m for the orbit-hold method and 305 K room temperature, which
2 Glenn Research Center (GRC)
and Lockheed Martin Space Systems [4,5]. The multi-purpose hydrogen test bed (MHTB)
less than when no MLI is used. The ground-hold method showed a heat leakag
included an 18.09 m3 tank, insulated with a 3.53-cm-thick spray-on foam insulation (SOFI)
and a ,3.75-cm-thick
W/m 2 which is 238 times that of the orbit-hold method. This result shows that th
multi-layer insulation (MLI) with 45 aluminum blankets and spacers,
conduction
as shown in Figureat the1. suspension betweenusing
The MLI was simulated the layer-by-layer,
two vessels,Lockheed,which isand themodified
structure that
Lockheed models, and the differences in the methods are described
tains the distance between the two vessels and fixes them, is more excellent than th in the following
Section 2.3 [6]. A vacuum gap separated the outer and inner vessels, and the two vessels
leakage of the wall surrounded by insulation schemes. Therefore, this result show
were fixed through the ground-hold or orbit-hold method. The results showed a heat
necessity
leakage of to study
0.22 W/mthe 2 formethod of fixing
the orbit-hold methodtwoandvessels withtemperature,
305 K room minimal structure
which is to avo
conduction.
50% less than Another
when no MLI example
is used.ofThelong-term
ground-holdLHmethod
2 storage othera than
showed NASAofis the fi
heat leakage
2 , which is 238 times that of the orbit-hold method. This result shows that the heat
fueling
63 W/mstation built at Munich Airport in Germany in 1999. However, the LH2 tan
conduction
not directlyat used
the suspension
to fuel LH between the two vessels, which is the structure that maintains
2, and a vaporizer was used to fuel the gas tank vehicle
the distance between the two vessels and fixes them, is more excellent than the heat leakage
Following the stationary LH2 storage, mobile LH2 vessels have been research
of the wall surrounded by insulation schemes. Therefore, this result showed the necessity
1999,
to studyBMW showed
the method the first
of fixing LH2 vehicle,
two vessels 750hl,
with minimal in Germany.
structure to avoid heatIn conduction.
2018, Hylium bu
first mobile
Another example refueling truck
of long-term LHin Korea.
2 storage In than
other 2021,NASALinde provided
is the LH2 tanks
first LH2 fueling stationto Hjelm
built at Munich
Ferry in Norway. Airport in Germany in 1999. However, the LH 2 tank was not directly used
to fuel LH2 , and a vaporizer was used to fuel the gas tank vehicles [7].

Figure 1. Insulation concept of the multi-purpose hydrogen test bed [4,5].

Figure 1. Insulation concept of the multi-purpose hydrogen test bed [4,5].
Following the stationary LH2 storage, mobile LH2 vessels have been researched. In
1999,This
BMWstudyshowed reviewed
the first LHthe literature
2 vehicle, 750hl,necessary
in Germany.for the system
In 2018, Hylium modeling
built the firstof LH2 s
tanks for application in the transportation sector. First, static modelingFerry
mobile refueling truck in Korea. In 2021, Linde provided LH 2 tanks to Hjelmeland studies, inc
in Norway.
heat transfer coefficients and thermal conductivity, were reviewed. Second, dy
This study reviewed the literature necessary for the system modeling of LH2 storage
modeling studies were
tanks for application in the reviewed, including
transportation phase
sector. First, staticchanges
modelingand values
studies, with chang
including
time. Third, systematic
heat transfer modeling
coefficients and thermalstudies for suitable
conductivity, sizing and
were reviewed. arrangement
Second, dynamic of the
modeling
system forstudies were were
a vehicle reviewed, includingLastly,
reviewed. phase changes
the futureand values withneeds
research changes oversumma
were
time. Third, systematic modeling studies for suitable sizing and arrangement of the power
system for a vehicle were reviewed. Lastly, the future research needs were summarized.
Energies 2022, 15, 4357 4 of 13

2. Static Modeling
2.1. Wall, Vacuum Gap, Suspension, and Sensors
Various tank shapes have been studied for ships carrying LNG, and the expression
“1st/2nd barrier of the tank” is used instead of “inner/outer vessel” of the LH2 tank. The
tanks are independent or integrated with the ship, which GTT operates in France.
Most materials become brittle under cryogenic conditions, such as stainless steel,
aluminum alloy, titanium alloy, and composite materials, whose behaviors have been
studied. Grade 316L stainless steel, which has a low carbon content, is commonly used [8].
Stainless steel has a thermal conductivity of approximately 10 W/(m·K) at 20 K in contact
with LH2 and approximately 16 W/(m·K) in contact with the air at room temperature. The
thermal resistance equation can represent the inner wall for conduction in a cylindrical
form [9].
Other materials, such as copper, which has a poorer insulation effect than steel, can be
considered for the inner vessel. The existing studies have only focused on the insulation
effect. Instead, the total efficiency should be considered when designing an LH2 tank for a
vehicle. The loss rate through vaporization during storage is approximately 0.1–5% per
day, but the loss rate during charging is approximately 20% per charge if the vessel is at
room temperature. Moreover, the storage period for a vehicle can be as short as a few days.
Therefore, an economic–energetic approach must be formulated to determine the optimal
point. The case of charging the vessel containing LH2 at cold temperatures instead of the
empty vessel at room temperature can also be considered.
The air layer in the gap between the inner and outer vessel walls can be configured as
the heat conduction of gases and heat radiation between solids. The outer vessel wall was
cooled via air convection, and radiation to the surroundings was assumed to be in the same
condition as the air. The emissivity of stainless steel is 0.6 (dimensionless), the convective
heat transfer coefficient of the air is approximately 2 W/(m2 ·K), and the Stefan–Boltzmann
number is 5.667 × 10−8 W/(m2 ·K4 ). The air and surrounding temperatures are generally
assumed to be equal.
The suspension structure connects the inner and outer walls, and heat conduction
appears. G10, G10-CR, and G11, which are fiberglass, can be used for suspension and
supports for small-sized stationary tanks [10]. Their thermal conductivity is approximately
0.2~0.3 W/(m·K). However, 316L stainless steel can be used for a big size of tank for
mobility, and it will enlarge the heat leakage.
Sensors also connect the inner and outer walls. Level, pressure, and temperature
sensors are typical. Their outer material can be 316L stainless steel, but the wire material can
be copper. For insulating the wire, the polyimide film can be used as the shell material, and
its thermal conductivity is approximately 0.1~0.7 W/(m·K) by product and temperature.

2.2. Inner Materials

As the inner materials (IMs), porous foam, SOFI, fiber-reinforced plastic (FRP), aerogel,
glass bubble, and hollow glass microsphere (HGM) have been used. Polyurethane foam
has been used to insulate liquified natural gas (LNG), but a higher insulation performance
is required for the insulation of LH2 . A thickness of several meters is required to insulate
LH2 by using polyurethane foam.
Unlike the thermal conductivity of MLI, which changes significantly with the change
in the degree of vacuum, the vacuum pressure had little effect on that of SOFI. Fesmire
researched the effects of spray foam and rigid foam in non-vacuum conditions for the
insulation of the rocket fuel tanks [11]. Johnson et al. tested the heat flux according to the
vacuum pressure in MLI for the long-term storage of liquid methane on the lunar surface
and suggested a correlation formula [12].
SOFI has a thermal conductivity of 10−3 –10−2 W/m·K and can be simplified to a
hypothetical layer in which the thermal conductivity ranges from 0.005 to 0.04 W/(m·K).
Tseng et al. measured the thermal conductivity of polyurethane foam and suggested the
shape of the test chamber and the experimental method [13]. Mekonnen et al. described
, x FOR PEER REVIEW 5 of 14
et al. measured the thermal conductivity of polyurethane foam and suggested the shape
of the test chamber and the experimental method [13]. Mekonnen et al. described the pro-
cess from fabrication to coating aerogels and showed the thermal conductivity range [14]
Energies 2022,et
15,al.
4357measured theMLI thermal
showsconductivity
a 35–50% better of polyurethane foam and than
insulation performance suggested the shape
SOFI [15]. 5 of 13 HGM
By contrast,
of the test chamber and the conductivity
has a thermal experimentalof method
10–3–10[13]. Mekonnen
–4 W/m·K, which et is
al.even
described
betterthe pro-
than MLI [16,17]. A
cess from fabrication to coating aerogels and showed the thermal conductivity
group of microspheres is easily assumed to be the same size and uniform arrangement, as range [14].
MLI shows athe
shown
processbetter
35–50%
in Figure
from insulation
fabrication to
2. However, the
coating aerogels
performance
particle sizethan and
SOFI
is all
showed
[15]. By
different
thecontrast,
thermal conductivity
in an actualHGM
experiment, and
range [14].
has a thermaltheconductivity
accumulation of 10 –3 –10 –4 W/m·K, which is even better than MLI [16,17]. A
MLI shows is random.
a 35–50% better insulation performance than SOFI [15]. By contrast, HGM
group of microspheres is easily assumed to be the same size and uniform arrangement, as
has a thermal conductivity of 10−3 –10−4 W/m·K, which is even better than MLI [16,17]. A
shown in Figure group
2. However, the particle
of microspheres size
is easily is all different
assumed in ansize
to be the same actual experiment,
and uniform and as
arrangement,
the accumulationshownis random.
in Figure 2. However, the particle size is all different in an actual experiment, and
the accumulation is random.

Figure 2. Cubic arrangement of a hollow glass microsphere for insulation [16].

2.3. Multi-Layer Insulation

Figure 2. Cubic arrangement of a hollow glass microsphere for insulation [16].
Figure 2. Cubic arrangement of a hollow glass microsphere for insulation [16].
MLI and variable density MLI (VDMLI) were studied. MLI consists of a reflector with
2.3. Multi-Layer Insulation
low thermal emissivity and a spacer with low thermal conductivity. Spacers prevent re-
2.3. Multi-Layer Insulation
MLI and variable density MLI (VDMLI) were studied. MLI consists of a reflector
flectors from direct contact and avoid heat conduction. Alumina-coated polyester sheets
variablelow
with
MLI andalternating thermal
density
layers
emissivity
MLI (VDMLI)
of aluminum
andwere
a spacer with low
studied. MLIthermal conductivity.
fiber,consists
Spacers
of a reflector orwith
prevent
reflectors from direct contact andfoilavoidand glass
heat conduction. aluminum,
Alumina-coatedsilica, perlite
polyester particles
sheets,
low thermal were
emissivity and a
used for layers spacer
MLI [18]. with
Typically,low thermal conductivity. Spacers prevent re-
alternating of aluminum foilMylar or Kapton
and glass can be used
fiber, aluminum, silica,asoraperlite
reflector, and Dacron
particles,
flectors fromnets
direct
werecontact
can used
be used and
for MLI avoid
as the
[18]. heat
spacer conduction.
[19].Mylar
Typically, Alumina-coated
The external
or Kaptonfilm canused
can be polyester
be coated sheets,
by painting
as a reflector, or material
and Dacron
alternating layers of glass-reinforced
suchnets
as aluminum
can be used asfoil
theand
spacerglass
cloth. Anfiber,
[19]. aluminum,
The external
example of film
MLIcan silica,
for or perlite
beinsulating
coated particles,
by painting
spacecraftor material
is the same in
were used forFigure
MLI [18].
such3 as Typically, Mylar
glass-reinforced
[20]. or An
cloth. Kapton canofbe
example used
MLI for as a reflector,
insulating and Dacron
spacecraft is the same in
Figure 3 [20].
nets can be used as the spacer [19]. The external film can be coated by painting or material
such as glass-reinforced cloth. An example of MLI for insulating spacecraft is the same in
Figure 3 [20].

Figure 3. MLI configuration for insulating spacecraft [20].

Figure 3. MLI configuration for insulating spacecraft [20].
Previous studies have shown that the ratio of these heat flux components depends
Previous
on studies distribution
the temperature have shown that
[21]. the VD-MLI,
Thus, ratio of which
these allows
heat flux components
for the thickness ofdepends
Figure 3. MLI on
configuration
the
spacers for
temperature insulating
to be modified, spacecraft
distribution
provided a[21]. [20].
Thus,
better VD-MLI,
insulation which than
performance allows for the
the basic MLIthickness
[22]. of
Many
spacers tostudies have suggested
be modified, provided the aoptimization of the layer
better insulation density andthan
performance arrangement
the basicvia
MLI [22]
PreviousMany thermodynamic
studies havehave
studies shownmodeling
that the
suggested [23–25].
ratio
the Optimization
of these heat
optimization based
of the on
flux a combination
components
layer density depends
and of arrangement
foam and via
on the temperature MLI was also researched,
distribution [21]. and
Thus, the combination
VD-MLI, of
which foam and
allows MLI
for showed
the a 10–50%
thickness higher
of
thermodynamic modeling [23–25]. Optimization based on a combination of foam and MLI
performance than MLI alone [26,27].
spacers to bewas
modified,
alsoMLI provided aand
researched, better
the insulation performance than the basic MLI [22]. higher per-
can be simplified to combination of foam
a hypothetical layer and
in which MLI
the showed
thermal a 10–50%
conductivity varies
Many studiesformance
have suggested
than MLI
from 0.00001 the optimization
alone
to 0.072 W/(m [26,27]. of the layer density and arrangement
·K), owing to material differences and pressure variations via [28].
thermodynamic modeling
MLIcan
MLI can [23–25].
bebe Optimization
simplified
simulated based
to a hypothetical
by layer-by-layer, onlayer
a combination
Lockheed, in modified
and ofLockheed
which the foam and
thermal MLI The
conductivity
models. var-
was also researched,
ies from and
Lockheed the combination
model
0.00001 tois0.072 of foam
based W/(m·K), and MLI
on a semi-empirical
owing showed
formula
to and can
material a quickly
10–50%
differences higher
solve
andthe per- heat
overall
pressure variations
formance than MLI
[28]. alone
MLI can[26,27].
be simulated by layer-by-layer, Lockheed, and modified Lockheed models
MLI can be simplified to a hypothetical layer in which the thermal conductivity var-
ies from 0.00001 to 0.072 W/(m·K), owing to material differences and pressure variations
[28]. MLI can be simulated by layer-by-layer, Lockheed, and modified Lockheed models.
 Energies 2022, 15, x FOR PEER REVIEW 6 of 14

Energies 2022, 15, 4357 6 of 13

The Lockheed model is based on a semi-empirical formula and can quickly solve the over-
all heat flux, but the selection of empirical parameters depends on the conditions. By con-
flux, but the selection of empirical parameters depends on the conditions. By contrast, the
trast, the layer-by-layer method is based on an iterative algorithm and can calculate the
layer-by-layer method is based on an iterative algorithm and can calculate the temperature
temperature of each layer in MLI. The layer-by-layer method divides the heat transfer into
of each layer in MLI. The layer-by-layer method divides the heat transfer into three forms:
three forms: solid heat conduction, gaseous heat conduction, and radiation.
solid heat conduction, gaseous heat conduction, and radiation.
2.4.Phase
2.4. PhaseChange
Change of of Liquid
Liquid Hydrogen
Hydrogen
Theheat
The heattransfer
transfer coefficient
coefficient (HTC),
(HTC), critical
critical heat
heat flux
flux (CHF),
(CHF), andand thermal
thermal conductivity
conductivity
of the hydrogen have been studied. Kumar et al. solved a homogeneous two-phase
of the hydrogen have been studied. Kumar et al. solved a homogeneous two-phase model model
and showed
and showed that that the
the effect
effect ofof evaporation
evaporation on on stratification
stratification increases
increases as as the
the aspect
aspect ratio,
ratio,
whichmeans
which means the the ratio
ratio ofof width
width andand height
height of of the
the tanks,
tanks, increases
increases by by Ansys
Ansys Fluent
Fluent [29].
[29].
Horieet
Horie etal.
al. experimentally
experimentally summarized
summarized the the film
film boiling
boiling heat heat transfer
transfer coefficient
coefficient of of LH
LH22,,
accordingtotothe
according thepressure
pressureand and thethe subcooling
subcooling temperature
temperature [30].
[30]. Baldwin
Baldwin et reviewed
et al. al. reviewed the
the relations
relations related
related to the to pool
the pool nucleate
nucleate boiling
boiling andandpoolpool
filmfilm boiling
boiling of LHof 2LH 2, liquid
, liquid oxy-
oxygen,
gen,methane
and and methane [31]. Ganesan
[31]. Ganesan et al. et al. derived
derived a universal
a universal CHFCHF correlation
correlation fromfrom a cryo-
a cryogenic
genic flow database from 1959 at Purdue University Laboratory
flow database from 1959 at Purdue University Laboratory [32]. Kuang et al. summarized [32]. Kuang et al. summa-
rized
the the nucleate
nucleate flow boiling
flow boiling HTC of HTCtheofLH the LH2suggested
2 and and suggested the artificial
the artificial neural neural
networknetworkthat
thatderive
can can derive
HTC [33].HTCWang [33]. Wang et al. performed
et al. performed a numerical
a numerical study of study
pool of pool
film film specific
boiling boiling
specific
to to the horizontal
the horizontal cylinder cylinder
[34]. [34].
Kang et al.
Kang al.suggested
suggested a 0-D
a 0-Dquasi-steady
quasi-steady thermal network
thermal model,model,
network as shown as inshown
Figurein 4
[35]. In4Figure
Figure [35]. In4,Figure
unlike 4, other studies
unlike otherthat modeled
studies that only
modeled the insulation system, thesystem,
only the insulation liquid
andliquid
the gaseousandhydrogen were assumed
gaseous hydrogen weretoassumed
be a single
to benode with node
a single mass,with
temperature,
mass, temper-pres-
sure, and volume. The thermal resistance diagram, including
ature, pressure, and volume. The thermal resistance diagram, including hydrogen, can hydrogen, can be expanded
to expanded
be dynamic modeling
to dynamicwith venting,
modeling withfilling, andfilling,
venting, vaporizing, with level with
and vaporizing, and level
pressure
and
changes changes
pressure in the tank. in theOntank.
the other
On thehand,
otherthe study
hand, theof high-pressure
study vesselsvessels
of high-pressure with phase
with
changes
phase is similar
changes to the to
is similar steam generation
the steam heat pump
generation [36]. In[36].
heat pump the In
industrial sector,sector,
the industrial high-
temperature and high-pressure
high-temperature and high-pressure steam are used,
steam and heat
are used, and heatpumps with with
pumps higher energy
higher effi-
energy
ciency than
efficiency thanboilers
boilersare introduced.
are introduced.InIna aflashflashtank,
tank,water
waterand andsteam
steamcoexist,
coexist,and and water
water is
addedas
added asmuch
muchas assteam
steamisisused.
used.

Figure4.4.Thermal
Figure Thermalresistance
resistancediagram
diagram of
of aa LH
LH22 insulation tank [35].

3.3.Dynamic
DynamicModeling
Modeling
3.1. Changes due to Phase Change of the Hydrogen
3.1. Changes due to Phase Change of the Hydrogen
One of the characteristics of LH2 is that its sensible heat is more significant than its
One of the characteristics of LH2 is that its sensible heat is more significant than its
vaporization heat until it reaches a gaseous state at room temperature. The ratio of the
vaporization heat until it reaches a gaseous state at room temperature. The ratio of the
two increases with the increase in pressure, and thus, the effect of the tank pressure was
two increases with the increase in pressure, and thus, the effect of the tank pressure was
studied. The ratio is 7.82 at 100 kPa and 23.70 at 1.2 MPa [37]. The enthalpy changes by
studied. The ratio is 7.82 at 100 kPa and 23.70 at 1.2 MPa [37]. The enthalpy changes by
3860 kJ/kg while vaporizing from –253 to 20 ◦ C, as shown in Table 3 [38]. The heat of the
3860 kJ/kg while vaporizing from –253 to 20 °C, as shown in Table 3 [38]. The heat of the
phase change and sensible heat of hydrogen can be obtained using REFPROP (version 10,
National Institute of Standards and Technology).
 Energies 2022, 15, x FOR PEER REVIEW 7 of 14

phase change and sensible heat of hydrogen can be obtained using REFPROP (version 10,
Energies 2022, 15, 4357 National Institute of Standards and Technology). 7 of 13

Table 3. Properties of hydrogen.

Table 3. Properties
Temperature of hydrogen.
(°C) –253 –253 0 20 A 70 B
State ◦ Liquid Vapor Vapor Vapor Vapor
Temperature ( C) –253 –253 0 20 A 70 B
Enthalpy (kJ/kg) 0 480 3580 3860 4580
State Liquid Vapor Vapor Vapor Vapor
A Room temperature. B Operating temperature of fuel cell.
Enthalpy (kJ/kg) 0 480 3580 3860 4580
A Room temperature. B Operating temperature of fuel cell.
A commercial electrolyzer requires 45–50 kWh of electricity to produce 1 kg of hy-
drogen, higher than
A commercial the higherrequires
electrolyzer heating 45–50
value kWh(HHV) of of hydrogen
electricity to (39.4 kWh/kg),
produce 1 kg ofand
hy-
emits 5–10
drogen, kWh/kg
higher thanofthe heat. A commercial
higher heating valuefuel (HHV)
cell generates 15–20 kWh/kg
of hydrogen of electricity
(39.4 kWh/kg), and
from 15–10
emits kg of hydrogen,
kWh/kg whichAiscommercial
of heat. lower than the fuelhydrogen’s lower
cell generates heating
15–20 value of
kWh/kg (LHV)
elec-
(33.4 kWh/kg)
tricity from 1 kg andof emits 13–18which
hydrogen, kWh/kg of heat.
is lower Thethe
than vaporization
hydrogen’sheat of LH
lower 2 is insuffi-
heating value
cient to(33.4
(LHV) coolkWh/kg)
the fuel cell,
andbut it can
emits helpkWh/kg
13–18 insulate oftheheat.
LH2 tank.
The vaporization heat of LH2 is
Additionally, the expansion ratio of the vaporization
insufficient to cool the fuel cell, but it can help insulate the LH is 2approximately
tank. 848, which
means that 100 kPathe
Additionally, LH2 can beratio
of expansion vaporized
of the to 172 MPa of is
vaporization vapor hydrogen. The
approximately 848,typical
which
fuel cells
means thatfor
100vehicles
kPa of use
LH2300cankPa–1.6 MPa oftogaseous
be vaporized 172 MPa hydrogen.
of vapor Therefore,
hydrogen.appropriate
The typical
pressure
fuel control
cells for is required
vehicles use 300from the vessel
kPa–1.6 MPa of to gaseous
the vaporizer and the
hydrogen. fuel cell inlet.
Therefore, appropriate
pressure control is required from the vessel to the vaporizer and the fuel cell inlet.
3.2. Vapor-Cooled Shield or Vent Line
3.2. Vapor-Cooled
A vapor-cooledShield or Vent
shield Lineis a vent pipe made from the same material as the inner
(VCS)
A vapor-cooled
vessel and located between shield (VCS) is a vent
the inner wallpipeand made
the endfromof the
MLI. same
For material
production as the inner
conven-
vessel
ience, and located
the VCS canbetween
be placed thejust
inner wall and
outside the the
innerendwall
of MLI. For production
by winding around convenience,
it. The VCS
the
canVCS can be placed
be simplified just outside
to a single layer and theconsidered
inner wallasbyintercepting
winding aroundthe heat it. penetration
The VCS can at
be simplified
the middle. to a single layer and considered as intercepting the heat penetration at
the middle.
The VCS, which involves recovering the sensible heat of the self-evaporated gaseous
The VCS,
hydrogen owingwhich involves
to heat recovering
penetration, is athe sensible heat
technology of the
at the self-evaporated
boundary between gaseous
passive
hydrogen
insulation owing to heat
and active penetration,
insulation. is a technology
The insulation effect ofatthe
theVCS
boundary
with LHbetween
2, nitrogen,passive
oxy-
insulation and active
gen, and methane was insulation.
studied usingThe insulation effect of against
a model validated the VCSthe with LH2 , nitrogen,
existing experimental oxy-
gen, and methane was studied using a model validated against
data [39]. Subsequently, the optimal location of the VCS within the foam and MLI and itsthe existing experimental
data [39].behavior
thermal Subsequently, the optimal
was studied location
[37,40,41]. In of the VCSanother
addition, within material,
the foam and liquidMLI and its
nitrogen,
thermal behavior was studied [37,40,41]. In addition, another
which has a higher boiling point than LH2, was studied for use as an insulation shield material, liquid nitrogen,
which
aroundhas theaLHhigher boiling point than LH2 , was studied for use as an insulation shield
2 tank [37].
around the LH tank
Based on 2the energy [37]. conservation, the heat transfer in the VCS is the same as the
difference in heat enteringconservation,
Based on the energy into and exiting theout
heatof transfer
the VCS,in asthe VCS is in
illustrated theFigure
same 5. asThe
the
difference in heatinto
heat penetration entering intotank
the inner and(Qexiting out of the VCS, as illustrated in Figure 5.
in) is assumed to be the latent heat of the vaporiza-
The
tion of hydrogen. In contrast, the heat transfer) through
heat penetration into the inner tank (Q in is assumed to be(Q
the VCS the latent heat of the
vcs) is assumed to be
vaporization of hydrogen. In contrast, the heat transfer through
sensible heat caused by the difference between the saturated vapor temperature the VCS (Qvcs ) is assumed
and the
to be sensible heat
outside temperature. caused by the difference between the saturated vapor temperature and
the outside temperature.

Figure5.5. The
Figure The example
example of
of thermal
thermal equilibrium
equilibrium and temperature nodes [31–33].

3.3. Boil-Off Ratio

The Boil-off Ratio (BOR) means the ratio of BOG generated in a day from the size of
the entire tank. The analysis conditions for the calculation of the BOR are shown in Figure 6.
For example, Jeon et al. predicted the BOR of the stationary tanks by a multiphase thermal
simulation [42]. A C-type liquid nitrogen tank and membrane LNG tank were analyzed by
the computational fluid dynamics (CFD) software STAR-CCM+, and the results showed
that heat conduction through suspension is significant.
 The Boil-off Ratio (BOR) means the ratio of BOG generated in a day from the size of
the entire tank. The analysis conditions for the calculation of the BOR are shown in Figure 6.
For example, Jeon et al. predicted the BOR of the stationary tanks by a multiphase thermal
simulation [42]. A C-type liquid nitrogen tank and membrane LNG tank were analyzed
Energies 2022,
by15,the
4357 computational fluid dynamics (CFD) software STAR-CCM+, and the results 8 of 13

showed that heat conduction through suspension is significant.

Figure 6. Analysis conditions: (a) shape,

Figure 6. Analysis (b) structure,
conditions: and
(a) shape, (b) (c) vibration
structure, (sloshing).
and (c) vibration (sloshing).

However, the mobile However,

vesselsthe for
mobile vessels show
transport for transport show
a higher a higher
BOR due BOR due to sloshing
to sloshing by by
vibration compared to the stationary vessels. Liu et al. did a series of studies on numerical
vibration compared to the stationary vessels. Liu et al. did a series of studies on numerical
models of dynamic meshes in sloshing conditions [43–45]. The authors used the software
models of dynamicICEM meshes in sloshing conditions [43–45]. The authors used the software
surf with the volume of fluid (VOF) method and applied the horizontal sinusoidal
ICEM surf with thevibration
volumetoofthe fluid
LH2(VOF) method
tank. Slosh and
baffles in applied the horizontal
the tank decreased sinusoidal
the sloshing, and BOR in-
vibration to the LH 2 tank.
creased Slosh
the bafflesthickness.
stratification in the tank
The decreased the sloshing,
authors additionally studiedand BORmomentum,
the force, in-
pressure,thickness.
creased the stratification and amplitude of the vibration
The authors in sloshing.
additionally studiedWei the
et al. analyzed
force, momen- the effect of
tum, pressure, andhorizontal
amplitude sinusoidal sloshing oninvertical
of the vibration tanksWei
sloshing. usingetthe
al. software
analyzed Fluent 17.0 (ANSYS
the effect of Inc.,
Pittsburgh, PA, USA) [46]. The authors showed that the pressure–time curve increases if
horizontal sinusoidal sloshing on vertical tanks using the software Fluent 17.0 (ANSYS
the vibration condition exceeds the critical amplitude and frequency.
Inc., Pittsburgh, PA, USA) Ghafri[46]. The authorsLHshowed
et al. summarized that the pressure–time curve in-
2 boil-off modeling and implemented it in the software
creases if the vibration condition
package BoilFASTexceeds
[47]. Thethe critical
software is amplitude
validated byand frequency.
multiple sets of industrially relevant
Ghafri et al. summarized
data, includingLH 2 boil-off
NASA, modeling
and provides and
static andimplemented
dynamic modeling. it in Besides,
the software
the pathway’s
package BoilFASTheat [47].penetration,
The software including pipes and
is validated cryopump,sets
by multiple wasofstudied. Petitpas
industrially et al. showed
relevant
data, including NASA,that the
andlosses
providesalongstatic
the pathway of LH2 modeling.
and dynamic could vary Besides,
by up to the 15%pathway’s
and calculated an
economic-energetic analysis [48].
heat penetration, including pipes and cryopump, was studied. Petitpas et al. showed that
the losses along the4.pathway
Systematic ofModeling
LH2 could vary by up to 15% and calculated an economic-
energetic analysis [48].
4.1. Capacity of the Power Required and the Fuel Tank
The capacity of the fuel cell and fuel tank are determined according to the purpose
4. Systematic Modeling
of the hydrogen vehicle. The typical capacity ranges of the hydrogen vehicle are listed in
Table 4. The Hyundai group in Korea modularized its fuel cells to 100 kW and produced
4.1. Capacity of the Power Required and the Fuel Tank
more than 20,000 fuel cells per year. Buses, trucks, and trams require two to four fuel cell
The capacity of the fuel
modules, cell and
whereas fuel tank
locomotives are determined
require according
one to two modules to of
for each thethepurpose
six wheels. With
the increase
of the hydrogen vehicle. The in the weight,
typical including
capacity rangescargo andhydrogen
of the passengers,vehicle
the fuelare
efficiency decreases,
listed in
Table 4. The Hyundai group in Korea modularized its fuel cells to 100 kW and produced to the
and the size of the fuel tank is determined to meet the required distance according
purpose of the vehicle.
more than 20,000 fuel cells per year. Buses, trucks, and trams require two to four fuel cell
modules, whereasTable
locomotives require
4. Typical capacity oneoftothetwo
ranges modules
fuel cells and fuelfor each
tanks of the vehicles.
of hydrogen six wheels.
With the increase in the weight, includingCar cargo and passengers,
Car
the fuel
Train
efficiency de-Train
creases, and the size of the fuel tank is determined
(Personal) A to meet the required
(Bus, Truck) B (Tram) C according
distance (Locomotive) D
to the purpose of the vehicle.
Fuel cell (kW) 100 100–200 200–400 2000–4000
Fuel tank (kg) 6–8 30–40 50–250 500–1000
Efficiency (km/kg) 90–100 5–20 1–3 1–2
Distance (km) 600–800 200–1000 200–1000 500–1000
A Hyundai Nexo (2018). B Hyundai Elec-city Bus (2018) and Xcient Truck (2020). C Alstom (2016), CRRC (2017),
and Hyundai-Rotem (2021). D KRRI (predicted value).

Apart from the vehicles, the LH2 tank for drones weighs 0.3–1 kg, the ground tank for
refueling stations is 20–450 kg, and the trailer tank for transport weighs 1000–3000 kg. The
Energies 2022, 15, 4357 9 of 13

Saturn V rocket used in the space race carried 69 metric tons of LH2 , but this is not relevant
to this study, because its tank was meant for immediate launch and not for insulated storage.

4.2. Arrangement of the Fuel Tank

Smith et al. designed a conceptual LH2 carrier ship from a conventional LNG carrier
ship [49]. Ship sizing, power consumption, including electrified propulsion system and
reliquifying the hydrogen system, and sloshing are reviewed. The authors predicted that
only 40.2% of energy could be loaded at the same volume, and the fuel loss rate could be
reduced by 38.7% through reliquefication.
Similarly, KRRI is considering the arrangement of the LH2 locomotive from a conven-
tional diesel locomotive. GE PowerHaul is a major diesel–electric locomotive with 2.8 MW
power since 2007, and it is called Class 70 in the UK, DE 36000 in Turkey, and Class 7600 in
Korea. Its standard subclasses include PH37ACi in mainland Europe and PH37ACmi in
the UK. A sample of its configuration is shown in Figure 7. The components related to the
engine and mass flow will be changed to the fuel cell and the hydrogen supply system. The
heating, ventilation, and air-conditioning (HVAC) system will remain. However, its sizing
and arrangement will be changed, because the amount of heat generation for power
Energies 2022, 15, x FOR PEER REVIEW gener-
10 of 14
ation (engine to fuel cell) is different. New technologies, including liquid- or phase-change
cooling, can be applied to decrease the volume of the HVAC system.

Figure 7. Sample of the configuration of the PH37ACi diesel–electric locomotive and components
Figure 7. Sample of the configuration of the PH37ACi diesel–electric locomotive and components
that will be changed.
that will be changed.
A conventional diesel fuel tank is located under the diesel engine at the bottom of the
A conventional diesel fuel tank is located under the diesel engine at the bottom of
vehicle. The side view of the tank is trapezoidal, and its capacity is 8600 L of diesel, as
the vehicle. The side view of the tank is trapezoidal, and its capacity is 8600 L of diesel, as
shown
shown inin Figure
Figure 8.
8.Firstly,
Firstly,the
thelocation
locationofofthe
thefuel
fueltank
tankcan
canbebeconsidered.
considered.Hydrogen
Hydrogenleaks
leaks
upwards in the case of fire, so it may be good to put the tank on top of the vehicle.
upwards in the case of fire, so it may be good to put the tank on top of the vehicle.
 that will be changed.

A conventional diesel fuel tank is located under the diesel engine at the bottom of
vehicle. The side view of the tank is trapezoidal, and its capacity is 8600 L of diesel
Energies 2022, 15, 4357 shown in Figure 8. Firstly, the location of the fuel tank can be considered. 10
Hydrogen
of 13 le
upwards in the case of fire, so it may be good to put the tank on top of the vehicle.

Figure 8. Fuel tank design: (a) size of a conventional diesel fuel tank at the bottom of the vehicle,
Figure 8. Fuel tank design: (a) size of a conventional diesel fuel tank at the bottom of the vehicle
(b) arrangement of 1-m-diameter tanks, and (c) arrangement of 0.72-m-diameter tanks.
arrangement of 1-m-diameter tanks, and (c) arrangement of 0.72-m-diameter tanks.
Secondly, the design of the LH2 tank can be considered. A vessel with a low surface-
Secondly,
to-volume the design
ratio shows a goodofinsulation
the LH2 effect.
tank can be considered.
However, the cylinderAtype
vessel with a to
is preferred low surfa
the sphere type because of the efficiency in space utilization. The cylindrical
to-volume ratio shows a good insulation effect. However, the cylinder type is preferredtank with a
diameter
the sphere of type
0.72 mbecause
and a length of 2.7
of the m has a capacity
efficiency in spaceofutilization.
500 L or 35 kg.
TheIf the tank spacetank wit
cylindrical
is a rectangular parallelepiped, 14 tanks (490 kg) can be stored. To enlarge the volumetric
diameter of 0.72 m and a length of 2.7 m has a capacity of 500 L or 35 kg. If the tank sp
and energetic efficiency, the diameter of the cylinder tank can be 1 m or more. A larger
tank can also offer better insulation (for example, 0.1 and 100 m3 tanks show boil-off ratios
of 2 and 0.06%, respectively [50]). The tank can be made in a spherical shape and placed
inside the vehicle to maximize the thermal insulation effect. David et al. briefly reviewed
the surface–volume ratio, boil-off ratio, and underlying assumptions of the LH2 tank [51].

4.3. Structure and Efficiency of the Fuel Tank

The LH2 tank design for aircraft and heavy-duty trucks (HDV) has been studied.
Gomez et al. suggest structural sizing and stress analysis for EASA CS-25 aircraft [52]. Abu
Kasim et al. suggest a performance and failure analysis for Cessna 208 aircraft [53]. Choi
et al. reviewed the ISO 13985 tank for HDV [54]. The authors suggested a design procedure
involving material selection and a structural and thermal analysis.
Thermodynamic analysis and CFD modeling of the LH2 tank venting under micrograv-
ity have been studied. Shuang et al. suggested pressure control by using the microgravity of
the tank, which showed that it can enhance the efficiency [55]. Jiang et al. did a CFD simula-
tion of the tank’s heat transfer and phase change under microgravity conditions. [56] Zheng
et al. also did a CFD simulation of the tank under microgravity, but the authors added the
dynamic results of the venting system [57]. Zuo et al. suggested a thermodynamic model
of the venting process of an on-orbit structure liquid hydrogen tank [58].

5. Future Research Needs

In future studies, the approaches, as follows, are predicted to be available:
(1) The optimal suspension design between vessels and baffles in the inner vessel can be
studied to minimize the BOG. Various CFD can be studied depending on the structure
and vibration direction.
(2) The optimal location and method of filling and drain can be studied. Depending
on the circumstances, the BOG in the charging process can be larger than the BOG
in storage.
Energies 2022, 15, 4357 11 of 13

(3) Materials and arrangement of the vessel wall and insulation system can be studied.
Former studies only focused on the insulation effect, but the optimal arrangement
can exist depending on the operation, including filling and draining. A 4E analysis
(Energy, Economic, Enthalpy, and Entropy) can be used.
(4) Dynamic control of the pressure, mass flow, and vaporizer (heat exchanger) system
before the fuel cell can be studied.
(5) Utilizing the heat of the fuel cell or battery for heating the hydrogen in the vaporizer
can also be studied.
(6) The arrangement of new components can be researched. For example, fuel cells and
battery modules require a large cooling capacity. Liquid- or phase-change cooling can
be applied.
(7) Various LH2 vessel shapes, locations, and arrangements can be studied. The charger
design is needed to research because of the large loss and BOG generation in the
filling process.

6. Conclusions
Heavy-duty mobilities, including trains, are expected to change their fuel from fossil
fuel to hydrogen, renewable energy for carbon-neutralization and solving global warming.
The use of liquid hydrogen is one way to solve the low energy density of gaseous hydrogen.
However, liquid hydrogen storage requires a high level of thermal insulation to minimize
the boil-off gas.
This paper shows the recent hydrogen train research trend and reviews the thermo-
dynamic modeling of the hydrogen tanks by dividing them into static, dynamic, and
systematic levels. In static studies, insulation schemes, including IMs, MLIs, vessels, and
vacuums, were researched. Thermal resistance diagrams involving hydrogen can be directly
applied to dynamic studies. Regarding dynamic studies, the phase change of hydrogen
was calculated. The shape, structure, and direction of the vibration of the tank were studied
to check the boil-off ratio. Finally, systematic studies reviewed the surface–volume ratio
and arrangement. The change of gravity and venting according to the arrangement also
improved the system efficiency.

Author Contributions: Methodology, data curation, validation, and writing of the original draft,
D.K.; conceptualization and investigation, S.Y.; and review and editing, B.-k.K. All authors have read
and agreed to the published version of the manuscript.
Funding: This research was funded by a grant from the R&D program (PK2203F1) of the Korea
Railroad Research Institute (KRRI).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Wijayanta, A.T.; Oda, T.; Purnomo, C.W.; Kashiwagi, T.; Aziz, M. Liquid hydrogen, methylcyclohexane, and ammonia as potential
hydrogen storage: Comparison review. Int. J. Hydrogen Energy 2019, 44, 15026–15044. [CrossRef]
2. Wijayanta, A.T.; Nakaso, K.; Aoki, T.; Kitazako, Y.; Fukai, J. Effect of pressure, composition and temperature characteristics on
thermal response and overall reaction rates in a metal hydride tank. Int. J. Hydrogen Energy 2011, 36, 3529–3536. [CrossRef]
3. Aziz, M. Liquid Hydrogen: A Review on Liquefaction, Storage, Transportation, and Safety. Energies 2021, 14, 5917. [CrossRef]
4. Hastings, L.J.; Hedayat, A.; Brown, T.M. Analytical Modeling and Test Correlation of Variable Density Multilayer Insulation for Cryogenic
Storage; NASA: Washington, DC, USA, 2004.
5. Hastings, L.J.; Martin, J.J. Experimental testing of a foam/multilayer insulation (FMLI) thermal control system (TCS) for use on a
cryogenic upper stage. AIP Conf. Proc. 1998, 420, 331–341.
6. Hedayat, A.; Hastings, L.J.; Brown, T. Analytical modeling of variable density multilayer insulation for cryogenic storage. AIP
Conf. Proc. 2002, 613, 1557–1564.
Energies 2022, 15, 4357 12 of 13

7. Pehr, K.; Sauermann, P.; Traeger, O.; Bracha, M. Liquid hydrogen for motor vehicles-the world’s first public LH2 filling station.
Int. J. Hydrogen Energy 2001, 26, 777–782. [CrossRef]
8. Qiu, Y.; Yang, H.; Tong, L.; Wang, L. Research progress of cryogenic materials for storage and transportation of liquid hydrogen.
Metals 2021, 11, 1101. [CrossRef]
9. Bergman, T.L.; Lavine, A.S.; Incropera, F.P.; DeWitt, D.P. Fundamentals of Heat and Mass Transfer; John Wiley & Sons: Hoboken, NJ,
USA, 2011.
10. Polinski, J. Materials in Cryogenics; European Course in Cryogenics; CERN: Geneva, Switzerland, 2010.
11. Fesmire, J.E. Layered composite thermal insulation system for nonvacuum cryogenic applications. Cryogenics 2016, 74, 154–165.
[CrossRef]
12. Johnson, W.; Fesmire, J. Thermal performance of low layer density multilayer insulation using liquid nitrogen. AIP Conf. Proc.
2012, 1434, 39–46.
13. Tseng, C.; Yamaguchi, M.; Ohmori, T. Thermal conductivity of polyurethane foams from room temperature to 20 K. Cryogenics
1997, 37, 305–312. [CrossRef]
14. Mekonnen, B.T.; Ding, W.; Liu, H.; Guo, S.; Pang, X.; Ding, Z.; Seid, M.H. Preparation of aerogel and its application progress in
coatings: A mini overview. J. Leather Sci. Eng. 2021, 3, 25. [CrossRef]
15. Liu, Z.; Li, Y.; Zhou, G. Insulation performance of foam during the terrestrial and ascent period. Appl. Therm. Eng. 2018, 145,
364–374. [CrossRef]
16. Wang, P.; Liao, B.; An, Z.; Yan, K.; Zhang, J. Measurement and calculation of cryogenic thermal conductivity of HGMs. Int. J. Heat
Mass Transf. 2019, 129, 591–598. [CrossRef]
17. Wang, P.; Ji, L.; Yuan, J.; An, Z.; Yan, K.; Zhang, J. Modeling and optimization of composite thermal insulation system with HGMs
and VDMLI for liquid hydrogen on orbit storage. Int. J. Hydrogen Energy 2020, 45, 7088–7097. [CrossRef]
18. Andersson, J.; Grönkvist, S. Large-scale storage of hydrogen. Int. J. Hydrogen Energy 2019, 44, 11901–11919. [CrossRef]
19. Wang, P.; Ji, L.; Yuan, J.; An, Z.; Yan, K.; Zhang, J. The influence of inner material with different average thermal conductivity on
the performance of whole insulation system for liquid hydrogen on orbit storage. Int. J. Hydrogen Energy 2021, 46, 10913–10923.
[CrossRef]
20. Kim, T.Y.; Lee, J.H. Electromagnetic Shielding Effectiveness of CFRP Panels, MLI, and RF Screens for Spacecraft Applications. J.
Korean Inst. Electromagn. Eng. Sci. 2020, 31, 947–955. [CrossRef]
21. Liu, Z.; Li, Y.; Xie, F.; Zhou, K. Thermal performance of foam/MLI for cryogenic liquid hydrogen tank during the ascent and on
orbit period. Appl. Therm. Eng. 2016, 98, 430–439. [CrossRef]
22. Hedayat, A.; Brown, T.M.; Hastings, L.J.; Martin, J. Variable density multilayer insulation for cryogenic storage. In Proceedings of
the 36th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit, Las Vegas, NV, USA, 24–28 July 2000.
23. Johnson, W.L. Thermal Performance of Cryogenic Multilayer Insulation at Various Layer Spacings. Master’s Thesis, University of
Central Florida, Orlando, FL, USA, 2010.
24. Johnson, W.L. Optimization of layer densities for multilayered insulation systems. AIP Conf. Proc. 2010, 1218, 804–811.
25. Wang, B.; Huang, Y.H.; Li, P.; Sun, P.J.; Chen, Z.C.; Wu, J.Y. Optimization of variable density multilayer insulation for cryogenic
application and experimental validation. Cryogenics 2016, 80, 154–163. [CrossRef]
26. Zheng, J.; Chen, L.; Cui, C.; Guo, J.; Zhu, W.; Zhou, Y.; Wang, J. Experimental study on composite insulation system of spray on
foam insulation and variable density multilayer insulation. Appl. Therm. Eng. 2018, 130, 161–168. [CrossRef]
27. Huang, Y.; Wang, B.; Zhou, S.; Wu, J.; Lei, G.; Li, P.; Sun, P. Modeling and experimental study on combination of foam and
variable density multilayer insulation for cryogen storage. Energy 2017, 123, 487–498. [CrossRef]
28. Fesmire, J.E. Standardization in cryogenic insulation systems testing and performance data. Phys. Procedia 2015, 67, 1089–1097.
[CrossRef]
29. Kuma, S.P.; Prasad, B.V.S.S.S.; Venkatarathnam, G.; Ramamurthi, K.; Murthy, S.S. Influence of surface evaporation on stratification
in liquid hydrogen tanks of different aspect ratios. Int. J. Hydrogen Energy 2007, 3, 21954–21960.
30. Horie, Y.; Shirai, Y.; Shiotsu, M.; Matsuzawa, T.; Yoneda, K.; Shigeta, H.; Tatsumoto, H.; Hata, K.; Naruo, Y.; Kobayashi, H.; et al.
Film boiling heat transfer properties of liquid hydrogen in natural convection. Phys. Procedia 2015, 67, 643–648. [CrossRef]
31. Baldwin, M.; Ghavami, A.; Ghiaasiaan, S.M.; Majumdar, A. Pool boiling in liquid hydrogen, liquid methane and liquid oxygen: A
review of available data and predictive tools. Cryogenics 2021, 115, 103240. [CrossRef]
32. Ganesan, V.; Patel, R.; Hartwig, J.; Mudawar, I. Universal Critical Heat Flux (CHF) Correlations for Cryogenic Flow Boiling in
Uniformly Heated Tubes. Int. J. Heat Mass Transf. 2021, 166, 120678. [CrossRef]
33. Kuang, Y.; Han, F.; Sun, L.; Zhuan, R.; Wang, W. Saturated hydrogen nucleate flow boiling heat transfer coefficients study based
on artificial neural network. Int. J. Heat Mass Transf. 2021, 175, 121406. [CrossRef]
34. Wang, J.; Li, Y.; Wang, L. Numerical Study on Pool Film Boiling of Liquid Hydrogen over Horizontal Cylinders. Energies 2022,
15, 1044. [CrossRef]
35. Kang, B.H.; Kim, M.S. Prediction of evaporation loss through integrated modeling of liquid hydrogen storage tank. In Proceedings
of the Thermal Engineering Division Spring Conference, Jeju, Korea, 20–22 April 2022; 22TE-Th06C12; Korean Society of Mechanical
Engineers: Seoul, Korea, 2022.
36. Kang, D.H.; Na, S.I.; Yoo, J.W.; Lee, J.H.; Kim, M.S. Experimental study on the performance of a steam generation heat pump with
the internal heat exchanging effect. Int. J. Refrig. 2019, 108, 154–162. [CrossRef]
Energies 2022, 15, 4357 13 of 13

37. Zheng, J.; Chen, L.; Wang, P.; Zhang, J.; Wang, J.; Zhou, Y. A novel cryogenic insulation system of hollow glass microspheres and
self-evaporation vapor-cooled shield for liquid hydrogen storage. Front. Energy 2020, 14, 570–577. [CrossRef]
38. Lemmon, E.W.; Bell, I.H.; Huber, M.L.; McLinden, M.O. NIST Standard Reference Database 23: Reference Fluid Thermodynamic and
Transport Properties-REFPROP; Version 10.0; National Institute of Standards and Technology, Standard Reference Data Program:
Gaithersburg, MD, USA, 2018.
39. Jiang, W.B.; Zuo, Z.Q.; Huang, Y.H.; Wang, B.; Sun, P.J.; Li, P. Coupling optimization of composite insulation and vapor-cooled
shield for on-orbit cryogenic storage tank. Cryogenics 2018, 96, 90–98. [CrossRef]
40. Jiang, W.; Sun, P.; Li, P.; Zuo, Z.; Huang, Y. Transient thermal behavior of multi-layer insulation coupled with vapor cooled shield
used for liquid hydrogen storage tank. Energy 2021, 231, 120859. [CrossRef]
41. Zheng, J.; Chen, L.; Wang, J.; Zhou, Y.; Wang, J. Thermodynamic modelling and optimization of self-evaporation vapor cooled
shield for liquid hydrogen storage tank. Energy Convers. Manag. 2019, 184, 74–82. [CrossRef]
42. Jeon, G.M.; Park, J.C.; Choi, S. Multiphase-thermal simulation on BOG/BOR estimation due to phase change in cryogenic liquid
storage tanks. Appl. Therm. Eng. 2021, 184, 116264. [CrossRef]
43. Liu, Z.; Feng, Y.; Lei, G.; Li, Y. Fluid thermal stratification in a non-isothermal liquid hydrogen tank under sloshing excitation. Int.
J. Hydrogen Energy 2018, 43, 22622–22635. [CrossRef]
44. Liu, Z.; Li, C. Influence of slosh baffles on thermodynamic performance in liquid hydrogen tank. J. Hazard. Mater. 2018, 346,
253–262. [CrossRef]
45. Liu, Z.; Feng, Y.; Lei, G.; Li, Y. Fluid sloshing dynamic performance in a liquid hydrogen tank. Int. J. Hydrogen Energy 2019, 44,
13885–13894. [CrossRef]
46. Wei, G.; Zhang, J. Numerical Study of the Filling Process of a Liquid Hydrogen Storage Tank under Different Sloshing Conditions.
Processes 2020, 8, 1020. [CrossRef]
47. Ghafri, S.Z.S.A.; Swanger, A.; Jusko, V.; Siahvashi, A.; Perez, F.; Johns, M.L.; May, E.F. Modelling of Liquid Hydrogen Boil-Off.
Energies 2022, 15, 1149. [CrossRef]
48. Petitpas, G. Boil-off Losses along LH2 Pathway; LLNL-TR-750685; Lawrence Livermore National Laboratory: Livermore, CA,
USA, 2018.
49. Smith, J.R.; Gkantonas, S.; Mastorakos, E. Modelling of Boil-Off and Sloshing Relevant to Future Liquid Hydrogen Carriers.
Energies 2022, 15, 2046. [CrossRef]
50. Godula-Jopek, A.; Jehle, W.; Wellnitz, Z. Hydrogen Storage Technologies: New Materials, Transport and Infrastructure; Wiley-VCH:
Weinheim, Germany, 2012.
51. David, B.; Stefania, G.; Simon, R.; Mari, V.; Yuki, I.; Petter, N. Liquid hydrogen as prospective energy carrier: A brief review and
discussion of underlying assumptions applied in value chain analysis. Renew. Sustain. Energy Rev. 2022, 154, 111772.
52. Gomez, A.; Smith, H. Liquid hydrogen fuel tanks for commercial aviation: Structural sizing and stress analysis. Aerosp. Sci.
Technol. 2019, 95, 105438. [CrossRef]
53. Abu Kasim, A.F.B.; Chan, M.S.C.; Marek, E.J. Performance and failure analysis of a retrofitted Cessna aircraft with a Fuel Cell
Power System fuelled with liquid hydrogen. J. Power Source 2022, 421, 230987. [CrossRef]
54. Choi, Y.; Kim, J.; Park, S.; Park, H.; Chang, D. Design and analysis of liquid hydrogen fuel tank for heavy duty truck. Int. J.
Hydrogen Energy 2022, 47, 14687–14702. [CrossRef]
55. Shuang, J.J.; Liu, Y.W. Efficiency analysis of depressurization process and pressure control strategies for liquid hydrogen storage
system in microgravity. Int. J. Hydrogen Energy 2019, 44, 15949–15961. [CrossRef]
56. Jiang, Y.; Yu, Y.; Wang, Z.; Zhang, S.; Cao, J. CFD simulation of heat transfer and phase change characteristics of the cryogenic
liquid hydrogen tank under microgravity conditions. Int. J. Hydrogen Energy 2022, in press. [CrossRef]
57. Zheng, Y.; Yang, P.; Liu, Y.; Yang, Q.; Yan, C.; Wang, X. Parametric optimization and analysis of thermodynamic venting system in
liquid hydrogen tank under microgravity. Int. J. Hydrogen Energy 2021, 46, 40041–40053. [CrossRef]
58. Zuo, Z.; Jiang, W.; Qin, X.; Huang, Y. Numerical investigation on full thermodynamic venting process of liquid hydrogen in an
on-orbit storage tank. Int. J. Hydrogen Energy 2020, 45, 27792–27805. [CrossRef]