International Journal of Hydrogen Energy 49 (2024) 1558–1571

Contents lists available at ScienceDirect

International Journal of Hydrogen Energy

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

Development of a condensing-type hydrogen liquefaction system for

improving cooling efficiency and long-term storage
Dong Woo Ha *, Hyun Woo Noh, Young Min Seo, Tae Hyung Koo, Rock Kil Ko
Hydrogen Electric Research Team, Korea Electrotechnology Research Institute, Changwon, 51543, Republic of Korea

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

Handling editor:Dr A Bhatnagar In this study, the experimental investigations for production of liquid hydrogen were conducted using a man­
ufactured condensing-type hydrogen liquefaction device. A direct-cooling type hydrogen liquefier was designed,
Keywords: fabricated, and tested using a commercially available cryo-coolers, heat pipe, ortho-para converter, LN2 shield,
Liquid hydrogen and LN2 pre-cooler. In order to effectively carry out the experimental investigations in the presence of two cryo-
Liquefaction system
coolers and LN2 shield in addition to the small-scale liquefaction system, this study introduced the new exper­
Cryo-coolers
imental system based on the cryogenic techniques. The production of liquid hydrogen could be achieved just by
LN2 shield
Long-term storage controlling the temperature and pressure inside the storage vessel. The cooling time of the vessel was reduced by
approximately 100 min using new experimental system. The total time required to produce the same amount of
liquid hydrogen, around 20L, was reduced by approximately 200 min. The experiments successfully demon­
strated the liquefaction of hydrogen gas in the vessel at a rate of approximately 3.5L/h, corresponding to a level
meter reading of 14.5 cm (6.95 % of the vessel’s height). Therefore, the cooling time of the vessel and the total
time required for liquid hydrogen production were found to be significantly influenced by the gas used to cool
the initial container and the presence of LN2 shield.

been offset by the harmful chemical effects of liquid hydrogen and

consequent restrictive handling constraints. Despite these limitations,
1. Introduction the liquid hydrogen fuel has been used extensively in the aerospace
industry to propel rockets, rocket boosters, missiles and space shuttles
Hydrogen is the simplest element on earth and it is an energy carrier, [7–9]. At standard boiling point, the liquid hydrogen has a density of
not an energy source. It can store and deliver useable energy, but it only 70.8 kg/m3, about 1/14 that of water, making it one of the lightest
doesn’t typically exist by itself in nature and must be produced from liquids. This liquid form of hydrogen is odorless and colorless, has a
compounds that contain it. The hydrogen can be used across different boiling point of − 252.87 ◦ C, and is a hazard for cold burns due to its cold
sectors to enable zero or near-zero emissions in other chemical and in­ temperatures. Due to its flammability, the liquid hydrogen should be
dustrial processes, integrated clean energy systems, and transportation. kept away from heat or flame unless ignition is intended [10,11].
Emerging hydrogen markets within these sectors include data centers, In line with the increasing interest in hydrogen, many researchers
ports, steel manufacturing, and medium- and heavy-duty trucks [1–4]. have been investigated the effects of various parameters on the lique­
Since the hydrogen is a carrier with a low volumetric energy density at faction systems in the cryogenic storage tank [12–15]. Additionally,
standard atmospheric pressure and temperature, it is usually stored as several studies have been conducted on small-scale hydrogen liquefac­
the compressed gas hydrogen at 35 MPa or 70 MPa or as the liquid tion systems. Baik and Notardonato [16] reported experimental results
hydrogen in cryogenic containers to increase the energy density. on hydrogen densification and zero boil-off storage testing. Baik et al.
Therefore, as the production and the usability of hydrogen increases, the [17] successfully designed and built a 1L/h scale hydrogen liquefier
proportion of liquid hydrogen become increases [5,6]. using a GM cryo-cooler. Garceau et al. [18] designed and fabricated a
Currently, the liquefaction processes for hydrogen gas are cost- GM cryo-cooler integrated hydrogen liquefier, low-loss LH2 transfer
intensive and require a significant amount of energy. Consequently, lines, an ortho-para hydrogen converter-integrated LN2 pre-cooler, and
mass production of liquid hydrogen has primarily been utilized in a low-loss 5L LH2 storage vessel. Xie et al. [19] designed a direction
cryogenic applications to minimize costs. However, this possibility has

* Corresponding author.
E-mail address: dwha@keri.re.kr (D.W. Ha).

https://doi.org/10.1016/j.ijhydene.2023.10.181
Received 6 July 2023; Received in revised form 21 September 2023; Accepted 18 October 2023
Available online 11 November 2023
0360-3199/© 2023 The Authors. Published by Elsevier Ltd on behalf of Hydrogen Energy Publications LLC. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

Table 1
Nomenclature Enthalpy calculation for cooling of liquid hydrogen depending on the presence
of LN2 pre-cooler.
Cold head-1 SD1 LN2 pre-cooler X LN2 pre-cooler O
Cold head-2 SD2 Sensible heat (300 K → 20 3500J/ Sensible heat (300 K → 80 3000J/
G10 Glass epoxy K) g K) g
GM Gifford-McMahon @ Cryo-cooler @ LN2 pre-cooler
Heat E.X - top SD3 Latent heat (80 K → 20 K) 500J/g
@ Cryo-cooler
Heat E.X - bottom SD4
Latent heat (gas → liquid) 448J/g Latent heat (gas → liquid) 448J/g
H2 Hydrogen gas @ Cryo-cooler @ Cryo-cooler
LH2 Liquid hydrogen Total enthalpy 3948J/ Total enthalpy 3948J/
LH2 Vessel-1 SD12 g g
LH2 Vessel-2 SD11 Amount of removed 3948J/ Amount of removed 948J/g
enthalpy @ Cryo-cooler g enthalpy @ Cryo-cooler
LH2 Vessel-3 SD10
LN2 Liquid nitrogen
MFC Mass flow controller experimental data in the development and operation of test equipment.
MFM Mass flow meter This study developed the new condensing-type hydrogen liquefac­
MLI Multi-layer insulation tion system that is simple and safe, encompassing scheme design, device
OFHC Oxygen-free high conductivity cooper development, operation processes, and safety measures with the aim of
S(SEN) Sensor enabling long-term storage of liquid hydrogen. In order to effectively
SD Silicon diodes carry out the experimental investigations in the presence of two cryo-
SEC Specific energy consumption coolers and LN2 shield in addition to the small-scale liquefaction sys­
SUS Steel use stainless tem, this study introduced the new experimental system based on the
cryogenic techniques. Experimental investigations have conducted
sequentially using the manufactured condensing-type hydrogen lique­
faction device to ascertain the production of liquid hydrogen. Further­
liquefaction scheme for micro-positive pressure with normal tempera­ more, this study investigated the production process of the liquid
ture hydrogen based on a two-stage GM cryo-cooler, and formulated the hydrogen on the temperature distribution inside the liquefaction system
hydrogen operation process and safety precautions. In order to produce and compared their performance depending on the experimental
the liquid hydrogen in the small-scale liquefaction system, many pre­ approach.
vious studies carried out their experimental investigations just by
applying the single cryo-cooler inside the liquid hydrogen storage vessel 2. Design of a condensing-type hydrogen liquefaction system
which includes the multi-layer insulations between inner and outer
walls. 2.1. Hydrogen liquefaction system
The cryocooler which has been verified at liquid hydrogen temper­
ature has advantages of long-life time, low level vibration and contin­ A key technology of all cryogenic vessels, including liquid hydrogen
uous refrigeration. In order to achieve the cooling capacity at liquid production equipment, is the insulation. It is important to block heat
hydrogen temperature, adequate capacity needs to be provided. Shen from the outside when designing a cryogenic container. For this reason,
et al. [20,21] and Chen [22] reported experimental and dynamic sim­ the liquid hydrogen container must be designed in consideration of the
ulations study of cryocooler working at liquid hydrogen temperature. cooling capacity of the freezer and the amount of liquid hydrogen to be
Additionally, the low boiling point, easy evaporation and difficult produced per hour.
storage of the liquid hydrogen limit its long-term use in orbit. Therefore, Table 1 shows the enthalpy calculation for cooling of liquid hydrogen
the research on the efficient thermal insulation technology of liquid depending on the presence of LN2 pre-cooler. When the LN2 pre-cooler is
hydrogen storage has become an important concern [23–26]. not used, the cryo-cooler must remove 3,500J/g of sensible heat and
In order to accurately produce the liquid hydrogen in the condensing 448J/g of latent heat, resulting in a total enthalpy of 3,948J/g. How­
type liquefaction system, the heat pipe which is connected through the ever, when the LN2 pre-cooler is employed, 3,000J/g of enthalpy from
cryo-cooler should be considered simultaneously. And then, the exper­ 300 K to 80 K can be removed by the LN2. In this case, the liquefaction is
imental investigation in the whole liquefaction system of the high vac­ achieved by removing only 948J/g of enthalpy. Therefore, it is very
uum jacketed storage tank including the multi-layer insulation should be efficient to use the LN2 pre-cooler in liquefying hydrogen in terms of
carried out. The generated amount of heat and the enthalpy of the time and cost.
hydrogen gas should also be calculated appropriately and should be In this study, a direct-cooling type hydrogen liquefier with a lique­
considered in the phase change for a sensible or latent heat. Addition­ faction rate of 3.5L/h using a commercially available system was
ally, in order to improve the cooling efficiency and the capacity in the designed, fabricated, and tested. The volume of liquid hydrogen is
hydrogen liquefaction system, two or more cryo-coolers should be approximately 1/780 of gaseous hydrogen. To produce 1L of liquid
considered simultaneously with a heat pipe. For the maximization of hydrogen, 780L of hydrogen gas at 300 K and 1 atm needs to be lique­
liquid hydrogen production, the use of multiple cryo-coolers has a great fied. The density of hydrogen gas at these conditions is approximately
advantage in terms of temperature and pressure control. Also, it is 0.08988 g/L. Thus, 70 g of hydrogen gas needs to be liquefied to produce
important to design the vessel using LN2 shield and to appropriately use 1L of liquid hydrogen. Considering the LN2 pre-cooler, the total enthalpy
catalysts with the working fluid for improving long-term storage capa­ required to liquefy 70 g of hydrogen gas is 66,360 J. To achieve the
bility of the liquefaction system. The LN2 shield prevents heat transfer by maximum output of 3.5L per hour, 245 g of hydrogen gas is required,
the conduction from the upper flange and the radiation between the resulting in a total enthalpy of 232,260 J for liquefaction. Since 1 W is
inner and outer walls, and consequently the temperature of liquid equivalent to 3,600J/h, a cooling capacity of approximately 65 W is
hydrogen storage vessel could be decreased to 80 K in advance. How­ necessary to produce 3.5L of liquid hydrogen per hour.
ever, there were relatively few studies on small-scale hydrogen lique­ Fig. 1 illustrates the overall appearance and the main components of
faction system with two cryo-coolers and LN2 shield in the storage the fabricated condensing type hydrogen liquefaction system. The liquid
vessel. In particular, there was a serious lack of experience and

1559
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

Fig. 1. Condensing-type hydrogen liquefaction system.

Fig. 2. Positions of temperature sensors.
hydrogen production vessel was manufactured based on the initial
design, and a helium leak detector was used to test for leaks after constructed using SUS316 to address hydrogen’s brittleness and miti­
fabrication. The model has two main compartments: a pre-cooler and a gate leakage issues caused by heat shrinkage. Support for fixing each
LH2 storage vessel as shown in Fig. 1(a). A heat pipe is incorporated part utilizes glass epoxy (G10), which has low thermal conductivity and
inside the vessel for the production of liquid hydrogen, and a pre-cooler no electrical conductivity. To prevent heat intrusion by convection, the
is installed on top of the vessel to enhance production efficiency. The container is designed as a double container with outer and inner walls,
liquid hydrogen container was designed with an upper flange structure, maintaining a vacuum state. Additionally, 20 to 40 layers of multi-layer
NW25 and NW40 ports, feed-throughs, and etc. The container was insulation (MLI) are utilized at various locations within the liquid

1560
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

Table 2 vessel neck space. A heat pipe is employed to increase the heat exchange
Detail information for position of temperature sensors. area of the cold heads of cryo-cooler. The heat pipe body is made of
Parameters Descriptions Parameters Descriptions stainless steel 304 (SUS304) with six oxygen-free high conductivity
copper (OFHC) fins on its external surface to increase the surface area
SD1 End parts of cryo- SD7 G10 bar at 10 cm intervals
SD2 coolers SD8 from bottom for convective heat transfer. The OFHC blocks are used at both ends to
SD3 Start and end parts of SD9 enhance heat transfer between the cryo-cooler cold head and the
SD4 heat pipe SD10 packed-bed of the ortho-para hydrogen converter. Further details on
SD5 Outlet of LH2 SD11 how to choose suitable ortho-para conversion reactor are described in
SD6 Pre-cooler SD12
Xu et al. [27], Qyyum et al. [28], Meng et al. [29] and Teng et al. [30].
As intended, the liquid hydrogen production vessel is a double
hydrogen container to minimize heat penetration by radiation. container consisting of an outer wall and an inner wall. In addition, the
Fig. 1(b) shows the main components inside the condensing-type LN2 shield is designed to minimize heat loss due to radiation between
hydrogen liquefaction system. Two cryo-coolers with a combined cool­ the inner and outer walls of the container in the present study. The LN2
ing capacity of about 90 W were utilized based on their specifications. shield prevents heat transfer by the conduction from the upper flange
To achieve the targeted production rate of 3.5L of liquid hydrogen per and the radiation between the inner and outer walls, and consequently
hour, the maximum heat penetration into the liquid hydrogen container the temperature of liquid hydrogen storage vessel which is made of
was designed to be below 15 W. The baffle and multi-layer insulation SUS316 is decreased to 80 K in advance. For this reason, the efficiency of
(MLI) were designed to block radiant heat, and the heat pipe was used hydrogen liquefaction is enhanced and the investment cost is reduced.
for liquefying hydrogen gas in the vessels. Multiple radiation baffles, LN2 level meters are also installed to monitor the water levels of LN2
consisting of thin plates covered with MLI, are installed in the inner shield.

Fig. 3. Positions of pressure sensors.

1561
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

The core technology for producing liquid hydrogen is to optimally

control the temperature and pressure inside the vessel. The temperature
monitoring inside the vessel is important for both control and safety
perspective. The pressure sensing and control throughout the entire
system are equally important because the pressure of the container is
directly related to safety.
Fig. 2 shows the position of temperature sensors for each port
attached to the upper flange. There are three NW 40 ports such as SEN1,
SEN2, and SEN3 in the upper flange of LH2 vessel. The 32-pin feed-
throughs connected these ports individually enable signal reception
from the temperature sensors, which provide information such as the
temperatures of the heat pipe connected to the cold head inside the
container as well as the injected hydrogen gas. The interior of the vessel
comprises various complex structures, including the insulation and the
heat pipe, which are critical for liquid hydrogen production. The tem­
perature measurements were performed using thermocouples posi­
tioned within the liquid hydrogen container. In the present study, the
silicon diodes (SD) sensors with a temperature measuring range of
approximately 1.4 K–500 K were employed. The total of 12 temperature
sensors from SD1 to SD12 were attached to different locations within the
vessel to measure the temperature during the production of liquid
hydrogen. Detailed information of locations for temperature sensors is
described in Table 2. The experimental investigations were conducted
using sensors that were verified at room temperature through liquid
nitrogen and liquid helium.
Fig. 3 demonstrates the installation of pressure sensors at different
positions. The explosion-proof certified pressure sensors were used for
checking pressure at each position to consider safety and control during
liquid hydrogen production. In this study, a verified pressure gauge was
used to measure from 14.5psi (1 bar) to the individual maximum value
of the pressure in 1 bar increments. The pressure sensors used in the
present study are from 1S to 4S, and the following roles is assigned to
each sensor:

1. Pressure sensor 1 (1S) (0~100psi): Measures hydrogen pressure

supplied to the mass flow controller (from gas storage).
2. Pressure sensor 2 (2S) (0~100psi): Measures hydrogen pressure
supplied to the vessel through the mass flow controller.
3. Pressure sensor 3 (3S) (0~870psi): Measures the pressure inside the
liquid hydrogen production vessel.
4. Pressure sensor 4 (4S) (0~30psi): Measures the pressure inside the
heat pipe.

2.2. Ortho-para hydrogen converter and LN2 pre-cooler Fig. 4. Cooling pipe in the pre-cooler.

The key factor in liquid hydrogen production and storage is the phase
change of hydrogen. At 300 K and 1 atm, the hydrogen exists in two Table 3
forms: ortho-hydrogen and para-hydrogen with a ratio of 75:25. When Design parameters of the cooling pipe in the pre-cooler.
the hydrogen is cooled to around 20 K to liquefy it, most of the ortho- Parameter Value Parameter Value
hydrogen converts to para-hydrogen, releasing a significant amount of Turn number 17 Pipe outer diameter [mm] 12.7
energy. This released energy exceeds 500J/g, which is approximately Turn diameter [mm] 136 Pipe inner diameter [mm] 9.3
15 % higher than the latent heat of vaporization of liquid hydrogen Height [mm] 230 Catalyst Iron(lll) oxide
(448J/g). Therefore, the ortho-para conversion using a catalyst should Length [m] 7.5 Catalyst volume [L] 0.51

be performed to prevent the rapid vaporization of the produced liquid

hydrogen. Commercially, an iron (III) oxide catalyst is commonly used in diameter, with a surface area of 0.06–0.07 mm2. The mesh size was
as catalysts based on their low cost and good performance in terms of determined based on the average catalyst size obtained from the optical
increasing the conversion rate. By referring to several previous papers, microscopic analysis.
we selected the iron (III) oxide catalyst considering excellent perfor­ Fig. 4 illustrates the cooling pipe which is made of copper inside the
mance, price, delivery time, and etc. pre-cooler. Detailed geometry and dimensional information for the
To increase the conversion rate from hydrogen gas to liquid cooling pipe are provided in Table 3. The pressurized test in the copper
hydrogen, the iron (III) oxide catalyst was injected into the LN2 pre- pipe of the pre-cooler was conducted to verify whether the catalyst
cooler. The catalyst was filled up to about half of the copper pipe in passes through the mesh. After checking the dimensions of the cooling
the pre-cooler, and a mesh smaller than the catalyst size was used to pipe in the pre-cooler, the amount of catalyst to be injected is deter­
prevent leakage. The iron (III) oxide catalyst had an irregular shape, and mined. In this case, 250 g of catalyst is injected, filling approximately
its size was determined through optical microscopic analysis to deter­ half of the pre-cooler. The mesh is installed inside the SUS union at the
mine the mesh size. The average catalyst size was approximately 200 μm

1562
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

end of the cooling pipe. And then, nitrogen gas is pressurized to test
whether the catalyst passes through the mesh. The gas flow is restricted
by increasing the nitrogen gas pressure from about 14.5psi to about
43.5psi. The test results confirm that no catalyst passes through the
mesh, and gas generation occurs even at the lowest pressure of 14.5psi
(1 bar).
Fig. 5 depicts the operating process according the injection of LN2
into the pre-cooler. During liquid hydrogen production, it is crucial to
cool the hydrogen gas from 300 K to 80 K in advance. To cool the
hydrogen gas, an industrial liquid nitrogen is supplied to the pre-cooler.
A calibrated level meter was used for monitoring the water levels of LN2
used in pre-cooling of hydrogen gas, and the supply begins when the
level is below 50 % and does not exceed 85 %. Heaters are also installed
inside the pre-cooler and the heat pipe, to refresh the iron oxide catalyst
during ortho-para conversion. The heaters are solely used for catalyst
refreshment, and the power is disconnected under normal operating
conditions. As mentioned, the use of a LN2 pre-cooler with the catalyst
significantly has reduced the heat load on the cryo-coolers.

2.3. Experimental setup

Fig. 6 presents a schematic diagram of the interconnected equipment

used in the liquid hydrogen production experiment. In this study, each
component was connected to facilitate liquid hydrogen production, and
the functionality of the manufactured equipment was verified under
normal operating conditions. The line supplying helium gas from the
Cryomech Inc. compressor cooling system was connected to two cryo-
coolers, which were attached to the liquid hydrogen production equip­
ment. Additionally, a line for supplying hydrogen from an external
hydrogen gas storage was connected to the gas controller using various
components such as valve, regulator pressure sensor, check valve, mass
Fig. 5. Operating process of pre-cooler.
flow meter (MFM), relief valve, pressure gauge, and etc. Valve and
sensor connections were established to control the gas injection amount
and monitor the temperature at various locations as specified in the
design drawings.

Fig. 6. Schematic diagram of the liquid hydrogen production process.

1563
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

Fig. 7. Liquid hydrogen production process using helium gas.

The cryo-cooler used in this study is designed to compress and relief valves and the check valves were installed in addition to the
expand helium gas to reach the cryogenic temperature of hydrogen gas. aforementioned components. Two relief valves were installed to prevent
Thus, a dedicated tool was used during assembly, to prevent the helium explosions by naturally venting excess pressure from the container in
leak during the connection of the cryo-coolers and the compressor situations such as power outages during liquid hydrogen production and
cooling system. The helium leak in the transfer line or other sections storage. Additionally, since 99.999 % ultra-high purity hydrogen is
could cause a decrease in pressure of the compressor cooling system. If required for producing liquid hydrogen, a check valve was installed to
the pressure of the helium gas drops too low, it leads to longer cooling prevent the injection of other gases, such as air, when supplying
times and decreased energy efficiency. If the pressure falls below a hydrogen. In general, the hydrogen gas used to produce liquid hydrogen
certain level, achieving the target temperature of 20 K becomes must be high purity hydrogen of 99.999 % or higher. At 20 K of lique­
impossible. Therefore, in the study, the low and high pressures of faction temperature for hydrogen gas, all the gas is solidified, except for
compressor cooling system were checked according to the operating helium gas. For this reason, the supplying gas used to produce liquid
mode in advance. When two compressor cooling systems are stopped, hydrogen is 99.999 % pure hydrogen gas. One relief valve was installed
the low and high pressures are compressed to about 210psi. On the other on the heat pipe with an operating pressure set at 85 psi, while the other
hand, when two compressor cooling systems are operated, the low and was installed in the H2 vessel where liquid hydrogen is stored, with an
high pressures are compressed to about 90psi and about 310psi, operating pressure of 10 psi.
respectively. This means that the cryo-coolers connected to the The valve opening sequence and the recommended pressure for each
compressor cooling systems can normally operate. section to ensure safety during liquid hydrogen production is visible in
The liquid hydrogen vessel in this system is made of stainless steel Fig. 6. During the production process, the heat exchanger (EX) line valve
316 (SUS 316), similar to other conventional cryogenic storage systems. and the H2 vessel line valve are shared and used through a T-union. The
As mentioned, the dedicated level meters are essential for monitoring initial conditions of pressure inside the heat pipe and H2 vessel are 15psi
the water levels of pre-cooling system, LN2 shield and the produced and 4psi. During the production of liquid hydrogen, the average pressure
liquid hydrogen, respectively. For this reason, the calibrated level me­ inside the heat pipe exceeds 30 psi, while the pressure inside the H2
ters are installed at the different positions such as LN2 pre-cooler, LN2 vessel is under 5 psi, resulting in a significant pressure difference be­
shield and H2 vessel to check water level as follows: tween two devices. Therefore, the heat exchanger line valve and the H2
vessel line valve should not be opened simultaneously.
① LN2 pre-cooler (N2 Level meter: 21.3 cm) The main features of the liquid hydrogen vessel manufactured in this
② LN2 shield (N2 Level meter: 120 cm) study are as follows:
③ H2 vessel (H2 Level meter: 100 cm)
○ Maximization of liquid hydrogen production by utilizing two units of
To produce liquid hydrogen, the instrument controllers and various cry-coolers
sensors were interconnected. Each component was connected to a ○ Design of a heat pipe for effective hydrogen liquefaction
dedicated controller to obtain experimental data from various sensors ○ Application of ortho-para converter (phase change) technology
such as temperature and pressure. Three liquid level instrument con­ ○ Design of an LN2 shield of the container to minimize heat loss caused
trollers were connected to control each of the three level meters. Two by radiation
Lake Shore 218 temperature monitors were used to read data from 12 ○ Design of an LN2 pre-cooler for efficient liquefaction of hydrogen gas
temperature sensors. Four controllers were installed to monitor the
attached pressure sensors at specific locations. A GMC 1200 controller 3. Experimental results and discussion
was connected to control the Mass Flow Controller (MFC), which reg­
ulates the amount of hydrogen gas supplied for liquid hydrogen pro­ 3.1. STEP 1: Use of helium gas for cooling inside LH2 storage vessel
duction. Two controllers were also installed to manage the heaters used
for refreshing the ortho-para converter. In the first liquid hydrogen production experiment, helium gas was
For safe operation of the liquid hydrogen production equipment, the used for initial cooling to ensure safety. This process involved that the

1564
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

Fig. 8. Temperatures and pressures during the pre-cooling process using helium gas.

temperature decreased to 20 K through the helium gas, and then the was operated, and hydrogen gas was injected into the heat exchanger
liquefaction of hydrogen gas was carried out by maintaining a vacuum while maintaining a pressure of 15psi. Simultaneously, helium gas was
and injecting hydrogen gas. The production of liquid hydrogen should injected into the hydrogen vessel at a pressure of 4psi. This resulted in a
be considered by controlling the temperature and pressure inside the monotonic decrease in temperature inside the heat exchanger and the
storage vessel. During the production of liquid hydrogen, various pa­ vessel. During the liquefaction process, once hydrogen gas began to
rameters such as hydrogen gas flow rate, temperature, pressure at liquefy in the heat exchanger, helium gas was evacuated using a vacuum
different locations, and water level were recorded during the pump in section (b), resulting in a rapid temperature decrease at SD6.
experiment. Subsequently, the supply of hydrogen gas to the heat exchanger was
Fig. 7 illustrates the overall process of liquid hydrogen production halted, and LN2 was introduced into the pre-cooler under vacuum con­
using helium gas. Based on the recorded data, the liquid hydrogen ditions in section (c). In this section, hydrogen gas was supplied to the
production process can be divided into two parts: the pre-cooling pro­ hydrogen vessel for liquid hydrogen production, leading to a slight
cess and the liquefaction process. During the pre-cooling process, the temperature increase inside the vessel. Simultaneously, the pressure
temperature inside the container decreases steadily to facilitate liquid difference generated by the ortho-para converter catalyst between 2S
hydrogen production. In contrast, the liquefaction process involves and 3S was checked. In the final part of section (c), all 1bt of liquid
injecting hydrogen gas into the cooled container to produce liquid hydrogen was utilized, and the supply of LN2 to the pre-cooler was
hydrogen. In the first liquid hydrogen production experiment, the heli­ stopped. In section (d), LN2 was reintroduced to the pre-cooler, and the
um gas was used in the cooling process due to its inert nature and low liquid hydrogen production was checked to calculate the liquefaction
liquefaction temperature of 4 K, ensuring safety. rate. Finally, the compressor cooling system was shut off after con­
The detailed experimental process is described below in four sections firming a constant temperature.
(a-d) as depicted in Fig. 7. In section (a), the compressor cooling system The pre-cooling process using helium gas involved approximately

Fig. 9. Temperature and hydrogen level meter during the liquefaction process using helium gas.

1565
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

Fig. 10. Liquid hydrogen production process using hydrogen gas.

450 min of heat exchange from 300 K to 20 K after operating the cryo- calculations based on the total amount of 300 K hydrogen gas injected
coolers. The hydrogen gas at 300 K room temperature is injected into the through the MFC (Mass Flow Controller) with the NIST-REFPROP
container after all the helium gas inside the container is removed database for hydrogen properties calculations. Typically, 1L of liquid
through the vacuum exhaust system for liquid hydrogen production and hydrogen is produced by liquefying 780L of hydrogen at 300 K and 1
subsequent liquefaction. As a result, the condensing-type liquefaction atm down to 20 K. However, the theoretical calculation method assumes
hydrogen system developed in this study produced approximately 20L of that all supplied hydrogen gas is completely liquefied, while in reality,
liquid hydrogen, despite the setbacks and various hardships. some gas remains in a cooled state without full liquefaction. Therefore, it
Fig. 8 presents the temperature and pressure at various positions in is necessary to exclude the amount of gas that is cooled to 20 K but still
the hydrogen liquefaction system during the pre-cooling process using exists in a non-liquefied state for a more accurate calculation.
helium gas. During the cooling of the liquefaction vessel for approxi­ Additionally, to assess the level of liquefied hydrogen, temperature
mately 450 min, the pressure in the heat exchanger was adjusted to a sensors were attached to the G10 bar at 10 cm intervals. The tempera­
maximum of 15psi. When helium gas was injected into the vessel for ture changes rapidly depending for the hydrogen gas on pressure,
heat exchange, the pressure in the vessel was adjusted to a maximum of whereas the temperature variation is uniform for the liquid hydrogen.
4psi. The temperature exhibited a rapid decrease at two cryo-coolers and By monitoring the temperature changes of the sensors at different lo­
the upper region of the heat pipe due to conduction, represented by the cations, it was possible to roughly estimate the amount of liquid
red region from SD1 to SD3. However, the temperature slowly decreased hydrogen produced inside the vessel. Finally, the most accurate method
at different locations within the vessel, depicted by the blue region from for checking the liquid hydrogen level is through the use of a level meter.
SD5 to SD12, except for SD6. The cooling temperature over time could By considering the inner volume of the container and the percentage
be monitored based on the compressor cooling system and external load indicated by the level meter, it becomes easier to determine the amount
conditions. It was observed that the condensing-type hydrogen lique­ of liquid hydrogen produced. In this study, the production of approxi­
faction system maintained stable cooling without an increase in vessel mately 20L of liquid hydrogen was confirmed using three methods.
pressure.
Fig. 9 illustrates the temperature and liquid level meter in the
liquefaction process using helium gas. After approximately 450 min of 3.2. STEP 2: Use of hydrogen gas for cooling inside LH2 storage vessel
the pre-cooling process, hydrogen gas was injected into the vessel, which
had been cooled to 20 K, to produce liquid hydrogen. The amount of In the primary liquid hydrogen production experiment, helium gas
hydrogen gas injected into the heat pipe was 50L, based on conditions of was chosen for heat exchange during the cooling process inside the
300 K and 1 atm, and the supply was then halted during the liquid vessel due to safety considerations. When the temperature inside the
hydrogen production. After helium gas was evacuated from the vessel vessel reaches 20 K, the liquefaction process must be carried out under
using a vacuum pump, liquid hydrogen was produced by supplying vacuum conditions. During the vacuum evacuation, the pressure inside
hydrogen gas cooled to 80 K through the LN2 pre-cooler. One such of the vessel becomes lower than the atmospheric pressure outside, and if
particular importance is the temperature measurement at SD12, which is there is a leak, other gases may enter the vessel, causing additional is­
closest to the bottom of the H2 vessel. This temperature remains rela­ sues such as mixing and freezing. Thus, taking these factors into account,
tively constant, with a variation of approximately 20 K. As a result, the in the secondary liquid hydrogen production experiment, hydrogen gas
reading of the H2 level meter steadily increases over time, enabling the was used instead of helium gas to cool the internal temperature of the
determination of the amount of liquid hydrogen produced as shown in vessel. Additionally, in the pre-cooling process, LN2 was injected into the
Fig. 9. Therefore, it was judged that the analysis of the hydrogen lever LN2 shield between the outer and inner walls of the vessel to reduce the
meter can be conducted with reasonable accuracy at over a certain level. cooling time.
The production of approximately 20L of liquid hydrogen was Fig. 10 presents an overview of the liquid hydrogen production
confirmed using three methods. The first method involved theoretical process using hydrogen gas. Similar to the primary experiment, the
production of liquid hydrogen should be considered by controlling the

1566
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

Fig. 11. Experimental data during the pre-cooling process using hydrogen gas.

temperature and pressure inside the storage vessel. The detailed In section (c), when the pre-cooling process is completed, 100L of
experimental process is divided into four sections from (a)–(d) as indi­ hydrogen gas is injected into the heat pipe. Simultaneously, LN2 starts to
cated in Fig. 10. supply at the pre-cooler, and the hydrogen gas starts to supply at the
In section (a), the compressor cooling system is operated. Hydrogen hydrogen vessel through the pre-cooler. Since hydrogen gas is used
gas is injected into the heat exchanger and the storage vessel, main­ instead of helium gas for cooling the internal temperature of the vessel,
taining pressures of 40 psi and 5 psi, respectively. The temperature the temperature stabilizes at around 20 K. From section (c) to (d),
gradually decreases inside the heat exchanger and the vessel, resulting hydrogen gas is continuously injected at a rate of approximately 45–47
in a cooling rate similar to that of helium gas. In section (b), LN2 is standard liters per minute (SLM) for liquid hydrogen production. The
injected into the LN2 shield between the outer and inner walls of the liquid hydrogen production is monitored, and the liquefaction rate is
vessel to shorten the pre-cooling time. As a result, the temperature in the calculated. Finally, the compressor cooling system is shut off while
secondary experiment using hydrogen gas decreases rapidly compared maintaining a constant temperature in section (d).
to the primary experiment using helium gas. Fig. 11 shows the gas flow rate, pressure, and temperature in the

1567
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

Fig. 12. Experimental data during the liquefaction process using hydrogen gas.

hydrogen liquefaction system on pre-cooling process using hydrogen heat exchange, the pressure in the vessel was adjusted to a maximum of
gas. Here, cold head-1 and cold head-2 represent SD1 and SD2, and heat 5psi. The pressure of the heat exchanger and the LH2 vessel was slightly
E.X-top and heat E.X-bottom represent SD3 and SD4. In addition, LH2 affected by the gas flow rate of hydrogen gas. The value of gas accu­
vessel-1, LH2 vessel-2, and LH2 vessel-3 represent SD12, SD11, and mulation of hydrogen injected into the heat pipe have been reached to
SD10, respectively. When cooling the inner liquefaction vessel for about 100L in the pre-cooling process. As mentioned, in the pre-cooling pro­
350 min, the pressure of the heat exchanger was adjusted to be under a cess, the temperature at two cryo-coolers and the upper region of heat
maximum of 40psi. When hydrogen gas was injected into the vessel for pipe is very quickly decreased, and consequently the temperature of LH2

1568
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

Fig. 13. Comparison of the cooling effect of helium and hydrogen gas.

1569
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

vessel is slowly decreased. Because of the injection of refrigerant to LN2 approximately 100 min using new experimental system. Also, the total
shield, the temperature of the secondary experiment using hydrogen gas time required to produce the same amount of liquid hydrogen, around
is more rapidly decreased compared to that of the primary experiment 20L, was reduced by approximately 200 min. As a result, the cooling
using helium gas. time of the vessel during the pre-cooling process and the total produc­
Fig. 12 illustrates the gas flow rate, pressure, temperature, and liquid tion time of liquid hydrogen were significantly influenced by the choice
level meter in the liquefaction process using hydrogen gas. The duration of gas used for the initial cooling and the presence of LN2 Shield. Ulti­
of the liquefaction process at a certain level is determined based on the mately, these experiments successfully demonstrated the liquefaction of
primary experiment. Hydrogen gas is injected at a constant flow rate hydrogen gas in the vessel at a rate of approximately 3.5L/h, corre­
ranging from 45 to 48 SLM, which affects the pressure in the heat sponding to a height of 14.5 cm (6.95 %) in the liquid level meter.
exchanger and vessel. After cooling the internal temperature of the In order to effectively carry out the experimental investigations for
vessel to 20 K in the pre-cooling process, hydrogen gas cooled to 80 K is the whole system in the presence of two cryo-coolers and LN2 shield in
supplied through a LN2 pre-cooler to produce liquid hydrogen. The LN2 addition to the small-scale liquefaction system, this study introduced the
is injected into the pre-cooler at a percentage ranging from 50 % to 80 % new experimental system based on the cryogenic techniques. Thus, this
to cool the supplied hydrogen gas to 80 K. The hydrogen gas at 80 K is experimental system will be able to be used in obtaining the optimal
further cooled to 20 K using two cryo-coolers and heat pipes. The re­ design conditions for the production of liquid hydrogen in the lique­
frigerators have a maximum heat capacity of 45L of hydrogen gas per faction system, depending on various design parameters and operation
minute, which corresponds to a maximum liquefaction amount of 3.5L conditions. Using this experimental system, future demonstrations will
per hour. The liquid level meter shows a linear increase over time with include the recovery of boil-off hydrogen gas by incorporating auto­
the presence of a slight fluctuation and uniformity of the temperature matic on-off control of heat exchangers into the system, aiming to ach­
distribution. ieve complete zero-boil off in the storage system.

3.3. Comparison of cooling process between STEP 1 and STEP 2 Declaration of competing interest

Fig. 13 presents the liquid hydrogen production data comparing the The authors declare that they have no known competing financial
cooling effects of helium gas and hydrogen gas. Two differences exist in interests or personal relationships that could have appeared to influence
the experimental conditions: gas used to cool the initial container, and the work reported in this paper.
presence of LN2 shield to shorten the cooling time. The cooling process
of the vessel during the primary liquid hydrogen production took Acknowledgements
approximately 450 min, including the vacuum exhaust after the injec­
tion of helium gas. On the other hand, the cooling process during the This research was supported by Korea Electrotechnology Research
secondary liquid hydrogen production took about 350 min using Institute (KERI) Primary research program through the National
hydrogen gas under the same conditions without the need for additional Research Council of Science & Technology (NST) funded by the Ministry
steps such as vacuum exhaust. As a result, the second liquid hydrogen of Science and ICT (MSIT) (No. 23A01043).
production experiment was able to reduce the cooling time in the vessel
by approximately 100 min compared to the first experiment. Moreover, References
the total time required to produce the same amount of liquid hydrogen,
around 20L, was reduced by approximately 200 min, as shown in [1] Krasae-in S, Stang JH, Neksa P. Development of large-scale hydrogen liquefaction
processes from 1898 to 2009. Int J Hydrogen Energy 2010;35:4524–33.
Fig. 13. Ultimately, these experiments successfully demonstrated the [2] Ohlig K, Decker L. The latest developments and outlook for hydrogen liquefaction
liquefaction of hydrogen gas in the vessel at a rate of approximately technology. AIP Conf Proc 2014:1311–7.
3.5L/h, corresponding to a height of 14.5 cm (6.95 %) in the liquid level [3] Aasadnia M, Mehrpooya M. Large-scale liquid hydrogen production methods and
approaches: a review. Appl Energy 2018;212:57–83.
meter. Therefore, the cooling time of the vessel during the pre-cooling [4] Faye O, Szpunar J, Eduok U. A critical review on the current technologies for the
process and the total production time of liquid hydrogen were signifi­ generation, storage, and transportation of hydrogen. Int J Hydrogen Energy 2022;
cantly influenced by the choice of gas used for the initial cooling and the 47:13771–802.
[5] Hunt JD, Montaniari PM, Hummes DN, Taghavi M, Zakeri B, Romero OJ, Zhou W,
presence of LN2 Shield. Freitas MAA, Castro NJ, Schneider PS, Wada Y. Solid air hydrogen liquefaction, the
missing link of the hydrogen economy. Int J Hydrogen Energy 2023;48:29198–208.
4. Conclusion [6] Taghavi M, Salarian H, Ghorbani B. Economic evaluation of a hybrid hydrogen
liquefaction system utilizing liquid air cold recovery and renewable energies.
Renew Energy Res Appl 2023;4:125–43.
A condensing-type hydrogen liquefaction system has been developed [7] Krasae-in S. Optimal operation of a lqrge-scale liquid hydrogen plant utilizing
with the aim of enabling long-term storage of liquid hydrogen. In order mixed fluid refrigeration system. Int J Hydrogen Energy 2014;39:7015–29.
[8] Seyam S, Dincer I, Agelin-Chaab M. Analysis of a clean hydrogen liquefaction plant
to effectively carry out the experimental investigations in the presence
integrated with a geothermal system. J Clean Prod 2020;243:118562.
of two cryo-coolers and LN2 shield in addition to the small-scale lique­ [9] Yilmaz C. Optimum energy evaluation and life cycle cost assessment of a hydrogen
faction system, this study introduced the new experimental system liquefaction system assisted by geothermal energy. Int J Hydrogen Energy 2020;45:
based on the cryogenic techniques. By using the two cryo-coolers, the 3558–68.
[10] Lee H, Haider J, Qyyum MA, Choe C, Lim H. An innovative high energy efficiency-
temperature and pressure in the whole system was effectively based process enhancement of hydrogen liquefaction: energy exergy, and economic
controlled. In addition, by using the LN2 shield, the heat transfer perspectives. Fuel 2022;320:123964.
generated by the conduction from the upper flange and the radiation [11] Kim H, Haider J, Qyyum MA, Lim H. Mixed refrigerant-based simplified hydrogen
liquefaction process: energy, exergy, economic, and environmental analysis.
between the inner and outer walls was reduced, and hence the efficiency J Clean Prod 2022;367:132947.
of hydrogen liquefaction was enhanced. [12] Shirai Y, Shiotsu M, Matsumoto T, Kobayashi H, Naruo Y, Nonaka S, Inatani Y.
In the primary liquid hydrogen production experiment, helium gas Heat transfer characteristics of liquid hydrogen flowing inside of a vertical heated
pipe under quasi-stationary heat input. Cryogenics 2021;113:103230.
was chosen for heat exchange during the cooling process inside the [13] Zuo Z, Jiang W, Qin X, Huang Y. Numerical investigation on full thermodynamic
vessel due to safety considerations. However, in the secondary liquid venting process of liquid hydrogen in an on-orbit storage tank. Int J Hydrogen
hydrogen production experiment, hydrogen gas was used to cool the Energy 2020;45:27792–805.
[14] Krasae-in S, Bredesen AM, Stang JH, Neksa P. Simulation and experiments of a
internal temperature of the vessel without the need for additional steps hydrogen liquefaction test rig using a multi-component refrigerant refrigeration
such as vacuum exhaust. In addition, LN2 was injected into the LN2 system. Int J Hydrogen Energy 2011;36:907–19.
shield between the outer and inner walls of the vessel to reduce the
cooling time. The cooling time of the vessel was reduced by

1570
D.W. Ha et al. International Journal of Hydrogen Energy 49 (2024) 1558–1571

[15] Qiu C, Chen S, Shen Y, Tao X, Gan Z. Numerical and experimental study on a tube- [24] Zheng J, Chen L, Xu X, Guo L, Zhou Y, Wang J. A novel insulation system based on
in-tube heat exchanger working at liquid-hydrogen temperature with a large active cooling without power input for liquid hydrogen storage. Energy 2019;182:
temperature span. Int J Heat Mass Tran 2023;208:124089. 1–10.
[16] Baik JH, WU Notardonato. Initial test results of laboratory scale hydrogen [25] Zheng J, Chen L, Wang J, Xi X, Zhu H, Zhou Y, Wang J. Thermodynamic analysis
liquefaction and densification system. AIP Conf Proc 2006;823:1530. and comparison of four insulation schemes for liquid hydrogen storage tank.
[17] Baik JH, Karng SW, Garceau N, Jang YH, Lim CM, Kim SY, Oh IH. Development of Energy Convers Manag 2019;186:526–34.
1 L hr-1 scale hydrogen liquefier using Gifford-McMahon (GM) cryocooler. AIP [26] Shi C, Zhu S, Wan C, Bao S, Zhi X, Qiu L, Wang K. Performance analysis of vapor-
Conf Proc 2014;1573:1357. cooled shield insulation integrated with para-ortho hydrogen conversion for liquid
[18] Garceau NM, Baik JH, Lim CM, Kim SY, Oh IW, Karng SW. Development of a small- hydrogen tanks. Int J Hydrogen Energy 2023;48:3078–90.
scale hydrogen liquefaction system. Int J Hydrogen Energy 2015;40:11872–8. [27] Xu P, Lei G, Xu Y, Wen J, Wang S, Li Y. Study on continuous cooling process
[19] Xie F, Xia S, Zhu Y, Ma Y, Li Y. Experimental study on small-scale hydrogen coupled with ortho-para hydrogen conversion in plate-fin heat exchanger filled
liquefaction of 0.5 L/h. Int J Hydrogen Energy 2022;47:38258–70. with catalyst. Int J Hydrogen Energy 2022;47:4690–703.
[20] Shen Y, Qiu C, Liu D, Tao X, Wan A, Zhang Z, Gan Z. Experimental study on a [28] Qyyum MA, Riaz A, Naquash A, Haider J, Qadeer K, Nawaz A, Lee H, Lee M. 100%
closed-cycle Joule-Thomson cryocooler working at liquid hydrogen temperature. saturated liquid hydrogen production: mixed-refrigerant cascaded process with
Appl Therm Eng 2023;234:121291. two-stage ortho-to-para hydrogen conversion. Energy Convers Manag 2021;246:
[21] Shen Y, Liu D, Liu L, Qiu C, Zhuan R, Gan Z. Experimental study on a simplified 114659.
precooled JT cryocooler for liquid hydrogen zero boil-off storage. Appl Therm Eng [29] Meng C, Qin X, Jiang W, Pu L, Liu W, Huang Y. Cooling effect analysis on para-
2022;216:119087. ortho hydrogen conversion coupled in vapor-cooled shield. Int J Hydrogen Energy
[22] Chen S, Qiu C, Shen Y, Tao X, Zhang Z, Wan A, Gan Z. Experimental and dynamic 2023;48:15600–11.
simulation study of the Joule-Thomson cryocooler working at liquid hydrogen [30] Teng J, Wang K, Zhu S, Bao S, Zhi X, Zhang X, Qiu L. Comparative study on
temperature. Appl Therm Eng 2023;235:121285. thermodynamic performance of hydrogen liquefaction processes with various
[23] Zheng J, Chen L, Liu X, Zhu H, Zhou Y, Wang J. Thermodynamic optimization of ortho-para hydrogen conversion methods. Energy 2023;271:127016.
composite insulation system with cold shield for liquid hydrogen zero-boil-off
storage. Renew Energy 2020;147:824–32.

1571