i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6

Available online at www.sciencedirect.com

ScienceDirect

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

Performance analysis of open-loop cycles for LH2

regasification

Ahmed Laouir*
Department of Mechanical Engineering, University M.S.Benyahia, BP 98, 18000 Jijel, Algeria

article info abstract

Article history: The paper addresses the subject of liquid hydrogen (LH2) regasification by means of a
Received 10 June 2018 technology that takes advantage of hydrogen coldness to produce mechanical energy.
Received in revised form Liquid hydrogen undergoing regasification plays the role of the heat sink, combined with
8 December 2018 an external heat source, it becomes possible to run a thermal machine. A theoretical
Accepted 21 December 2018 analysis is first given to discuss the ideal regasification. Then some original processes are
Available online 1 February 2019 proposed based on cycles using hydrogen at the same time as the working fluid. Two types
of cycles are considered depending on whether the regenerative loop is super or tran-
Keywords: scritical. Simulation results demonstrate that recompression is an effective means for
Liquid hydrogen improving regasification efficiency. Compared to direct expansion process, the output is
Cold energy about 2e3 times higher. The second law efficiency is in general greater than 30% and can
Power cycles exceed 45%.
© 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

Hydrogen is expected to be a major energy vector in the

Introduction future. In this perspective, the development of efficient pro-
duction methods is subject to intensive research activity [9].
Regasification is the process intended to turn a cryogenic fluid This is also the case for liquefaction, which will play an
back to the gaseous form. Nowadays, regasification technol- important role in the hydrogen economy. In general, end-
ogy is mostly encountered in the field of the liquefied natural users require hydrogen in the gaseous form [10] making
gas (LNG) industry. Various designs have been addressed in regasification as important as the other processes composing
the literature [1e3] and LNG regasification terminals produc- the hydrogen industrial chain. Liquefaction leads to a reduc-
ing power have been built and operated since decades [4]. LNG tion of the hydrogen volume by a factor of 870, which makes
can play the role of a heat sink for a single or a combination of storage and transport easier and economical especially for
power cycles. In more complex configurations, the regasifi- large quantities. Likewise, producing and storing hydrogen is
cation system is integrated with other industrial processes a possible solution to address large-scale storage of the energy
requiring low-temperature cold like air separation, hydrogen produced from renewable resources [11,12]. The storage is
liquefaction and CO2 capturing systems [5]. Regarding the possible a priori in the form of compressed gas in geological
hydrogen, there has been a constant research effort to formations or in tanks, or as a cryogenic liquid [13e15]; the
improve liquefaction methods as reviewed in Refs. [6e8]. On comparative advantages and disadvantages are still under
the contrary, publications on liquid hydrogen (LH2) regasifi- debate. The principal constraint cited regarding the liquid
cation using thermodynamic cycles are far less numerous. storage option is the significant power consumption during

* Corresponding author.
E-mail addresses: [email protected], [email protected].
https://doi.org/10.1016/j.ijhydene.2018.12.204
0360-3199/© 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
22426 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6

the liquefaction stage. Actual hydrogen liquefiers consume up understanding of the factors that can weaken the perfor-
to 47 MJ/kg, the objective with improved designs is to reduce mance of regasification in general and to gain insight into the
power consumption to less then 25 MJ/kg [7]. To this end, the possible solutions that can be considered to design effective
most promising technology is that based on mixed fluid methods. In a second phase, the approach is adapted to the
refrigeration cycles. In a recent work, it is shown that power hydrogen case to define practical processes for this fluid. The
consumption can be reduced to a value as low as 16 MJ/kg [16]. patented method [27] is general and applicable to other cryo-
Despite this drawback, the storage of hydrogen in the form of genic fluids. In a previous work, it was evaluated for LNG and
a cryogenic liquid instead of a compressed gas is thought to be the expected advantages were discussed [28]. Beyond the
advantageous at very large-scale of energy storage [17]. similarities due to comparable thermodynamic behaviour, the
As in the LNG case, to take advantage of the coldness practical feasibility of the method in relation to a given fluid
available in LH2 and produce valuable mechanical energy, should examine the specific features of that fluid. Compared
there is a need for effective regasification methods. Such a to LNG, hydrogen presents lower temperature levels, higher
system will contribute to make the whole chain, from pro- heat capacities, lower gas densities and more critical safety
duction to consumption, more profitable as a portion of the considerations. The central problem to be addressed in order
energy consumed during the liquefaction stage can be recov- to improve the efficiency concerns the internal heat transfer,
ered. For energy storage applications, a regasification system that is, the regeneration. An effective regeneration is achiev-
producing power is crucial to improve the roundtrip efficiency able if some rules are observed. Among these, phase transi-
and the attractiveness of LH2 energy storage option. LH2 tions as well as working over isobars with sharp heat capacity
coldness offers an opportunity to run a thermal machine and variation must be avoided. In addition, flow rates of the
to produce power. By absorbing some heat from the environ- heated and the heating steams must be properly chosen. The
ment or from other available waste heat sources, a fraction of derived processes comport the specific feature to put into play
it is converted as useful work and the other is discarded from recompression operations arranged in the regenerative loop
the process in accordance with the second law of thermody- to significantly improve the efficiency.
namics. The pressure at which the gaseous hydrogen may be
produced depends on the context; long-distance pipeline
transport involves gas pressures ranging between 70 and Some features regarding the ideal process
100 bar [18]. Distribution for end-users requires medium to
low-pressure levels depending on the type of use. On board A regasification performed without any entropy production at
applications for LH2 fuelled vehicles is another situation the surroundings temperature T0 produces the maximum
where the technology can be used. possible quantity of work Wid given by Ref. [29],
From the fundamental viewpoint, the thermodynamics of


LNG and LH2 regasification are quite similar. To some extent, Wid ¼ hg  hl  T0 sg  sl (1)
possible regasification methods for LH2 could be inspired from where h and s are the enthalpies and entropies, subscripts g
those known for LNG. The simplest regasification method that and l refer to the gas and the liquid, respectively. The quantity
can be applied to the hydrogen is the direct expansion (DEx) T0 ðsg  sl Þ corresponds to the heat Q0 absorbed by the fluid.
[19]. It consists of three successive stages: The pumping of the The graphical representation of Wid in the temperature (T)-
liquid (elevating pressure), followed by a heating (elevating entropy (s) diagram is shown in Fig. 1a. Fig. 2 shows the plot of
temperature) and then the expansion of the gas to the desired
pressure. Other methods have been evaluated based on closed
Brayton, Stirling or Rankine cycles cooled by the LH2 stream
undergoing regasification [20e25]. Because LH2 temperature is
very low (20 K at 1 bar), helium is practically the only fluid
convenient as the working medium in closed cycles. Other
fluids like nitrogen, argon and methane have been tested in
closed Rankine cycles [21]. But in this case, in order to prevent
the freezing of the working fluid, the condensing temperature
is maintained relatively high compared to that of LH2. By
doing so, the potential of the heat sink is not completely
exploited. Closed cycles are also combined to direct expansion
to achieve higher efficiency [22]. In a study [26], LH2 serves to
perform a deep cooling of the gas entering the compressor of a
gas turbine. The gas mainly nitrogen with a low content of
oxygen is cooled down to 83 K. In other concepts, LH2 is used
to cool a magnetohydrodynamic [17] and a thermoelectric [23]
generator.
In this work, we first present a general analysis of the
regasification process, focusing on open-loop methods using Fig. 1 e a/Ideal work of regasification Wid represented as a
the fluid to be regasified as the working fluid. The theoretical surface. b/Principle of an open-loop regasification cycle
study to be presented examines in detail the reversible i.e. with an isothermal expansion between pressures PH and
ideal regasification. This preliminary task allows the PL.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6 22427

this process is not feasible because of the extremely high

pressure needed for increasing the fluid temperature from
saturation to T0 during the first stage.
In place of the previous method, the ideal process can be
imagined as a loop, fed with the liquid and delivering the gas.
This constitutes an open-loop cycle in which a part fid of the
fluid is recirculated to close the loop as shown in Fig. 1b. The
only information given to define the cycle are: it comprises an
isothermal expansion absorbing heat between a high pressure
PH and a lower pressure PL, the cycle is adiabatic elsewhere,
the stream fid is recirculated at pressure PL. As an ideal pro-
cess, the cycle must absorb a quantity of heat equal to Q0 and
produce a net work output equal to Wid. The isothermal
expansion between PH and PL produces a work WE while
another quantity of work WC can be involved elsewhere in the
loop. The two equality conditions are then,

 

T0 sg þ fid sðPL Þ  1 þ fid sðPH Þ ¼ T0 sg  sl (2)

WE þ WC ¼ Wid (3)

From (2) the expression of the ideal flow recirulated fid is

Fig. 2 e Ideal work of regasification Wid of parahydrogen, sðPH Þ  sl

fid ¼ (4)
plotted versus gas pressure Pg at different surroundings sðPL Þ  sðPH Þ
temperature T0.
The energy balance written for the isothermal expansion is
WE þ Q0  fid hðPL Þ  hg þ ðfid þ 1ÞhðPH Þ ¼ 0 and having for all the
system Wid þ Q0  hg þ hl ¼ 0, WC expression is then,
(1) as a function of the pressure Pg of the produced gas and T0.
Liquid hydrogen is considered saturated at 1.1 bar and rega- WC ¼ hl  hðPH Þ  fid ðhðPH Þ  hðPL ÞÞ (5)
sified as 100% parahydrogen. The thermodynamic properties
Eqs. (4) and (5) are independent of the properties of the
are calculated using the modified Benedict-Webb-Rubin
produced gas; fid and WC depend principally on the choice of PH
equation of state for parahydrogen [30].
and PL. The plot of (4) is shown in Fig. 3; the curves indicate
Actually, liquid hydrogen at equilibrium is 99.8% para and
0.02% ortho. At room temperature, the composition is 75%
para and 25% ortho, known as normal hydrogen. The para to
ortho conversion is an endothermic process; hence, the total
ideal work of regasification, including an ideal para-ortho
conversion, is greater than the values indicated in Fig. 2. The
difference can be estimated as follows. The total (i.e. including
ortho-para conversion) ideal work for liquefying normal
hydrogen at 21 bars and 298 K is 10.404 MJ/kg [31]. This work is
also equal to that involved in the reverse process, namely the
ideal regasification from LH2 state to normal hydrogen at
21 bars and 298 K. The value taken from Fig. 2 for para-
hydrogen regasification is 8.645 MJ/kg. The difference between
the two, 1.759 MJ/kg, is the ideal work related to a reversible
para-ortho conversion. In summary, an additional work of
1.759 MJ/kg can be produced compared to that given by Fig. 2 if
the contribution of the reversible para-ortho conversion is
taken into account. The only approximation made for this
estimation is that LH2 is supposed to be 100% parahydrogen.
Because the conversion is a very slow process and takes days
without the use of a catalyst [32], ignoring it is realistic as long
as the regasification system is considered without conversion
capability.
As represented in Fig. 1a, the ideal regasification can be
performed using a once-through process composed of an Fig. 3 e Ideal flow fid recirculated within a hypothetical
isentropic pressure elevation until the temperature T0 then ideal open-loop regasification cycle, in kg per kg-of-gas.
followed by an isothermal expansion at T0 at which the fluid Plotted versus high pressure PH at different values of low
absorbs a quantity of heat Q0 from the surroundings. Actually, pressure PL (T0 ¼ 300 K).
22428 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6

that increasing the difference between PH and PL leads to a stream at PH can be heated beyond T0 by using a higher-grade
significant reduction in fid. Assuming a similar trend with heat source [19]. But in downing so, the efficiency from the
actual systems, this effect could favour a reduction in thermodynamic point of view is not necessarily improved. In
component size and capital cost as the quantity of fluid this work, some novel methods are proposed giving the
needed to be recirculated is lowered. The plot of (5) is shown in possibility of exploiting at the maximum the energy potential
Fig. 4; the negative values indicate that WC is absorbed in the available between the surroundings at T0 and the cryogenic
form of pumping and/or compression work. This demon- fluid. Such efficient systems will also insure superior perfor-
strates that the loop must include some work-absorbing mance if any other heat source at higher temperature can be
transformations in order to accomplish the regasification. used. The central problem to be resolved concerns the heat-
The plot indicates that increasing PH and/or PL leads to an ing of the H2 stream so as to reduce the heat transfer entropy
increase in WC. production.
The LH2 at PH can be heated by internal heat transfer,
namely by regeneration, using a part of the expanded fluid at
Regenerative cycles comprising parallel T0. This option does not insure in all cases a significant
recompressions improvement of the efficiency, unless certain conditions are
met. Fig. 5b represents and explains the method to be followed
As stated before, the once-through process shown in Fig. 1a is in order to perform and favour an effective regeneration. LH2
unpractical. A simple way to produce some work is to operate at PH and T3 is heated up to T0 by a stream f1 at PL. Assuming an
according to the method of direct expansion (DEx) [19]: ideal heat transfer, f1 is cooled down to T3. The same situation
increasing LH2 pressure to an acceptable value PH then as before arises: how to heat f1 again to T0 without directly
heating the stream to the temperature T0 and then expanding using the external heat source. Compression can be used as
the gas to the desired pressure as represented in Fig. 5a. To another means permitting to elevate the fluid temperature
simplify presentation, isobars are drawn as straight lines in without the need of a heating medium. The stream f1 at PL and
the T-s diagram. Compared to the ideal process shown in T3 once compressed to PH, warms up to T2. To complete the
Fig. 1a, direct expansion causes important thermodynamic heating up to T0 another stream f2 at PL is used, then recom-
losses due to the direct and irreversible heat transfer from the pressed to attain T1. The process is repeated until the last
external heat source to the hydrogen stream at a much lower recompression alone suffices to elevate the fluid temperature
temperature. The lost work corresponds to the dotted trian- to T0. This gives the spiral construction, which may be called
gular surface in Fig. 5a. Direct expansion presents the the ideal regenerative path. As shown, the superposed loops
advantage of simplicity but low efficiency is its principal so arranged are equivalent to an association of a number of
drawback. To increase the quantity of work that can be pro- cycles producing together a total work represented by the
duced per unit mass of regasified hydrogen, the hydrogen triangular surface. The total surface corresponds to the ideal
work of regasification shown in Fig. 1a. Because the heat
transfer is supposed ideal and causes the same temperature
variation at both sides for a pair of streamsf ðiÞ and f ði  1Þ
having respective heat capacities cpH and cpL, this leads to the
relationf ðiÞ=f ði  1Þ ¼ cpH =cpL ¼ r; i ¼ 1; 2; 3. Having f 0 ¼ 1,
the total flow rate used for the regeneration is then ftotal ¼ f 1 þ
f 2 þ f 3 ¼ r þ r2 þ r3 . According to the analysis presented in this
section and Section Some features regarding the ideal process,
ftotal and fid must be equal. As a consequence, for a given value
of fid there is a corresponding value of the ideal heat capacity
ratiorid . This represents the condition that the fluid properties
must satisfy to make an ideal regasification possible according
to the method represented in Fig. 5b.

Adapting the method to the hydrogen

The analysis given in Section Regenerative cycles comprising

parallel recompressions demonstrates the central role that
regeneration plays in improving efficiency. On the other hand,
the recompression as a work-consuming operation is in fact
beneficial for elevating the output as the fluid, after being
heated at higher temperature, is expanded at more important
flow rates. The question is how to adapt in the best possible
Fig. 4 e Compression work WC absorbed by a hypothetical way the approach to the actual fluid for which the isobars are
ideal open-loop regasification cycle; plotted versus high of irregular shapes. The two isobars PH and PL to be chosen
pressure PH at different values of low pressure PL must make it possible to perform an effective countercurrent
(T0 ¼ 300 K). heat exchange over the entire range of temperatures. From
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6 22429

Fig. 5 e a/Direct expansion (Dex) causes thermodynamic losses due to the irreversible heat transfer between the heat source
and the fluid at cryogenic temperature, b/Correcting direct expansion drawback by judicious internal heat exchange
(regeneration) involving recompression operations.

this perspective, a certain “compatibility” of the two isobars In the light of the foregoing, there are two possible options
should be fulfilled. To this end, the considered isobars should to transpose acceptably the method into the real fluid case.
present a heat capacity ratio as stable as possible in relation to
temperature i.e. cpH =cpL ycste; this also means that the corre-  In the first option, the admitted liquid is pumped to a su-
sponding curves in T-s diagram show a graphical similarity percritical pressure PH then heated countercurrently by
and comparable curvatures. In the actual fluid case, this another supercritical stream at lower pressure PL. As
condition can partially be fulfilled. The least favorable situa- explained before, the magnitude of the heat capacity
tions to be avoided concern the isobars leading to phase variation should be mitigated by preferring pressure values
transitions or passing close to the critical point. Under such well above the supercritical pressure P>>12.8 bar. In this
conditions, a temperature pinch can easily occur and case, the regenerative loop is entirely situated in the su-
compromise the regeneration. Fig. 6a shows qualitatively the percritical (SC) region.
type of isobars not appropriate to perform an effective  In the second option, the admitted liquid is pumped to a
regeneration. Fig. 6b gives the normalized heat capacity of supercritical pressure PH not necessarily as high as in the
supercritical parahydrogen. A sharp and detrimental increase previous case. The heating stream is a gas at relatively low
of the heat capacity occurs near the critical point. pressure PL to eliminate the risk of condensation. The

Fig. 6 e a/Typical isobars shape in actual Temperature (T)-Entropy (s) diagram, a decrease in isobar slope denotes a local
increase in heat capacity value, at saturation heat capacity is infinite, b/Ratio of parahydrogen heat capacity to that of the
gas at 1 bar, cp =cpg ð1barÞ.
22430 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6

regenerative loop is then transcritical (TC), namely Cr1, Cr2 and Cr3 are arranged within the regenerative loop
distributed between the super and the subcritical zone. according to the approach presented in Section Regenerative
Thanks to the greater compressibility of the gas, the cycles comprising parallel recompressions.
recompression participates more strongly in increasing the As mentioned before, the thermodynamic properties are
fluid temperature. Therefore, an open-loop cycle com- calculated using a BWR equation of state for parahydrogen.
porting TC regeneration needs fewer recompression oper- The minimum temperature difference is set to 3  C for heat
ations than a cycle with SC regeneration. exchangers E. For regenerator Er1 the minimum temperature
difference can vary between 3 and 4  C, and for or re-
generators Er2, Er3, and Er4 between 3 and 7  C depending on
the flow rates f1, f2, and f3 giving the maximum net work
Simulated examples output. Due to the curvature of the isobars, the minimum
temperature difference can occur inside the regenerators
To illustrate the method, two typical cases are considered and rather then at the sides. The written Fortran program contains
analysed by simulation. A SC cycle working at 250/70 bar and a routine intended to prevent the occurrence of a “hidden”
delivering the gas at 70 bar intended for pipeline transport. temperature pinch lower than the fixed values.
The second is a TC cycle working at 60/4 bar delivering the gas The analysis is carried out in such a way that the isentropic
at 4 bar for consumers. Fig. 7 represents the system layout. In efficiency IsE is taken as the principal parameter. This is
place of the theoretical isothermal expansion, four adiabatic because the system is highly sensitive to the quality of the
expansions and three reheats are arranged to extract work. pump, compressors, and turbines. To this end, IsE is varied
The expansion is divided over the turbines with an expansion between 0.75 and 0.90 and a same value is given to all these
ratio per turbine of about (PL/PH)1/4. The working pressures for components. The pressure drop at different parts is neglected.
SC and TC cycles are 250/182/132/96/70 bar and 60/30/15/7.5/ The efficiency of the system is evaluated by calculating the net
4 bar, respectively. The heat source is supposed at 20  C. The work output per one kg of hydrogen produced,
regenerative loop contains three compressions for the SC
cycle and two for the TC cycle. X
4 X
3

Liquid hydrogen is fed at point 1, the pressure is then W¼ WTbi  WCrj  WP0 (6)
i¼1 j¼1
elevated to the value PH with the pump P0. Between points 2
and 6, the fluid is successively heated in regenerators Er. Be- In relation to the layout of Fig. 7, direct expansion DEx re-
tween points 6 and 14, the fluid absorbs heat from the external fers to the system without compressors Cr1, Cr2 and Cr3, or to
heat source at heat exchangers E and endergoes expansions in the situation where f 1 ¼ f 2 ¼ f 3 ¼ 0; the corresponding net
P
turbines Tb in an alternating manner. The role of heat output is. WDEx ¼ 4i¼1 WTbi  WP0
exchanger E5 is to warm up the gas if necessary prior to de- The quality of the regasification from the thermodynamic
livery. At point 15, a total flow rate f1þf2þf3 at PL is directed to point of view is evaluated by calculating the second law effi-
regenerators Er to heat the admitted hydrogen. Compressors ciency SLE, defined as the ratio of the actual work to the ideal

Fig. 7 e Calculated regasification system. E: heat exchanger, Er: regenerator, Tb: turbine, Cr: compressor, P0: pump.
Transcritical (TC) process comprises only compressors Cr1 and Cr2.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6 22431

work of regasification, SLE ¼ W=Wid . The imperfection of the needed in theory with the SC cycle and three with the TC
regeneration is evaluated by calculating the relative thermo- cycle. In real case, the temperature differences at regenerators
dynamic losses ThL expressed as the ratio of the lost work [33] and friction in compressors are factors reducing the number
(or the destroyed exergy) to the ideal work of regasification, of the possible compressions that can be used. As represented
ThL ¼ sT0 =Wid . In the previous expression, s is the entropy by the dotted segment in Fig. 8, a fourth compression can be
production at each regenerator Er, and T0 is the surroundings added to the SC cycle.
temperature, that is, the heat source at 293 K. The TC cycle can work with two compressions at a
For IsE ¼ 0.80, taken as a reference value, Table 1 gives the maximum. In this case, the system layout of Fig. 7 does not
thermodynamic states at different points of Fig. 7. Table 2 comprise compressor Cr3 and regenerator Er4. Points 15 and
gives the amounts of mechanical work exchanged at 16 are at the same thermodynamic state; this is also the case
different components. The cycles are represented in the T-s for points 5 and 6. Under certain conditions, temperature T(4),
diagram as shown in Fig. 8. The stepped construction in the which is equal to T(21), may be greater than T(15) making the
red dashed line indicates that six isentropic compressions are use of regenerator Er3 unnecessary. In this case, states 17 and
16 become identical as well as states 4 and 6. With the chosen
operating conditions, the isentropic efficiency IsE is the factor
Table 1 e Thermodynamic states at the different points determining whether Er3 is necessary or not. For the case
indicated in Fig. 6. Results for IsE ¼ 0.80. First line for the IsE ¼ 0.80, T(4) ¼ 258.34 K and T(15) ¼ 253.52 K meaning that Er3
SC process, second line for the TC process. The cycles are should not to be used.
represented in Fig. 8. The practical realisation of a regasification system like one
Point Temperature Pressure P Enthalpy h Entropy s of those to be analysed can fully use the available hydrogen
T (K) (bar) (kJ/kg) (kJ/kg.K) technology. In this sense, specialised companies have exten-
1 20.40 1.10 52.46 2.974 sive experience gained from the development of hydrogen
20.00 1.20 48.71 2.782 liquefiers. This includes components design and
2 37.09 250.00 453.32 5.418 manufacturing (compressors, expanders, heat exchangers,
24.58 60.00 149.05 3.636 valves, safety devices…) as well as the operation and main-
3 58.79 250.00 704.45 10.704 tenance of the plants. The new aspect that may probably need
91.19 60.00 1191.46 23.709
a peculiar design effort is related to the compressors involved,
4 101.08 250.00 1315.33 18.453
258.34 60.00 3906.08 40.576
especially in the transcritical configuration. Because the
5 162.19 250.00 2374.90 26.617 compression ratio is relatively high, this will lead to use a
258.34 60.00 3906.08 40.576 great number of compression stages making the compressor
6 234.45 250.00 3613.92 32.953 more expensive. Nevertheless, because the gas to be com-
258.34 60.00 3906.08 40.576 pressed is at lower temperatures and/or higher pressures, the
7 290.15 250.00 4496.66 36.334
gas densities are higher than encountered in liquefiers at the
290.15 60.00 4391.39 42.348
compression stage. This factor can significantly facilitate the
8 271.29 182.00 4164.94 36.643
250.25 30.00 3769.78 42.981 design of compressors meeting the required specifications.
9 290.15 182.00 4456.52 37.682
290.15 30.00 4376.70 45.232 Supercritical process 250/70 bar
10 270.88 132.00 4132.60 37.984
250.19 15.00 3763.70 45.857 The obtained results are shown in Figs. 9 and 10. The regen-
11 290.15 132.00 4428.74 39.040
eration is affected in different ways by the isentropic effi-
290.15 15.00 4369.52 48.105
ciency IsE. Values of flows f1, f2, and f3 giving the maximum
12 270.91 96.00 4115.68 39.332
250.16 7.50 3760.79 48.725 work production vary sensitively with IsE as shown in Fig. 9a.
13 290.15 96.00 4409.70 40.380 Temperatures T(19), T(18) and T(17) at which f1, f2, and f3 are
91.19 60.00 1191.46 23.709 admitted for compression depend strongly on IsE value. The
14 270.99 70.00 4104.82 40.664 higher the IsE, the lower is the temperature at each
253.52 4.00 3811.12 51.527 compressor inlet, and the lower is the specific work absorbed
15 270.99 70.00 4104.82 40.664
during compression; the reverse is also true. Fig. 9a shows that
253.52 4.00 3811.12 51.527
optimal values of f1 and f2 slightly increase as IsE is increased;
16 165.84 70.00 2403.49 32.681
253.52 4.00 3811.12 51.527 on the contrary, f2 diminishes significantly. The improvement
17 107.02 70.00 1422.98 25.403 in the output by associating three successive recompressions
253.52 4.00 3811.12 51.527 to direct expansion DEx is detailed in Fig. 9b. The curves show
18 63.01 70.00 750.19 17.293 that such an association can considerably enhance the effi-
112.67 4.00 1573.21 38.656 ciency. The gain obtained by adding compressor Cr3 is modest
19 40.09 70.00 351.57 9.442
at low IsE values. In general the quantity of work consumed by
28.58 4.00 551.24 22.234
20 58.79 250.00 704.45 10.704
compressor Cr3 is quite important compared to those at
91.19 60.00 1191.46 23.709 compressors Cr1 and Cr2. This is because at T(17) the specific
21 101.08 250.00 1315.33 18.453 volume of the fluid is relatively large. The situation is aggra-
258.34 60.00 3906.08 40.576 vated when IsE is low causing additional temperature eleva-
22 162.19 250.00 2374.90 26.617 tion by friction. Near IsE ¼ 0.75 the backwork i.e. the work
/ / / /
needed to drive compressor Cr3, is practically as high as the
22432 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6

Table 2 e Work absorbed/produced at indicated components in MJ/kg for IsE ¼ 0.80. Flow rates (f1, f2, f3) are (0.63, 0.85, 1.20)
and (1.02, 1.27, 0.00) for SC and TC cycle respectively.
Component P0 Cr1 Cr2 Cr3 Tb1 Tb2 Tb3 Tb4
SC process 0.401 0.222 0.480 1.142 1.221 1.192 1.152 1.122
TC process 0.100 0.653 2.963 / 2.045 2.017 2.003 1.825

Fig. 8 e Supercritical (left) and transcritical (right) cycles represented in T-s diagram of parahydrogen, isentropic efficiency
IsE ¼ 0.80.

expansion work of the flow f3 in the turbines making the third (and a fifth regenerator), the temperature difference at Er4 will
loop useless. The second law efficiency SLE, calculated on the remain high causing the destruction of a substantial quantity
basis of an ideal work of regasification Wid ¼ 7.009 MJ/kg for of work.
the gas produced at 70 bar and 20  C, is shown in Fig. 9c. At
IsE ¼ 0.80, the cycle produces 2,44 MJ/kg with an SLE value Transcritical process 60/4 bar
equal to 35%.
The thermodynamic losses ThL at each regenerator are The obtained results are shown in Figs. 11 and 12. Optimal f1
shown in Fig. 9d. The curves indicate that the regeneration is and f2 values are given by Fig. 11a. There is a moderate
quite effective at Er1, Er2, and Er3 at which the losses remain decrease in f1 and slight increase in f2 with increasing IsE. As
low. On the contrary, Er4 can be responsible of significant can be observed from Fig. 11b, for IsE lower than 0.785, the use
losses that become more important as IsE is increased. Tem- of the second compression causes a reduction in the total
perature profiles at two different IsE values are represented in output. Around the reference value IsE ¼ 0.80, the use of the
Fig. 10. For IsE ¼ 0.75 the regeneration is globally satisfactory second compression is also not justified. The system working
as indicated by the narrow gap i.e. temperature difference only with a single compression is quite effective and presents
between the two streams. This allows concluding that the practically similar performance to the case with two com-
relatively poor performance at IsE ¼ 0.75 is due to factors other pressions, which produces 4.17 MJ/kg with an SLE value equal
than the quality of heat transfer at the regenerators. The to 39% (Wid ¼ 10.572 MJ/kg for the gas produced at 4 bar and
reason is mainly linked to friction in the pump, compressors 20  C). The second compression might be envisaged with
and turbines. For IsE ¼ 0.90 the plot show that at regenerator higher quality turbines and compressors, but the improve-
Er4, the temperature difference between the two streams is ment would be quite moderate as shown in Fig. 11b.
important as well as the quantity of heat transferred. This Compared to the SC case, the TC cycle shows better values of
causes a loss of about 14% indicated in Fig. 9d. This represents second law efficiency: 0:32 < IsE < 0:53versus 0:26 < IsE < 0:46;
practically 1 MJ/kg of destroyed work. As stated before, high but the SC cycle with four compressions is likely to achieve
IsE values lead to a reduction in temperatures at compressors higher performance. The thermodynamic losses at re-
and make possible the use of additional compressions to generators Er1 and Er2 are shown in Fig. 11d. As explained
extract more work. In the absence of a fourth compression before, the use of regenerator Er3 is required only for high IsE
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6 22433

Fig. 9 e Results for supercritical (SC) process 250/70 bar, as functions of isentropic efficiency IsE. a/Optimal flow rates at
compressors in kg per kg-of-gas. b/Work output W improvement by means of recompressions. The curve “DEx þ 3
compressions” gives the total output. c/Second law efficiency SLE. d/Thermodynamic losses ThL at regenerators Er.

Fig. 10 e Temperature profile of SC regeneration at two different values of isentropic efficiency IsE.
22434 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6

Fig. 11 e Results for transcritical (TC) process 60/4 bar, as functions of isentropic efficiency IsE. a/Optimal flow rates at
compressors in kg per kg-of-gas. b/Work output improvement by means of recompressions, the curve “Dex þ 2
compressions” gives the total output. c/Second law efficiency SLE. d/Thermodynamic losses ThL at regenerators Er.

Fig. 12 e Temperature profile of TC regeneration at two different values of isentropic efficiency IsE.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6 22435

values. But in general its role remains marginal; at IsE ¼ 0.90,

Er3 causes a small ThL of about 0.9% (not noticeable in Acknowledgements
Fig. 11d). Temperature profiles, represented in Fig. 12, show
relatively larger temperature differences between the two The author expresses his thanks to Mr. Omar Laghrouche for
streams. This explains the significant losses at Er1 and Er2 providing useful references.
indicated in Fig. 11d. The total loss can attain 20% represent-
ing 2.11 MJ/kg of destroyed work. Nomenclature

cp Heat capacity, kJ/kg.K

Conclusion f1, f2, f3 Relative flow rates at compressors
h Specific enthalpy, kJ/kg
The theoretical study presented about the reversible regasifi- IsE Isentropic efficiency
cation when conducted thought an open cycle led to the P Pressure, bar
identification of two ideal quantities in addition to the well- r Heat capacities ratio
known ideal work of regasification. The ideal flow recircu- s Specific entropy, kJ/kg.K
lated and the ideal work of compression are derived and SLE Second law efficiency
quantified for the hydrogen. The two quantities depend T Temperature, K
mainly on the low and high pressure in the loop, and are in- ThL Thermodynamic losses, Exergy destruction
dependent of the properties of the produced gas. A method to W Mechanical work, MJ/kg
correct the drawback of direct expansion is defined based on s Entropy production, kJ/kg.K
the use of internal heat exchange and the transfer of quanti- Subscripts
ties of fluid from the low-pressure side to the high-pressure C Compression
side by recompression. The first advantage of the regasifica- Cr Compressor
tion method described in the paper is that the quantity of DEx Direct expansion; E Expansion
work that can be extracted from the temperature difference g Gas
between a given heat source and cryogenic hydrogen is H High, At high pressure
maximised. The second advantage is the regasification is i Summation index
based on hydrogen as working fluid and no special fluid id Ideal
should be selected. This is particularly important in the j Summation index
hydrogen case for which there is a very limited choice of l Liquid
suitable fluids that can be used as working medium in closed L Low, At low pressure
regasification cycles. In addition, an open cycle eliminates the 0 Surroundings conditions
risk of failure related to possible leakage and loss of working P0 Pump
medium contained in a closed regasification system. Tb Turbine
The results related to the two case studies demonstrate, as
expected from the theoretical analysis, the improvement in
performance in comparison to the direct expansion. The cy- references
cles can be combined in a cascade with power cycles using
high-grad heat sources. The complexity of the system can be
modulated and optimised to take into account the techno- [1] Invernizzi CM, Iora P. The exploitation of the physical exergy
economic constraints by varying the number of compres- of liquid natural gas by closed power thermodynamic cycles.
An overview. Energy 2016;105:2e15.
sors, regenerators and turbines. Another advantage of the
[2] Liu M, Lior N, Zhang N, Han W. Thermoeconomic analysis of
regasification systems constructed according to this method
a novel zero-CO2-emission high-efficiency power cycle using
is versatility: the gas can be produced at different pressures LNG coldness. Energy Convers Manag 2009;50:2768e81.
levels at the same time. For example, a supercritical regasifi- [3] Yu G, Jia S, Dai B. Review on recent liquefied natural gas cold
cation system, like the one presented, gives the possibility to energy utilization in power generation cycles. Adv Geo-
extract some gas at the high pressure to fill hydrogen tubes Energy Res 2018;2:86e102.
and cylinders in addition to the quantity intended to pipeline [4] Hisazumi Y, Yamasaki Y, Sugiyama S. Proposal for a high
efficiency LNG power generation system utilizing waste heat
transmission. The transcritical configuration is suitable for
from the combined cycle. Appl Energy 1998;60:169e82.
delivering hydrogen at low to medium pressures to meet the [5] Mehrpooya M, Moftakhari Sharifzadeh MM, Katooli MH.
needs of different categories of clients. It is also possible to Thermodynamic analysis of integrated LNG regasification
elevate the high pressure of the cycle to a value suitable for process configurations. Prog Energy Combust Sci 2018;69:1e27.
pipeline transmission making the transcritical option even [6] Krasae-in S, Stang JH, Neksa P. Development of large-scale
more versatile. hydrogen liquefaction processes from 1898 to 2009. Int J
Hydrogen Energy 2010;35:4524e33.
[7] Cardella U, Decker L, Klein H. Economically viable large-scale
hydrogen liquefaction. IOP Conf Ser Mater Sci Eng
Funding 2017;171:012013.
[8] Aasadnia M, Mehrpooya M. Large-scale liquid hydrogen
This research did not receive any specific grant from funding production methods and approaches: a review. Appl Energy
agencies in the public, commercial, or not-for-profit sectors. 2018;212:57e83.
22436 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 2 4 2 5 e2 2 4 3 6

[9] Dincer I, Acar C. Review and evaluation of hydrogen [22] Bisio G, Massardo A, Agazzani A. Combined helium and
production methods for better sustainability. Int J Hydrogen combustion gas turbine plant exploiting liquid hydrogen
Energy 2015;40:11094e111. (LH2) physical exergy. J Eng Gas Turbines Power
[10] Demir ME, Dincer I. Cost assessment and evaluation of 1996;118:257e64.
various hydrogen delivery scenarios. Int J Hydrogen Energy [23] Fyke A, Li D, Crane P, Scott DS. Recovery of
2018;43:10420e30. thermomechanical exergy from cryofuels. Int J Hydrogen
[11] Gahleitner G. Hydrogen from renewable electricity: an Energy 1997;22:435e40.
international review of power-to-gas pilot plants for [24] Agazzani A, Massardo AF, Korakianitis T. An assessment of
staionary applications. Int J Hydrogen Energy the performance of closed cycles with and without heat
2013;38:2039e61. rejection at cryogenic temperatures. J Eng Gas Turbines
[12] Maroufmashat A, Fowler M. Transition of future energy Power 1999;121:458e65.
system infrastructure; through power-to-gas pathways. [25] Cheng K, Qin J, Dang C, Lv C, Zhang S, Bao W.
Energies 2017;10:1089. Thermodynamic analysis for high-power electricity
[13] Hirscher M. Handbook of hydrogen storage. Wiely-VCH generation systems based on closed-Brayton-cycle with
Verlag GMBH&Co.KGaA; 2010. ISBN: 978-3-527-32273-2. finite cold source on hypersonic vehicles. Int J Hydrogen
[14] Taylor JB, Alderson EA, Kalyanam KM, Lyle AB, Philips LA. Energy 2018;43:14762e74.
Technical and economic assessment of methods for the [26] Zhang N, Lior N. A novel Brayton cycle with the integration of
storage of large quantities of hydrogen. Int J Hydrogen liquid hydrogen cryogenic exergy utilization. Int J Hydrogen
Energy 1986;11:5e22. Energy 2008;33:214e24.
[15] Steward D, Saur G, Penev M, Ramsden T. Lifecycle cost [27] Laouir A. PCT patent application N PCT/DZ2018/050001,
analysis of hydrogen versus other technologies for electrical WIPO organisation.
energy storage. National Renewable Energy Laboratory; [28] Laouir A. LNG regasification cycles with supercritical/
November 2009. Technical Report NREL/TP-560-46719. transcritical regenerative loop. In: Proc. 30th international
[16] Sadaghiani MS, Mehrpooya M. Introducing and energy conference on efficiency, cost, optimization, simulation and
analysis of a novel cryogenic hydrogen liquefaction process environmental impact of energy systems (ECOS),; 2-6 july
configuration. Int J Hydrogen Energy 2017:1e18. XXX. 2017 [San Diego, California, USA].
[17] Trevisani L, Fabbri M, Negrini F, Ribani PL. Advanced energy [29] Szargut J, Szczygiel I. Utilization of the cryogenic exergy of
recovery systems from liquid hydrogen. Energy Convers liquid natural gas (LNG) for the production of electricity.
Manag 2006;48:146e54. Energy 2009;34:827e37.
[18] Yang C, Ogden J. Determining the lowest-cost hydrogen [30] Reynolds WC. Thermodynamic properties in SI. Department
delivery mode. Int J Hydrogen Energy 2007;32:268e86. of Mechanical Engineering. Stanford University; 1979.
[19] Voth RO, Parrish WR. Studies of hydrogen liquefier efficiency [31] Krasae-in S. Efficient hydrogen liquefaction processes [PhD
and the recovery of the liquefaction energy. Cryogenic thesis]. Norway: NTNU Trondheim; 2013.
division. National Bureau of Standards; 1977. p. 77e862. Vol. [32] Petitpas G, Aceves SM, Matthews MJ, Smith JR. Para-H2 to
NBSIR. ortho-H2 conversion in a full-scale automotive cryogenic
[20] Oshima K, Ishizaki Y, Kameyama S, Akiyama M, Okuda M. pressurized hydrogen storage up to 345 bar. Int J Hydrogen
The utilization of LH2 and LNG cold for generation of electric Energy 2014;39:6533e47.
power by a cryogenic-type Stirling engine. Cryogenics [33] Szargut J, Morris DR, Stewart FR. Exergy analysis of thermal,
1978:617e20. chemical and metallurgical processes. New York:
[21] Furuhama S, Nakajima T, Honda T. Rankine cycle engines for Hemisphere Publishing; 1988.
utilization of LH2 car fuel as a low-temperature source. Int J
Hydrogen Energy 1993;18:149e55.