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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7

Available online at www.sciencedirect.com

ScienceDirect

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

Optimization of Liquid Organic Hydrogen Carrier

(LOHC) dehydrogenation system

Nihal Rao a, Ashish K. Lele b, Ashwin W. Patwardhan a,*

a
Department of Chemical Engineering, Institute of Chemical Technology, Mumbai, 400019, India
b
National Chemical Laboratory, Dr.Homi Bhabha Road, Pune, 411008, India

highlights graphical abstract

Integration of perhy-
drodibenzyltoluene dehydrogena-
tion reactor with the energy
supplied by hydrogen burner.

Modelling and simulation have
been carried out integrating
DWSim and Python.

0.3 wt% Pt/Al2O3 kinetic data is
used to simulate dehydrogenation
reactor.

Different dehydrogenation reactor
conditions and hydrogen burner
efficiency were optimized.

Cost analysis was performed
based on the catalyst and heat
exchanger for maximum hydrogen
production.

article info abstract

Article history: In this paper, the perhydrodibenzyltoluene dehydrogenation flowsheet has been simu-
Received 30 December 2021 lated. Modelling of the dehydrogenation reactor has been performed using the 1-D model.
Received in revised form External and internal mass transfer resistances are also considered. Non-isothermal pellet
6 April 2022 condition has been considered for simulating the dehydrogenation reactor. The flowsheet
Accepted 20 June 2022 simulation has been carried out in DW-Sim v 6.5.2 integrated with the reactor model coded
Available online 20 July 2022 in Python. NET. The dehydrogenation reactor is operated at a feed temperature between
523 K 613 K, a wall temperature of 623 K and 653 K, and a reactor pressure maintained at
Keywords: 1.2 atm. The amount of catalyst required for the perhydrodibenzyltoluene (PDBT) dehy-
LOHC drogenation reactor is evaluated such that the conversion reaches 99%. The process
Perhydrodibenzyltoluene flowsheet has been simulated to produce 10 Nm3/hr of industrial-grade hydrogen. The
Simulation effects of feed temperature, wall temperature, and hydrogen burner efficiency on various
DWSim system requirements, including catalyst weight, energy supplied to the dehydrogenation
reactor, areas of the heat exchanger, and hydrogen production from the reactor, have been
discussed. Preliminary cost optimization based on the heat exchangers and catalyst at

* Corresponding author.
E-mail address: aw.patwardhan@ictmumbai.edu.in (A.W. Patwardhan).
https://doi.org/10.1016/j.ijhydene.2022.06.197
0360-3199/© 2022 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7 28531

Python various feed temperatures, reactor wall temperature, and hydrogen burner efficiency has
Dehydrogenation been carried out.
© 2022 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

storage capacity of 5.7 wt% 7.1 wt%, which is higher than the
Introduction US DOE (2020) mandate of about 5.5 wt% [16e22]. In a
LOHC þ compound, hydrogen can be stored for a long time
Hydrogen is considered to be one of the energy sources. It has without self-discharge, thus favouring seasonal energy storage
the highest gravimetric energy density of 120 MJ/kg and does and transportation to remote areas where dehydrogenation
not emit greenhouse gases compared to other carbon-based can be performed [23]. This reversible process allows rapid
fuels such as coal, gasoline [1,2]. One of the challenges faced hydrogenation and dehydrogenation without consuming the
by the hydrogen infrastructure is the storage and trans- compounds forming the LOHC system, and the released prod-
portation of hydrogen. Due to its low volumetric density, uct only consists of hydrogen.
hydrogen is commonly transported using pressurized The most important consideration for the dehydrogena-
hydrogen (200 bare700 bar) and liquefied hydrogen (252  C), tion of the LOHC is the nature of the compound, the catalyst
but these are pretty expensive and require unique infra- used, and the reaction mechanism between the catalyst and
structure [2,3]. Hydrogen is also stored using a metal-organic compound. Transportation of hydrogen from the production
framework (MOF), hyper cross-linked polymers, and carbon- site to the consumer site without compromising the ecological
based materials. The hydrogen stored by this method is and biological factors is complicated [24]. LOHC is considered
maximum at elevated pressures and low temperatures. one of the ways for transporting hydrogen efficiently and
However, hydrogen release occurs at low pressure and high safely. The only limitation with the LOHC dehydrogenation is
temperatures [4e7]. Since the system is solid, special infra- the high energy requirement for hydrogen release.
structure is required for transportation, mainly designed for Some aspects of LOHC property have to be considered,
liquid systems. Metal hydrides are also considered to be the such as high storage density, low vitality, low reaction
potential hydrogen storage system with a high storage ca- enthalpy and temperature, high reversibility, and low envi-
pacity of 20.9 wt%. Hydrogen is adsorbed and desorbed using ronmental hazard. Many hydrogen carriers have been studied
metals and metal alloys, which include simple hydrides and evaluated based on the Liquid hydrogen carrier's avail-
(MgH2, LiH), borohydrides (NaBH4, LiBH4), alanates (LiAlH4, ability, cost, and toxicological profile. It is considered one of
NaAlH4), transition metal hydrides, and complex hydrides the ways where hydrogen is transported at ambient temper-
[2,8e14]. The problem faced by the metal hydride is the harsh ature. The Liquid hydrogen carrier has a low energy demand
operating conditions, decrease in storage capacity due to and also has a lower cost estimation.
contamination, and difficulty in achieving reversibility due to The most commonly used LOHC, which was investigated
limitations in thermodynamics [15]. Metal hydrides also thoroughly, are N-ethycarbalzole (NECZ) -perhydro-N-ethyl-
require special infrastructure due to their solid nature. carbazole, methylcyclohexane (MCH)-toluene, dibenzylto-
Hydrogen transportation using the current infrastructure is luene (DBT) eperhydrodibenzyltoluene (PDBT) [9,25,26]. Many
economical and minimizes the need for additional equip- other LOHC's were also discovered, consisting of eutectic
ment. Liquid organic hydrogen carriers (LOHC) have proved to mixtures like biphenyl-diphenylmethane and heteroaromatic
be an efficient hydrogen storage technique [16]. The previous compounds such as 1-methylindole, 2-methylindole, 1-ethyl
studies [17e20] show that the Liquid Organic Hydrogen Carrier indole, etc. These were not studied extensively for commer-
(LOHC) is the most suitable option for hydrogen storage and cial application [27e32]. The physical and chemical properties
release in terms of energy consumption and cost of trans- of the commonly used LOHC are given in Table 1.
portation after compressed hydrogen and liquefied hydrogen From Table 1, the dibenzyltoluene e perhy-
storage. drodibenzyltoluene is the suitable LOHC for hydrogen storage
The technology consists of two processes, the hydrogen lean and transportation. It has a low melting point, high boiling
LOHCe and hydrogen-rich LOHCþ, which can be transported point, and high hydrogen storage capacity, making them
using the existing infrastructure like pipelines and trucks. A suitable for storage and transportation in pipelines and stor-
complete LOHC system consists of a reversible cycle of hydro- age tank trucks.
genation and dehydrogenation for sequential hydrogen storage The published literature in Table 2 shows various LOHC
and release. The amount of hydrogen released in the dehy- dehydrogenation process simulated in different simulation
drogenation system depends on various operational parame- software. Mizsey et al. [33] simulated the pilot plant scale of
ters. Many LOHC's have been used for the transportation and the MCH-toluene system based on the fixed bed reactor
storage of hydrogen, i.e., Benzene-cyclohexane, Methyl- 
operated at the temperature range between 260 C- 320  C, and
cyclohexane (MCH)-Toluene, Naphthalene-Decalin, N-ethyl- the hydrogen yield was obtained to be 98%. Hamayun et al.
carbazole (NECZ)-perhydro-N-ethylcarbazole and [16,18] simulated the MCH-toluene system in ASPEN Plus
Dibenzyltoluene (DBT)- perhydro-dibenzyltoluene (PDBT) software based on steam generation using hydrogen from the
system. Various literature cited the use of LOHC having the
28532 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7

Table 1 e Physical and Chemical properties of the commonly used LOHC [9,24,25].
Methylcyclohexane N-ethylcarbazole Dibenzyltoluene
Melting point (K) 147 341 247
Boiling point (K) 374 543 663
Hydrogen storage capacity 6.1 5.8 6.2
(wt %)
Enthalpy for hydrogenation/ 68.3 55 65.3
dehydrogenation (KJ/molH2)
Advantages
High availability
Requires lower enthalpy for
High thermal stability due to

Higher hydrogen storage hydrogenation and lower vapor pressure
capacity dehydrogenation.
High hydrogen storage
capacity
Disadvantages
Lower boiling point results in
High melting points make
High energy requirement
the additional usage of transportation difficult
condensation and
Lack of availability
purification.

High energy requirement

dehydrogenation process and the oxygen obtained from the DBT system to supply hydrogen for electricity generation
electrolysis process. The various catalyst used in the dehy- using fuel cells. The hydrogen released from the dehydroge-
drogenation process was also compared by varying operating nation reaction ranges between 1 and 5 bars.
temperature. Energy analysis was also performed using the Roland et al. [37] developed an integrated model based on
ASPEN Energy analyzer. The actual energy utilized by the the Dibenzyltoluene dehydrogenation reactor. The hydrogen
heating utilities, cooling utilities, carbon emission, and the produced is sent to a solid oxide fuel cell (SOFC) for power
total utilities was less than the target energy requirement. generation. The cascade model used in the simulation of DBT
Firman et al. [34] simulated the MCH dehydrogenation and and modeling of the SOFC was simulated using the Simulink
combined cycle for power generation and compared the sys- in MATLAB. Haupt et al. [38] integrated dehydrogenation of
tem efficiency with the Graz cycle. The system efficiency is DBT with the integration of Combined heat and power cycle,
calculated based on the power generated by the gas turbine which provide thermal energy not only for the dehydrogena-
and steam turbine to the difference in the calorific value of the tion process but also for providing energy to heat the building.
MCH as input material and toluene as an output material. The The sensitivity parameters include the electrical efficiency
proposed system has an efficiency of 54.6%, which is greater and self-consumption rate. The electrical efficiency is calcu-
than the Graz cycle-based system. Wang et al. [35] used the N- lated based on the amount of energy generated by the fuel cell
ethylcarbazole derivatives to develop the model for analyzing to the power used by the auxiliary. Another term known as the
the energy consumption in hydrogenation and dehydroge- self-consumption rate shows the relation between the elec-
nation processes. The system efficiency is calculated by the trical energy used and the electrical energy supplied by the
amount of energy that goes into the fuel cell to the amount of CHP system. The overall concept was simulated in Simulink
energy required for transportation, compression, and dehy- software in MATLAB. Nassem et al. [39] developed a compre-
drogenation reaction. The system efficiency was found to be hensive parametric investigation into the dehydrogenation of
69.17% and 88.74% with and without energy recycling- LOHC using various aromatic compounds and heteroaromatic
respectively. Brigljevic et al. [36] developed the process compounds. The effect of methanation reaction using pa-
simulation of the dehydrogenation process together with the rameters like temperatures, methane split ratio, and pressure.
economic analysis to develop the industrial-scale production The system efficiency is calculated as the amount of methane
of 1000 Nm3/hr of hydrogen to the pipeline system. The LOHC produced in the methanation reactor to the sum of the
used in the simulation includes biphenyl-diphenylmethane, hydrogen and dehydrogenated compound present in the
2-(N-MethylBenzyl)-pyridine, N-phenylcarbazole and N-eth- product. Dibenzyltoluene was found to have been selected
ylcarbazole. The first three LOHC's are still in the experimental from the list of LOHC material to have the highest system
stage and are not commercially produced on a large scale. The efficiency of 46.7% with a 100% medium recovery. The process
fourth LOHC N-ethylcatbazole is not produced on a massive was simulated using ASPEN Plus. Asif et al. [40] simulated the
scale, and since the melting point of N-ethyl carbazole is high, dehydrogenation of the dibenzyltoluene system to determine
transportation of N-ethyl carbazole is still a challenge. The the exergy efficiency and exergy destruction rate. Exergy ef-
cost of the system is calculated based on the capital cost and ficiency is the ratio of the thermal efficiency of the existing
operating cost. system to the reversible of the system. Two flowsheet
The LOHC which has the most potential to be used in schemes were studied based on different process configura-
commercial application is the DBT-PDBT system. Braukner tions of the heat exchanger. A conversion reactor was used to
et al. [41] had discovered this potential LOHC which already simulate the dehydrogenation reaction of perhy-
had an application in the heat transfer fluid. They are desig- drodibenzyltoluene where the hydrogen released was used to
nated under the industrial name Marlotherm LH and Marlo- run the turbine. The overall exergy efficiency was found to be
therm MH. Hydrogeneous technologies also used the PDBT- 91.06% and 94.86% for the two flowsheets, respectively.
Table 2 e Simulations reported in the literature for various LOHC dehydrogenation systems.
Liquid Organic hydrogen Simulation Operating Operating Dehydrogenation Catalyst used Reaction Hydrogen System Reference

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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7
carrier (LOHC) studied software dehydrogenation dehydrogenation reactor used in model flow rate and cost
temperature (K) pressure the simulation considered analysis
performed
MCH-toluene ASPEN Plus 573 Ke723 K 1.8 bar Isothermal plug flow Platinum on alumina Power law model/ e No [16]
reactor Langamuir
Heinshelwood
Hougen Watson
(LHHW)
MCH-toluene ASPEN Plus 573 Ke723 K 1 bare1.8 bar Isothermal plug flow Platinum on alumina Power law model/ e No [18]
reactor LHHW
MCH-toluene SIMUSOLV 573 Ke723 K 10 bar Adiabatic packed Sulfided noble metal Power law model e No [33]
bed reactor on alumina
MCH-toluene ASPEN-HYSYS 723 K 1.2 bare1.5 bar Isothermal plug flow Platinum on alumina Power law model 69,129 Nm3/hr No [34]
reactor
NECZ- perhydro-N- e 433.15 K 1.2 bar Isothermal plug flow Platinum on alumina Power law model e Yes [35]
ethylcarbazole reactor
2-(N-MethylBenzyl) Pyridine, ASPEN Plus 433 Ke613 K 1.5 bar Stoichiometric PdeC e 1000 Nm3/h Yes [36]
biphenyl-diphenylmethane, reactor
N phenylcarbzole, N-
ethylcarbazole
Dibenzyltoluene Simulink- 533 Ke583 K 1.2 bar Adiabatic packed Platinum on alumina Power law model 0.45 Nm3/hr No [37]
-perhydrodibenzyltoluene MATLAB bed rector
Dibenzyltoluene Simulink - 573 K 1 bar Isothermal packed Platinum on alumina e e No [38]
-perhydrodibenzyltoluene MATLAB bed reactor
Dibenzyltoluene ASPEN -Plus 573 Ke623 K 1 bar Isothermal plug flow Platinum on alumina Power law model e Yes [39]
eperhydrodibenzyltoluene reactor
Dibenzyltoluene ASPEN Plus 563 Ke593 K 2 bar Conversion reactor Platinum on alumina e e No [40]
-perhydrodibenzyltoluene
Dibenzyltoluene DWSim-Python 523 Ke613 K 1.2 bar Adiabatic packed Platinum on alumina Power law model 17 Nm3/hr- Yes Present
-perhydrodibenzyltoluene bed reactor 36 Nm3/hr study

28533
28534 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7

The previous studies were primarily focused on the in the supplementary material (Table S1). The mixing rule used
experimental analysis of the perhydrodibenzyltolune (PDBT) in estimating the physical property of the mixture, such as
dehydrogenation system. Many other well-defined systems density and viscosity is based on the molar average of the
with different LOHC were studied but the integrated 1-D components. The basic mixing rule of the Peng-Robinson
model with the quantity of catalyst, heat exchangers area, thermodynamic method has been used in the simulation.
the molar flow rate of perhydrodibenzyltoluene, and The process flow diagram of the system is given in Fig. 1.
hydrogen production with respect to wall temperature, feed The flowsheet is simulated in a steady state condition. The
temperature, and hydrogen burner efficiency were not studied blue arrow denotes the material stream, and the yellow arrow
previously. Furthermore, the cost analysis on the heat indicates the energy stream. Pehydrodibenzyltoluene is
exchanger and the catalyst that comprise the central part of pumped (Pump-01) at ambient temperature and is heated
the cost in this system has not been studied in detail with using the G-L heat exchanger (G-L HEX), L-L heat exchanger (L-
respect to feed temperature, wall temperature, and burner L HEX), and flue gas heat exchanger (Flue gas HEX). The G-L
efficiency. Therefore, perhydrodibenzyltoluene dehydroge- Heat exchanger (G-L HEX) transfers heat from the hot
nation system has been studied in detail using the combina- hydrogen gas produced from the dehydrogenation reactor
tion of DW-Sim and Python software with more focus on (PBR) to the cold perhydrodibenzyltoluene from the pump
earlier discussed points. (Pump-01). The L-L heat exchanger (L-L HEX) transfers heat
The simulation system consists of a DBT-PDBT dehydro- from the hot dibenzyltoluene (DBT), coming as a product from
genation system together with the hydrogen combustion the dehydrogenation reactor to the perhydrodibenzyltoluene
system. The simulation system incorporates a dibenzylto- (PDBT) from the G-L Heat exchanger (G-L HEX). The perhy-
luene dehydrogenation reactor, hydrogen burner, vapor- drodibenzyltoluene is heated to the required feed temperature
liquid separator, and a split controller. The catalyst consid- using the flue gas heat exchanger (Flue gas HEX). The flue gas
ered in the dehydrogenation system is 0.3 wt% Pt/Al2O3 [20]. heat exchanger (Flue gas HEX) transfers heat from the hot flue
From the previous study [42], which shows the Life cycle gas produced in the hydrogen burner to the perhy-
assessment (LCA) for the transportation of hydrogen from the drodibenzyltoluene from the Liquid-Liquid heat exchanger (L-
production site to the consumer site by comparing com- L HEX). The heated perhydrodibenzyltoluene is sent to the
pressed hydrogen (G-H2) and LOHC transportation. The con- packed bed reactor using the dehydrogenation reactor model
version of the LOHC dehydrogenation reactor for the LCA (PBR) coded in Python. The dehydrogenation reactor converts
analysis was taken as 99%. Achieving nearly 100% dehydro- perhydrodibenzyltoluene to hydrogen and dehydrogenated
genation conversion is technically challenging and non-viable product (dibenzyltoluene). The quantity of the catalyst
[43,44]. Therefore, we have considered 99% conversion of the required for the packed bed reactor (PBR) is estimated to
dehydrogenation reactor for our simulation. The flowsheet achieve 99% perhydrodibenzyltoluene (PDBT) dehydrogena-
study has been carried out using DW-Sim v 6.5.2 integrated tion reaction. The product is then sent to the liquid-gas
with the 1-D model coded in Python.NET for simulating the separator (V-L Separator). The hydrogen from the liquid-gas
dehydrogenation of the PDBT. separator contains a small quantity of vaporized dibenzylto-
luene which is further separated using the compound sepa-
rator (CS-01). The separated benzyltoluene from the
Process model and cost analysis compound separator (CS-01) is mixed with the benzyltoluene
from the liquid-gas separator (V-L Separator) using the stream
Process flowsheet mixer (MIX-01). The benzyltoluene is sent to the liquid-liquid
heat exchanger (L-L HEX) using the pump (PUMP-02) to
The simulation study for the process model is conducted in decrease the temperature of benzyltoluene. The hydrogen
DWSim v 6.5.2. A fixed bed reactor has been considered for the produced in the packed bed reactor (PBR) is then sent to the
dehydrogenation of PDBT, and the conversion reactor is used to gas-liquid heat exchanger (G-L HEX), which reduces the tem-
simulate the hydrogen burner. The 1-D model for the dehy- perature of hydrogen. A part of the hydrogen is used in the
drogenation reactor is coded in Python.NET and is integrated hydrogen burner (HTB), an exothermic reaction that will
with DW-Sim. The thermodynamic package used in the simu- provide the heat necessary for the dehydrogenation reaction.
lation is the “Peng-Robinson” model. Peng Robinson is a cubic The remaining hydrogen is sent to the fuel cell (HTF). The
equation of state which relates the temperature, pressure and hydrogen that is sent to the fuel cell and the hydrogen sent to
molar volume of a pure component or mixture at equilibrium. the burner is split using the stream splitter (SPLIT-01). The
The cubic equations are the simplest equation that simulta- amount of hydrogen that is split between the hydrogen sent to
neously represents the liquid and vapor phases behaviour. the fuel cell (HTF) and the hydrogen sent to the burner (HTB) is
Peng-Robinson is recommended for use in hydrocarbons and adjusted by the controller (ADJ-01). The controller adjusts the
non-condensables [45]. The binary interaction parameters of split operation in such a manner that the hydrogen supplied
the compounds available in the ChemSep library were derived to the hydrogen burner is the same as the energy required for
from the Chemsep PR BiP database [46]. The binary interaction the packed bed reactor to carry a dehydrogenation reaction.
parameters for the user-defined compounds such as dibenzyl- The air is compressed using the compressor (COMP-01) mixed
toluene and perhydrodibenzyltoluene were estimated using with the hydrogen in the stream mixer (MIX-02). The mixture
the group contribution method based on the critical volume of hydrogen and air (MHA) is sent to the hydrogen burner for
[47]. The tabular column for the binary interaction parameters the combustion reaction. The hydrogen burner is simulated
used in the Peng Robinson thermodynamic model is provided using the conversion reactor (RC-01). The outlet temperature
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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7 28535
Fig. 1 e Process flow sheet for the dehydrogenation of the perhydrodibenzyltoluene system.
28536 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7

of the flue gas is the same as the reactor wall temperature.

Table 3 e Properties of the catalyst [20].
Three recycle streams, termed recycle stream 1 (REC-1),
recycle stream 2 (REC-2), and recycle stream 3 (REC-3), are Catalyst Properties Value
used to circulate the perhydrodibenzyltoluene, dibenzylto- Catalyst 0.3 wt% Pt/Al2O3
luene, and flue gas to the heat exchangers. Particle diameter 3 mm
Particle density 1320 kg/m3
Particle void fraction 0.69
Physico-chemical Properties and kinetics
Tortuosity 1.31
shape factor 1
Since dibenzyltoluene and perhydrodibzytoluene are not Thermal conductivity of the catalyst 0.2073 W/m.K
found in the Chemsep library of the DW-Sim. The compound Pre-exponential factor (ko Þ 7.80Eþ04 m3/Kgcat.s
creation tool in the DW-Sim software is utilized to estimate Activation energy 118.17 kJ/mol
the property for the PDBT-DBT compounds using the combi-
nation of experimental data and Group contribution method.
Muller's experimental data [19,48] were used to specify the
Table 4 e Specification of the packed bed reactor in an
dibenzyltoluene (DBT) and perhydrodibenzyltoluene (PDBT) adiabatic mode of operation.
viscosity, specific heat, and density. The rest of the properties
PBR Specifications Value
like the heat of vaporization, vapor pressure, thermal con-
ductivity, surface tension are estimated using the Group Reactor Temperature 523 Ke613 K
Reactor diameter 0.1 m
Contribution Method. Joback technique is used for estimating
Bed void fraction 0.5
ideal gas heat capacity, Lee-Kesler for vapor pressure, Latini
Wall Temperature 623 Ke653 K
method for thermal conductivity of liquids [47]. The catalyst
considered for the process is a spherical 0.3 wt% Pt/Al2O3
(platinum on alumina) catalyst. The experimental perfor-
mance of the 0.3 wt% Pt/Al2O3 catalyst using the DBT-PDBT as The reaction is considered the endothermic reaction,
a Liquid hydrogen carrier was described by Bulgarin [20]. The which requires a certain amount of heat to be supplied to
reaction is considered as the first-order reaction with less initiate the reaction. This heat has to be supplied either by
than 10% deviation from the experimental data. using the external energy supply such as electricity, burners
or by using some form of an exothermic reaction, which can
Modeling of Dibenzyltoluene dehydrogenation reactor provide energy to the dehydrogenation reaction. Since the
kinetic mechanism follows the first-order reaction which is
The adiabatic packed bed reactor (PBR) is simulated using the expressed as
1-D model. The 1-D reactor model assumption is made to
r18HDBT ¼  kg C18HDBT (1)
simplify the numerical modeling and computation time for
DW-Sim. where kg is the overall diffusion resistance, which is the
reciprocal sum of the internal mass transfer resistance (Ri ),

Steady-state and external mass transfer resistance (Re ), is shown in Eq. (2).

Axial dispersion is considered to be negligible.
1

The rate of change of concentration and temperature in the kg ¼ (2)
Re þ Ri
radial direction is neglected.

The wall temperature is constant throughout the length of The external mass transfer coefficient is calculated from
the reactor the Sherwood number. The Sherwood number is calculated

Heat transfer takes place between the wall of the reactor from the Thoenes-Kramer correlation for the spherical cata-
and the fluid along the axial direction. lyst particle,

Pressure drop across the packed bed reactor is considered
Sh0 ¼ Re0:5 Sc0:33 (3)
negligible.
where Reynold number and Schmidt number are calculated
Tabulation for the properties of the catalyst, and specifica- based on the porosity of the catalyst, viscosity (ml ), density (rl ),
tion of the packed bed reactor, is given in Table 3 and Table 4. superficial velocity (us ), shape factor (gs ), and diffusion coef-
The dehydrogenation reaction for the perhy- ficient (DBC ). The Reynold number and Schmidt number are
drodibenzyltoluene is given by- computed using Eq. (4) and Eq. (5).
 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7 28537

therefore, the effectiveness factor will decrease. It has two

dp rl us factors associated while calculating the internal non-
Re ¼ (4)
ml isothermal effectiveness factor e the Arrhenius coefficient
(g) and the thermicity (b) shown in Eqs. (13) and (14).
The modified form of the Reynold number in terms of
catalyst void fraction (εc ) and shape factor (gs ) is given in Eq. Ea
g¼ (13)
(5). RTs

Re
Re0 ¼ (5) ðDHrxn ÞCs Deff
ð1  εc Þgs b¼ (14)
keff Ts
The Schmidt number in terms of liquid viscosity (ml Þ, liquid
The Maxwell equation gives the effective thermal con-
density (rl Þ, and diffusion coefficient (DBC ) is given in Eq. 6
ductivity for the spherical particle in terms of liquid conduc-
ml tivity (kf ), catalyst conductivity (km ) and catalyst porosity (εc ) is
Sc ¼ (6)
rl DBC described in Eq. (15).
Polson criteria is used to calculate the diffusion coefficient
km ð2km þ kf  2εc ðkm  kf ÞÞ
(DBC Þ between the bulk liquid, i.e. perhydrodibenzyltoluene keff ¼ (15)
2km þ kf þ εc ðkm  kf Þ
and catalyst, as shown in Eq. (7).
The equation used for estimating the non-isothermal in-
9:4  1015 T ternal effectiveness factor is calculated using Eq. (16) [49].
DBC ¼ 1 (7)
mb M3b
1 bg
h ¼ ð Þexpð Þ (16)
Effective diffusivity (Deff ) is calculated from diffusion co- 4 5
efficient (DBC ), porosity (εc ), and tortuosity (tt )
The internal mass transfer resistance (Ri ) show the actual
DBC εc rate on the catalyst surface, which is given in Eq. (17).
Deff ¼ (8)
tt
1
Ri ¼ (17)
The external mass transfer coefficient (kc ) is calculated hk
from the Sherwood number using Eq. (9).
1-D model is used for the simulation of the packed bed
0
Sh DBC ð1  εc Þgs reactor. The 1-D model describes the variation of concentra-
kc ¼ (9) tion and temperature in the axial direction. The general mass
dp εc
balance for the 1-D model is shown in Eq. (18).
The external resistance is calculated based on the mass
transfer coefficient and the external surface area of the cata- dCi r ri
¼  bulk (18)
lyst. The external surface area per mass of the solid is calcu- dz us
lated from the bed porosity (εb ), bulk density (rb ), and particle The heat transfer coefficient between the wall to the fluid is
diameter (dp ), which is given in Eq. (10). described by Dixon and Cresswell (1979) [47,50] in Eq. (19).
6ð1  εb Þ dp
ac ¼ (10) Nuwf ¼ 0:523ð1  ð ÞÞPr0:33 Re0:738 (19)
rb dp dt
The external mass transfer resistance from the bulk liquid The wall-to-fluid heat transfer coefficient is calculated
to the catalyst Re is given in Eq. 11 using Eq. (20) from Nusselts number (Nuwf ).

1 Nuwf kf
Re ¼ (11) hwf ¼ (20)
kc ac dp
The internal mass transfer resistance is calculated based
The wall temperature is considered constant, and thus
on the internal effectiveness factor, which depends on the
temperature variation across the reactor, which transfers heat
Thiele modulus. Thiele modulus is calculated from the effec-
from the wall to the fluid, is expressed in Eq. (21).
tive diffusivity (Deff ), particle diameter (dp ), and rate constant
(k). The Thiele modulus for the first-order reaction is given in 4hwf
dTi DHrxn ri rbulk dt
ðTi  Tw Þ
Eq. (12). ¼  (21)
dz rl us Cp rl us Cp
sffiffiffiffiffiffiffiffi
dp k Eq. (18) and Eq. (21) are coupled and are solved simulta-
4¼ (12) neously. The above equations simulate the 1-D model for the
2 Deff
dehydrogenation of perhydrodibenzyltoluene with enough
The analysis for the non-isothermal pallets for the endo- accuracy. The produced hydrogen is sent to the splitter,
thermic reactions leads to a reduction in the effectiveness where the hydrogen can be separated for the fuel cell and the
factor. For the endothermic reaction, the temperature inside energy required to supply energy to the dehydrogenation
the pallets decreases when the reaction occurs at the surface; reactor.
28538 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7

Flowsheet simulation and numerical technique used to of the heat exchanger is estimated from the cost data provided
simulate dehydrogenation reactor by Corripio et al. [52]. The estimated base cost of the heat
exchanger in USD in terms of heat exchanger area is given as
The flowsheet in Fig. 1 consists of various unit operations and
processes such as heat exchangers, pumps, vapor-liquid CBð1982Þ ¼ expð8:202 þ 0:01506  ln A þ 0:06811  ðln AÞ2 Þ (22)
separator, compound separator, reactor, and burners. The The exchange type cost factor for the fixed head heat
heat exchangers in the flowsheet are simulated by specifying exchanger in terms of the heat exchanger area is given in
the hot fluid or cold fluid outlet temperatures and the overall
heat transfer coefficient. The overall heat transfer coefficient FD ¼ expð  0:9003 þ 0:0906  ln AÞ (23)
considered for the G-L Heat exchanger, L-L heat exchanger, and
The material cost of the construction cost factor for the SS
flue gas heat exchanger is considered as 20 W/m2K, 100 W/m2K, 316 is given in Eq. (24).
20 W/m2K, respectively from the tabular column given by Perry
[51]. The pump is simulated by giving the outlet pressure and FM ¼ 1:4144 þ 0:239269  ln A (24)
efficiency to get the power required. The 1-D model for the
The overall cost of the heat exchanger in terms of the base
dehydrogenation reactor was solved using the odeint equation
cost, exchanger type cost, and material cost of construction is
solver module of the python software. The odeint module
given in Eq. (25).
computes the coupled first-order ordinary differential equation
for stiff and non-stiff systems. The odeint module allows CE ¼ CBð1982Þ  FD  FM (25)
switching between the Adam method for a non-stiff ordinary
Since the capital cost of the heat exchanger in the above
differential equation and the backward differential formula
equation is formulated in the year 1982. Therefore due to
(BDF) for stiff differential equations. Since the time taken for
inflation and other factors, there has been a change in the cost
computing the coupled ordinary differential equation remains
of the heat exchanger for the current year. The chemical en-
the same when switching between the stiff and non-stiff
gineering plant cost index can be used to calculate the cost of
solver. Therefore the Adam method for the non-stiff ordinary
the heat exchanger for the current year. The cost index ratio
differential equation is used for computing the 1-D model in
calculation is given in Eq. (26).
our system. The solution to the problems defined by the initial
conditions was obtained using the odeint solver with the CEð 1982Þ I1982
known variable step size. The tolerance used to solve the ode ¼ (26)
CEð2021Þ I2021
for each length step is 1.49e-08. The wall temperature of the
reactor is kept at 623 K and 653 K to minimize the vapor for- The cost index of the heat exchanger for 1982e1983 is
mation of benzyltoluene and dibenzyltoluene. Low vapor for- 372.1, while the cost index for 2020e2021 is 698.4.
mation helps in reducing condenser duty. The vapor-liquid The catalyst cost is calculated based on the weight of the
separator (V-L separator) is simulated by calculating the catalyst required for the dehydrogenation reactor, which is
average inlet pressure, separating 91%e96% of the hydrogen 0.3 wt% Pt/Al2O3, where only the weight percentage of plat-
mainly depending on the wall temperature. The rest is 4%e9% inum is considered as a part of the cost estimation. The cost of
dibenzyltoluene, and a relatively smaller percentage of perhy- the platinum catalyst varies according to the market demand.
drodibenzyltoluene. The pure hydrogen is obtained by the The current cost of platinum is $33340 per kilogram as per the
addition of a compound separator (CS-01), which acts like a market cost.
condenser and is simulated by giving the outlet mass per- The major capital cost is the sum of the heat exchangers
centage of hydrogen. The hydrogen from the compound is split and catalyst costs. The major capital cost of the system is
between the one to be supplied to the fuel cell (HTF) and the one given in Eq. (27).
to be supplied to the burner (HTB) to fulfil the dehydrogenation
CM ¼ CE þ CC (27)
reactor energy requirement. The split is adjusted using the
controller (ADJ-01), which has an error tolerance of 1e-05. Air is
compressed using the adiabatic compressor (COMP-1), which
has an efficiency of 75%. The air from the compressor and the Results and discussion
hydrogen from the HTB is mixed in the stream mixer (MIX-02).
Some part of the hydrogen that is sent to the hydrogen burner Temperature variation along the length of the reactor
is simulated using the conversion reactor. Hydrogen burner
efficiency is calculated as the amount of hydrogen fuel burned Since the reactor is adiabatic, the wall temperature influences
during the combustion reaction to the amount of hydrogen fuel the temperature of the feed across the length of the reactor.
supplied. The specification required for the conversion reactor The temperature variation across the reactor length at the
is the conversion of the combustion reaction and the outlet flue four different feed temperatures of 613 K, 573 K, 543 K, and
gas temperature. 523 K is represented in Fig. 2 (a)e(d) for the wall temperatures
of 623 K and 653 K, respectively. The change in reactor length
Cost analysis of the heat exchanger and catalyst based on the feed temperature and wall temperature for the
99% dehydrogenation conversion is also shown in
The cost considered in the flowsheet, shown in Fig. 1, are the Fig. 2(a)e(d).
heat exchangers and catalyst. The other auxiliary equipment As seen from the graphical trend in Fig. 2 (a)e(b), the
does not contribute to major additional costs. The capital cost temperature drops initially due to the dominance of
 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7 28539

Fig. 2 e Temperature variation along the length of the reactor for 99% dehydrogenation reactor conversion at various feed
temperatures (a) 613 K (b) 573 K (c) 543 K (d) 523 K with a constant wall temperature kept at 623 K and 653 K.

endothermicity, but after a certain length, the influence of the the wall temperature is 653 K compared to the wall tempera-
heat transfer across the reactor length increases. There is a ture of 623 K.
slight deviation from the trend for the feed temperature at
523 K and 543 K, as seen from Fig. 2 (c)e(d) as the influence of Energy supplied to dehydrogenation reactor and heat
heat loss due to endothermicity is minimal and the heat exchanger area requirement
transfer from the wall to the fluid increases.
The influence of heat transfer due to wall temperature The process flow diagram described in Fig. 1 consists of three
increases with a decrease in the feed dehydrogenation tem- heat exchangers for heating the feed to the desired feed
perature. From Fig. 2(a)e(b), it is clearly seen that at the wall temperature and a hydrogen burner where the energy is
temperature of 653 K, the temperature drop due to endo- supplied to the dehydrogenation reactor. The energy required
thermicity is slightly lower as compared to one when the wall for heating the dehydrogenation reaction for different feed
temperature is kept at 623 K. Similar trends also show for the temperatures at 523 Ke613 K with constant wall temperatures
feed temperature of 523 K and 543 K shown in Fig. 2(c)e(d), of 623 K and 653 K and at different hydrogen burner efficiency
where the higher wall temperature of 653 K has the more of 90% and 60% is given in Fig. 3.
significant influence of heat transfer from the wall to the fluid As the temperature of the feed increases, the energy sup-
as compared to the wall temperature of 623 K. plied by the hydrogen burner to the dehydrogenation reactor
The length of the reactor also changes depending on the decreases. The energy supplied by the burner also increases
feed temperature and wall temperature. The length of the with the increase in the wall temperature. The efficiency of
reactor increases when lowering the feed temperature and the hydrogen burner also influences the energy supplied to
increasing the wall temperature. The reactor length is lower the dehydrogenation reactor. A decrease in the hydrogen
when the feed temperature is 613 K, and the wall temperature burner efficiency increases the energy supplied to the dehy-
is 623 K. Similarly, the length of the reactor increases when drogenation reactor. The energy balance for the packed bed
28540 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7

Fig. 3 e Energy supplied to dehydrogenation reactor for different feed temperatures at the constant wall temperatures of
623 K and 653 K for the hydrogen burner efficiency (n) of 60% and 90%.

Fig. 4 e Hydrogen production from the reactor for different dehydrogenation feed temperatures at the constant wall
temperatures of 623 K and 653 K for the hydrogen burner efficiency (n) of 60% and 90%.

reactor indicates that increasing the feed temperature and temperature. The hydrogen burner efficiency also affects the
lowering the wall temperature lowers the enthalpy difference hydrogen production in the reactor, as seen in Fig. 4. The
between the inlet and outlet streams. The splitter (SPLIT-01) decrease in the hydrogen burner efficiency increases the
adjusts the hydrogen supply to the burner so that a sufficient hydrogen production in the reactor. The high energy required
amount of hydrogen is burned to supply heat for the dehy- to supply heat to the dehydrogenation reactor supplements
drogenation reactor. the need for a higher hydrogen requirement to be supplied to
The hydrogen production from the reactor with varying the burner, as shown in Fig. 3.
feed temperature and hydrogen burner efficiency at different Three heat exchangers were used to raise the feed tem-
wall temperatures is given in Fig. 4. The feed temperatures perature from 303 K to various feed temperatures varying
ranges from 523 K to 613 K, keeping the wall temperature from 523 K to 613 K. The three heat exchangers include
constant at 623 K and 653 K. The hydrogen burner efficiency is hydrogen-DBT heat exchanger (G-L Heat exchanger), PDBT-
kept at 60% and 90%. DBT heat exchanger (L-L Heat exchanger), and flue gas heat
The hydrogen production from the reactor increases with exchanger. The area required by the three heat exchangers at
the increase in the wall temperature and at lower feed various feed temperatures ranging from 523 K to 613 K with
 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7 28541

Fig. 5 e Area required for Gas-Liquid heat exchanger (G-L HEX) for different dehydrogenation feed temperatures at the
constant wall temperatures of 623 K and 653 K for the hydrogen burner efficiency (n) of 60% and 90%.

Fig. 6 e Area required for Liquid-Liquid heat exchanger (L-L HEX) for different dehydrogenation feed temperatures at the
constant wall temperatures of 623 K and 653 K for the hydrogen burner efficiency (n) of 60% and 90%.

different wall temperatures at 623 K and 653 K for the wall temperature of 623 K. The gas-liquid heat exchanger area
hydrogen burner efficiency of 60% and 90% is given in Fig. 5., also increases with the decrease in the hydrogen burner effi-
Fig. 6., and Fig. 7 (a)e(b). ciency. Decreasing the hydrogen burner efficiency increases
The area for the gas-liquid heat exchanger (G-L HEX) does hydrogen production, as shown in Fig. 4.
not decrease significantly with the increase in the feed tem- The area of the liquid-liquid heat exchanger (L-L HEX) in-
perature, as given in Fig. 5. The outlet temperature of the creases with the feed temperature, as shown in Fig. 6. The
hydrogen remains close to the wall temperature. Since the LMTD decreases with an increase in the feed temperature,
temperature of perhydrodibenzyltoluene remains at room therefore increasing the area of the heat exchanger. The liquid-
temperature, the LMTD (Logarithmic mean temperature dif- liquid heat exchanger (L-L HEX) area also decreases with the
ference) of the heat exchanger does not change significantly. increase in the wall temperature. This is attributed to the
Increasing the wall temperature to 653 K increases the gas- higher LMTD, and lower heat exchange duty for the wall tem-
liquid heat exchanger (G-L HEX) area, which is attributed to perature of 653 K as compared to the wall temperature of 613 K.
the higher heat duty of the heat exchanger as compared to the Decreasing the hydrogen burner efficiency increases the heat
28542 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7

Fig. 7 e Area required for Flue gas heat exchanger (Flue gas HEX) for different dehydrogenation feed temperatures at
hydrogen burner efficiency (n) of 60% and 90% with a constant wall temperature kept at (a) Tw ¼ 623 K (b) Tw ¼ 653 K.

exchanger area at various feed and wall temperatures due to an Weight of catalyst based on 99% dehydrogenation
increase in the flow rate of perhydrodibenzyltoluene. conversion
For the flue gas heat exchanger (flue gas HEX), which is
given in Fig. 7(a)e(b), the area increases with the increase in The amount of catalyst required for the dehydrogenation
the feed temperature. This is caused by the decrease in the reactor can be used to determine the catalyst cost. Catalyst
LMTD with the increase in the feed temperature. Similarly, weight depends on the reaction kinetics and hydrogen pro-
increasing the wall temperature decreases the area of the flue duction. The catalyst weight of 0.3 wt% Pt/Al2O3 required for
gas heat exchanger (flue gas HEX). The LMTD of the heat 99% dehydrogenation conversion is given in Fig. 8 for the feed
exchanger increases with the increase in the wall tempera- temperature ranging from 523 K to 613 K with the wall tem-
ture. Therefore, the flue gas heat exchanger area is lower for peratures of 623 K and 653 K at different hydrogen burner
the wall temperature of 653 K when compared with the wall efficiency of 60% and 90%.
temperature of 623 K. Decreasing the hydrogen burner effi- The catalyst weight required for 99% conversion of dehy-
ciency also increases the area of the flue gas heat exchanger drogenation reaction is based on the bulk density of the
with varying feed temperatures and wall temperature. This is catalyst. The larger the quantity of catalyst, the higher will be
due to the increase in hydrogen production from the reactor. the cost to procure catalyst. It is observed that a lower

Fig. 8 e Catalyst weight of 0.3 wt % Pt/Al2O3 for different feed temperatures at the constant wall temperatures of 623 K and
653 K for the hydrogen burner efficiency (n) of 60% and 90%.
 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7 28543

Fig. 9 e Cost of catalyst at various feed temperatures for the hydrogen burner efficiency at two different wall temperatures
623 K and 653 K for the hydrogen burner efficiency at (a) 90% (b) 60%; Cost of G-L Heat exchanger at various feed
temperatures for the hydrogen burner efficiency at two different wall temperatures 623 K and 653 K for the hydrogen burner
efficiency at (c) 90% (d) 60%; Cost of L-L Heat exchanger at various feed temperatures for the hydrogen burner efficiency at
two different wall temperatures 623 K and 653 K for the hydrogen burner efficiency at (e) 90% (f) 60%; Cost of flue gas heat
exchanger at various feed temperatures for the hydrogen burner efficiency at two different wall temperatures 623 K and
653 K for the hydrogen burner efficiency at (g) 90% (h) 60%.; Major capital cost at various feed temperatures for the hydrogen
burner efficiency at two different wall temperatures 623 K and 653 K for the hydrogen burner efficiency at (i) 90% (j) 60%.
28544 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7

quantity of catalyst is required at higher dehydrogenation The cost of the flue gas heat exchanger (Flue gas HEX), as
feed temperatures, which reduces the cost for loading the shown in Fig. 9(g)e(h), increases when the feed gas tempera-
catalyst compared to the one when the reactor is operated at a ture is increased. Increasing the wall temperature decreases
lower dehydrogenation feed temperature. The catalyst the cost of the flue gas heat exchanger due to higher loga-
requirement reduces since the reaction rate is higher at higher rithmic mean temperature difference (LMTD), which reduces
feed temperatures than at lower feed temperatures. The the area. Decreasing the hydrogen efficiency increases the
catalyst requirement at the wall temperature of 653 K is higher cost of the flue gas heat exchanger. Compared to the other
than when the wall temperature is kept at 623 K. Higher wall heat exchangers, the flue gas heat exchanger cost is compar-
temperature requires a higher energy requirement for the atively higher than other heat exchangers, especially at the
dehydrogenation reactor to be supplied by the burner, which wall temperature of 623K and the feed temperature of 613 K
means more hydrogen has to be produced. Higher hydrogen due to lower overall heat transfer coefficient (U) and LMTD.
production requires a higher quantity of catalyst for dehy- The major capital cost of the process, as shown in
drogenation reaction. For instance, when the feed tempera- Fig. 9(i)e(j), increases with an increase in the feed tempera-
ture is kept at 613 K, the weight of the catalyst is 115 kg when ture. This is attributed to the higher catalyst cost than the heat
the wall temperature is kept at 623 K and 121 kg when the wall exchangers cost, including G-L heat exchanger, L-L heat
temperature is kept at 653 K for the burner efficiency of 90%. exchanger, and flue gas heat exchanger. Increasing the wall
When the efficiency of the burner decreases, the weight of the temperature also increases the major cost except for the feed
catalyst for the corresponding feed temperature increases due temperature at 613K. This is due to the higher cost of the flue
to the increase in hydrogen production. Low hydrogen burner gas heat exchanger at the wall temperature of 623 K compared
efficiency means more hydrogen has to be supplied to the to the wall temperature at 653 K. Lower hydrogen burner ef-
burner to meet the energy demand for the dehydrogenation ficiency of 60% also contributes to higher major capital cost.
reactor. Therefore, the catalyst weight is lower when the
hydrogen burner efficiency is at 90% than when the hydrogen
burner efficiency is at 60%. Conclusions

Cost analysis of the heat exchanger and catalyst Detailed simulation of the dehydrogenation reactor based on
1-D model at different feed temperatures, wall temperature,
The major cost involved in this process consists of heat ex- and hydrogen burner efficiency to study the energy required
changers and catalysts. The estimated cost of the catalyst in by the dehydrogenation reactor, heat exchanger areas, and
the packed bed reactor, G-L heat exchanger, L-L heat quantity of catalyst required have been discussed. Optimi-
exchanger, flue gas heat exchanger and major capital cost zation of the cost based on the catalyst and heat exchangers
based on the change in feed temperature ranging from 523 K has also been studied. Lower feed temperatures at 523 K
to 613 K, wall temperature at 623 Ke653 K, and hydrogen favour a higher quantity of catalyst, a slightly higher Gas-
burner efficiency of 60% and 90% are given in Fig. 9(a)e(j). Liquid heat exchanger (G-L HEX), lower Liquid-Liquid heat
The cost of the catalyst with the change in feed tempera- exchanger (L-L HEX) area and lower flue gas heat exchanger.
ture with respect to wall temperature and hydrogen burner In contrast, a higher feed temperature at 613 K favours a
efficiency is given in Fig. 9(a)e(b). The cost of the catalyst de- lower quantity of catalyst, lower Gas-Liquid heat exchanger
creases when increasing the feed temperature since the lower (G-L HEX), higher Liquid-Liquid heat exchanger (L-L HEX) and
quantity of catalyst contributes to lower catalyst cost. The cost a higher flue gas heat exchanger area (Flue gas HEX).
of the catalyst increases when the wall temperature is Increasing the wall temperatures favours a higher quantity
increased since the hydrogen production is higher at a higher of catalyst, higher Gas-Liquid heat exchanger (G- L HEX),
wall temperature, as shown in Fig. 4. The lower hydrogen lower Liquid eLiquid heat exchanger (L-L HEX) and lower flue
burner efficiency also increases the cost of the catalyst. gas heat exchanger areas (Flue-gas HEX). Higher burner ef-
Lowering the burner efficiency increases the use of hydrogen, ficiency of 90% also favours a lower quantity of catalyst and
which is sent to the burner for supplying energy to the lower heat exchanger area. The major capital cost is mini-
dehydrogenation reactor. mum when the feed temperature is kept at 613 K with the
The cost of gas-liquid heat exchangers (G-L HEX) shown in wall temperature of 653 K and the hydrogen burner efficiency
Fig. 9(c)e(d) does not change significantly when increasing the at 90%. This system does not require additional utilities for
feed temperature. The cost of the gas-liquid heat exchanger heating and cooling. It utilizes the product from the reactor
(G-L HEX) increases when increasing the wall temperature and burner for heating the feed and the energy supplied to
since the heat duty of the heat exchanger is higher. The the dehydrogenation reactor, which minimizes the addi-
decrease in the hydrogen burner efficiency also increases the tional costs for utilities.
cost of the gas-liquid heat exchanger.
The cost of the liquid-liquid heat exchanger (L-L HEX) given
in Fig. 9(e)e(f) increases when the feed temperature increases Declaration of competing interest
due to the higher heat exchange area. Similarly, increasing the
wall temperature decreases the cost of the liquid-liquid heat The authors declare that they have no known competing
exchanger. Lower hydrogen burner efficiency also contributes financial interests or personal relationships that could have
to the higher cost of the liquid-liquid heat exchanger. appeared to influence the work reported in this paper.
 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7 28545

Greek symbols
Acknowledgment rc Density of the catalyst (kg/m3)
rbulk Bulk density (Kg/m3)
I would like to thank Reliance Industries for providing finan- rl Liquid density (Kg/m3)
cial support under Reliance fellowship.scheme. mb Viscosity of the fluid (Kg/ms)
εc Particle void fraction
εb Bed void fraction
Appendix A. Supplementary data tt tortuosity
h Effectiveness factor
Supplementary data to this article can be found online at 4 Theile modulus
https://doi.org/10.1016/j.ijhydene.2022.06.197. g Arrehenius coefficient
b Thermicity

Abbreviations
references
Nomenclature
A Heat exchanger areas (m2)
3
ra Rate of reaction (kgmcat sÞ [1] Vaidya PD, Rodrigues AE. Insight into steam reforming of
R Radius of the catalyst (m) ethanol to produce hydrogen for fuel cells. Chem Eng J
2006;117:39e49. https://doi.org/10.1016/j.cej.2005.12.008.
kc Mass transfer coefficient (m/s)
[2] Abe JO, Popoola API, Ajenifuja E, Popoola OM. Hydrogen
DBC Diffusion constant (m2/s) energy, economy and storage: review and recommendation.
Deff Effective diffusivity (m2/s) Int J Hydrogen Energy 2019;44:15072e86. https://doi.org/
n Order of the reaction 10.1016/j.ijhydene.2019.04.068.
Cs Surface concertation of the reactant (mol/m3) [3] Fikrt A, Brehmer R, Milella VO, Müller K, Bo€ smann A,
Ci Concentration of the reactant (mol/m3) Preuster P, et al. Dynamic power supply by hydrogen bound
to a liquid organic hydrogen carrier. Appl Energy
C0 Initial concentration of the reactant (mol/m3)
2017;194:1e8. https://doi.org/10.1016/j.apenergy.2017.02.070.
CB Base cost of the heat exchanger ($)
[4] Aakko-Saksa PT, Cook C, Kiviaho J, Repo T. Liquid organic
CE Overall cost of heat exchanger ($) hydrogen carriers for transportation and storing of
Cc Cost of the catalyst ($) renewable energy e review and discussion. J Power Sources
CM Major capital cost ($) 2018;396:803e23. https://doi.org/10.1016/
Sh' Sherwood number j.jpowsour.2018.04.011.
Re Reynold number [5] Shet SP, Shanmuga Priya S, Sudhakar K, Tahir M. A review on
current trends in potential use of metal-organic framework
Sc Schmidt number
for hydrogen storage. Int J Hydrogen Energy
Ti Temperature of the reactor 2021;46:11782e803. https://doi.org/10.1016/
Mb Molecular weight (Kg/kmol) j.ijhydene.2021.01.020.
dp Density of the particle (m) [6] Ren J, Musyoka NM, Langmi HW, Swartbooi A, North BC,
gs Shape factor Mathe M. A more efficient way to shape metal-organic
ac External surface area (m2/Kg) framework (MOF) powder materials for hydrogen storage
k Rate constant (m3/Kgcat.s) applications. Int J Hydrogen Energy 2015;40:4617e22. https://
doi.org/10.1016/j.ijhydene.2015.02.011.
Ri Internal rate constant (m3/Kgcat.s)
[7] Sawant SV, Yadav MD, Banerjee S, Patwardhan AW, Joshi JB,
Re External rate constant (m3/Kgcat.s) Dasgupta K. Hydrogen storage in boron-doped carbon
kg Overall resistance (m3/Kgcat.s) nanotubes: effect of dopant concentration. Int J Hydrogen
Ea Activation energy (KJ/mol) Energy 2021;46:39297e314. https://doi.org/10.1016/
R Universal Gas constant (KJ/mol.K) j.ijhydene.2021.09.183.
DHrxn Heat of reaction (J/mol) [8] Dalebrook AF, Gan W, Grasemann M, Moret S, Laurenczy G.
Hydrogen storage: beyond conventional methods. Chem
keff Effective thermal conductivity (W/mK)
Commun 2013;49:8735e51. https://doi.org/10.1039/
km Thermal conductivity of the catalyst (W/mK)
c3cc43836h.
kf Thermal conductivity of LOHC (W/mK) [9] Bourane A, Elanany M, Pham TV, Katikaneni SP. An overview
Ts Surface temperature of the catalyst (W/mK) of organic liquid phase hydrogen carriers. Int J Hydrogen
F0 Molar flow rate of dibenzyltoluene (mol/hr) Energy 2016;41:23075e91. https://doi.org/10.1016/
F Cost factor j.ijhydene.2016.07.167.
us Superficial velocity (m/s) [10] Faye O, Szpunar J, Eduok U. A critical review on the current
technologies for the generation, storage, and transportation
z Length of the reactor (m)
of hydrogen. Int J Hydrogen Energy 2022. https://doi.org/
Nu Nusselt number 10.1016/j.ijhydene.2022.02.112.
Pr Prandtl number [11] Al S, Yortanlı M, Mete E. Lithium metal hydrides (Li2CaH4 and
dt Tube diameter (m) Li2SrH4) for hydrogen storage; mechanical, electronic and
Cp Specific heat of compound (J/KgK) optical properties. Int J Hydrogen Energy 2020;45:18782e8.
h Heat transfer coefficient (W/m2K) https://doi.org/10.1016/j.ijhydene.2020.05.037.
28546 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7

[12] Chen H, Liang H, Dai W, Lu C, Ding K, Bi J, et al. ScienceDirect 3. Energy sources. Part A Recover Util Environ Eff
MgScH 15 : a highly stable cluster for hydrogen storage. Int J 2011;33:2264e71. https://doi.org/10.1080/15567036.2011.
Hydrogen Energy 2020;45:32260e8. https://doi.org/10.1016/ 585388.
j.ijhydene.2020.08.229. [27] Jang M, Jo YS, Lee WJ, Shin BS, Sohn H, Jeong H, et al. A high-
[13] Ri S Il, Hong S Il, Kim JC, Wi JH. Excellent destabilization capacity, reversible liquid organic hydrogen carrier: H 2
effect and ideal desorption temperature in the hydride -release properties and an application to a fuel cell. ACS
Mg2FeH6 substituted with yttrium; a first principles study. Sustainable Chem Eng 2019;7:1185e94. https://doi.org/
Int J Hydrogen Energy 2021;46:3319e28. https://doi.org/ 10.1021/acssuschemeng.8b04835.
10.1016/j.ijhydene.2020.11.048. [28] Jang M, Shin BS, Jo YS, Kang JW, Kwak SK, Yoon CW, et al. A
[14] Dong S, Lv E, Wang J, Li C, Ma K, Gao Z, et al. Construction of study on hydrogen uptake and release of a eutectic mixture
transition metal-decorated boron doped twin-graphene for of biphenyl and diphenyl ether. J Energy Chem 2020;42:11e6.
hydrogen storage: a theoretical prediction. Fuel https://doi.org/10.1016/j.jechem.2019.05.024.
2021;304:121351. https://doi.org/10.1016/j.fuel.2021.121351. [29] Dong Y, Yang M, Li L, Zhu T, Chen X, Cheng H. Study on
[15] Song Y. New perspectives on potential hydrogen storage reversible hydrogen uptake and release of 1,2-
materials using high pressure. Phys Chem Chem Phys dimethylindole as a new liquid organic hydrogen carrier. Int J
2013;15:14524e47. https://doi.org/10.1039/c3cp52154k. Hydrogen Energy 2019;44:4919e29. https://doi.org/10.1016/
[16] Hamayun MH, Maafa IM, Hussain M, Aslam R. Simulation j.ijhydene.2019.01.015.
study to investigate the effects of operational conditions on [30] Yang M, Cheng G, Xie D, Zhu T, Dong Y, Ke H, et al. Study of
methylcyclohexane dehydrogenation for hydrogen hydrogenation and dehydrogenation of 1-methylindole for
production. Energies 2020;13. https://doi.org/10.3390/ reversible onboard hydrogen storage application. Int J
en13010206. Hydrogen Energy 2018;43:8868e76. https://doi.org/10.1016/
[17] Teichmann D, Arlt W, Wasserscheid P. Liquid Organic j.ijhydene.2018.03.134.
Hydrogen Carriers as an efficient vector for the transport and [31] Bhamu KC, Chung JS, Kang SG. Dehydrogenation of a liquid
storage of renewable energy. Int J Hydrogen Energy organic hydrogen carrier compound 1-(3-cyclohexylpropyl)-
2012;37:18118e32. https://doi.org/10.1016/ 3-ethylcyclohexane: a density functional theory study. Int J
j.ijhydene.2012.08.066. Hydrogen Energy 2022;47:9635e42. https://doi.org/10.1016/
[18] Hamayun MH, Hussain M, Maafa IM, Aslam R. Integration of j.ijhydene.2022.01.047.
hydrogenation and dehydrogenation system for hydrogen [32] Kerscher M, Klein T, Schulz PS, Veroutis E, Dürr S, Preuster P,
storage and electricity generation e simulation study. Int J et al. Thermophysical properties of diphenylmethane and
Hydrogen Energy 2019;44:20213e22. https://doi.org/10.1016/ dicyclohexylmethane as a reference liquid organic hydrogen
j.ijhydene.2019.06.053. carrier system from experiments and molecular simulations.
[19] Müller K, Stark K, Emelyanenko VN, Varfolomeev MA, Int J Hydrogen Energy 2020;45:28903e19. https://doi.org/
Zaitsau DH, Shoifet E, et al. Liquid organic hydrogen carriers: 10.1016/j.ijhydene.2020.07.261.
thermophysical and thermochemical studies of benzyl- and [33] Mizsey P, Cuellar A, Newson E, Hottinger P, Truong TB, von
dibenzyl-toluene derivatives. Ind Eng Chem Res Roth F. Fixed bed reactor modelling and experimental data
2015;54:7967e76. https://doi.org/10.1021/acs.iecr.5b01840. for catalytic dehydrogenation in seasonal energy storage
[20] Bulgarin A, Jorschick H, Preuster P, Bo € smann A, applications. Comput Chem Eng 1999;23:S379e82. https://
Wasserscheid P. Purity of hydrogen released from the Liquid doi.org/10.1016/S0098-1354(99)80093-3.
Organic Hydrogen Carrier compound perhydro [34] Juangsa FB, Prananto LA, Mufrodi Z, Budiman A, Oda T,
dibenzyltoluene by catalytic dehydrogenation. Int J Aziz M. Highly energy-efficient combination of
Hydrogen Energy 2020;45:712e20. https://doi.org/10.1016/ dehydrogenation of methylcyclohexane and hydrogen-based
j.ijhydene.2019.10.067. power generation. Appl Energy 2018;226:31e8. https://
[21] Heublein N, Stelzner M, Sattelmayer T. Hydrogen storage doi.org/10.1016/j.apenergy.2018.05.110.
using liquid organic carriers: equilibrium simulation and [35] Wang H, Zhou X, Ouyang M. Efficiency analysis of novel
dehydrogenation reactor design. Int J Hydrogen Energy Liquid Organic Hydrogen Carrier technology and
2020;45:24902e16. https://doi.org/10.1016/ comparison with high pressure storage pathway. Int J
j.ijhydene.2020.04.274. Hydrogen Energy 2016;41:18062e71. https://doi.org/10.1016/
[22] Shi L, Qi S, Qu J, Che T, Yi C, Yang B. Integration of j.ijhydene.2016.08.003.
hydrogenation and dehydrogenation based on [36]  B, Byun M, Lim H. Design, economic evaluation,
Brigljevic
dibenzyltoluene as liquid organic hydrogen energy carrier. and market uncertainty analysis of LOHC-based, CO2 free,
Int J Hydrogen Energy 2019;44:5345e54. https://doi.org/ hydrogen delivery systems. Appl Energy 2020;274. https://
10.1016/j.ijhydene.2018.09.083. doi.org/10.1016/j.apenergy.2020.115314.
[23] Modisha PM, Ouma CNM, Garidzirai R, Wasserscheid P, [37] Peters R, Deja R, Fang Q, Nguyen VN, Preuster P, Blum L, et al.
Bessarabov D. The prospect of hydrogen storage using liquid A solid oxide fuel cell operating on liquid organic hydrogen
organic hydrogen carriers. Energy Fuel 2019;33:2778e96. carrier-based hydrogen e a kinetic model of the hydrogen
https://doi.org/10.1021/acs.energyfuels.9b00296. release unit and system performance. Int J Hydrogen Energy
[24] Acar C, Dincer I. Review and evaluation of hydrogen 2019;44:13794e806. https://doi.org/10.1016/j.ijhydene.2019.
production options for better environment. J Clean Prod 03.220.
2019;218:835e49. https://doi.org/10.1016/j.jclepro.2019.02.046. [38] Haupt A, Müller K. Integration of a LOHC storage into a heat-
[25] Usman MR, Aslam R. Dehydrogenation of controlled CHP system. Energy 2017;118:1123e30. https://
methylcyclohexane for on-board hydrogen use: initial rate doi.org/10.1016/j.energy.2016.10.129.
kinetics over 1.0 wt% Pt/g-Al2O3. Arabian J Sci Eng [39] Naseem M, Usman M, Lee S. A parametric study of
2014;39:615e20. https://doi.org/10.1007/s13369-013-0694-y. dehydrogenation of various Liquid Organic Hydrogen Carrier
[26] Usman MR, Aslam R, Alotaibi F. Hydrogen storage in a (LOHC) materials and its application to methanation process.
recyclable organic hydride: kinetic modeling of Int J Hydrogen Energy 2021;46:4100e15. https://doi.org/
methylcyclohexane dehydrogenation over 1.0 wt% Pt/-Al 2O 10.1016/j.ijhydene.2020.10.188.
 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 7 ( 2 0 2 2 ) 2 8 5 3 0 e2 8 5 4 7 28547

[40] Asif F, Hamayun MH, Hussain M, Hussain A, Maafa IM, [46] Medeiros D. Dwsim - process simulation, modeling and
Park Y. Performance analysis of the perhydro-dibenzyl- optimization technical manual. 2016.
toluene dehydrogenation system d a simulation study. 2021. [47] Poling BE, Prausnitz JM. Sciencedirect_articles_08Jan2016_02-
[41] Brückner N, Obesser K, Bo € smann A, Teichmann D, Arlt W, 19-13.101. 2011.
Dungs J, et al. Evaluation of industrially applied heat-transfer [48] Müller K, Aslam R, Fischer A, Stark K, Wasserscheid P,
fluids as liquid organic hydrogen carrier systems. ChemSusChem Arlt W. Experimental assessment of the degree of hydrogen
2014;7:229e35. https://doi.org/10.1002/cssc.201300426. loading for the dibenzyl toluene based LOHC system. Int J
[42] Bano S, Siluvai Antony P, Jangde V, Biniwale RB. Hydrogen Hydrogen Energy 2016;41:22097e103. https://doi.org/10.1016/
transportation using liquid organic hydrides: a comprehensive j.ijhydene.2016.09.196.
life cycle assessment. J Clean Prod 2018;183:988e97. https:// [49] Liu S-L. Stable explicit difference approximations to
doi.org/10.1016/j.jclepro.2018.02.213. parabolic partial differential equations. AIChE J
[43] Okada Y, Sasaki E, Watanabe E, Hyodo S, Nishijima H. 1969;15:334e8. https://doi.org/10.1002/aic.690150308.
Development of dehydrogenation catalyst for hydrogen [50] Dixon AG, Cresswell DL. Theoretical prediction of effective
generation in organic chemical hydride method. Int J heat transfer parameters in packed beds. AIChE J
Hydrogen Energy 2006;31:1348e56. https://doi.org/10.1016/ 1979;25:663e76. https://doi.org/10.1002/aic.690250413.
j.ijhydene.2005.11.014. [51] Perry S, Perry RH, Green DW, Maloney JO. Perry's chemical
[44] Shukla A, Karmakar S, Biniwale RB. Hydrogen delivery engineers' handbook, vol. 38; 2000. https://doi.org/10.5860/
through liquid organic hydrides: considerations for a choice.38-0966.
potential technology. Int J Hydrogen Energy 2012;37:3719e26. [52] Corripio AB, Chrien KS, Evans LB. Estimate costs of heat
https://doi.org/10.1016/j.ijhydene.2011.04.107. exchangers and storage tanks via correlations. Chemical
[45] Peng DY, Robinson DB. A new two-constant equation of engineering magazine. Access Intelligence; 1982.
state. Ind Eng Chem Fundam 1976;15:59e64. https://doi.org/
10.1021/i160057a011.