Fuel 214 (2018) 314–321

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Thermodynamic analysis of diesel hydrotreating reactions T

1 ⁎
Rajarshi Bandyopadhyay , Sreedevi Upadhyayula
Department of Chemical Engineering, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India

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

Keywords: Thermodynamic analysis of diesel hydrotreating reactions—Hydrodesulfurization (HDS), Hydrodenitrogenation

Thermodynamic analysis (HDN) and Hydrodearomatization (HDA)—were studied by Gibbs free energy minimization to understand the
Diesel hydrotreating effect of process variables such as temperature (200–600 °C), pressure (10–80 bar(a)) and H2/oil ratio (stoi-
Gibbs free energy minimization chiometric ratio – 5 times of stoichiometric ratio) on the product distribution. Dibenzothiophene (DBT),
Euro-V diesel
naphthalene and pyridine were considered as representative compounds to study the HDS, HDA and HDN re-
Hydrodesulfurization
actions respectively and the reactions were studied in both gas phase and two phase (gas–liquid). It was found
Hydrodenitrogenation
that the HDN reaction has no thermodynamic limitation while HDA reaction has significant thermodynamic
limitation at low pressures and high temperatures. Higher H2 concentration favors the HDA equilibria but not
very significantly. Overall HDS reaction has no thermodynamic limitation but for refractory sulfur compounds
like DBT and alkyl substituted DBTs, the preference for individual reaction pathways—Direct Desulphurization
(DDS) and Hydrogenation (HYD)—changes with change in process variables. It was shown that while the se-
lection of operating temperature and H2 concentration for a commercial diesel hydrotreating reactor are gov-
erned by the reaction kinetics, the reactor pressure is mostly governed by the HDA reaction equilibria.

1. Introduction maintaining a higher hydrogen to oil ratio.

Diesel hydrotreaters are one of the most common units in a refinery
and its importance is increasing day by day with the processing of
heavier feedstocks with higher sulfur, nitrogen and aromatic content on
the one hand and more stringent emission norms coming into force on
the other hand [1]. Main reactions that take place in diesel hydro-
treating are Hydrodesulfurization (HDS), Hydrodenitrogenation (HDN)
and Hydrodearomatization (HDA) [2]. Industrial diesel hydrotreating
reactors usually operate between 40–70 bar(g) and at 300–400 °C. For a
particular unit, the pressure slightly increases from Start of Run (SOR)
to End of Run (EOR), usually by 2–3 bar, in order to compensate for the
pressure drop in reactors because of fouling of catalyst. However, the
reactor operating temperature is slowly increased from SOR to EOR by
as much as 40–50 °C in order to compensate for the progressive deac-
tivation of the catalyst [3]. It is also a standard practice to maintain
around three times the stoichiometric requirement of H2 in the reactor
[4]. Pilot plant studies carried out at higher stoichiometric ratios sug-
gest that higher sulfur removal can be achieved by increasing the hy-
drogen to oil molar ratio up to ∼11 [5]. However, the differential in-
crease in sulfur removal at stoichiometric ratios higher than 3 is quite
marginal and has not been adopted commercially because of the sub-
stantially higher capital and operating expenses involved in
Diesel hydrotreating is a mature technology and a lot of research has
been done regarding the selection of catalyst and reactor for this pro-
cess. However, little work is available in open literature that discusses
the diesel hydrotreating reactions from a thermodynamic perspective.
This is mainly because of the lack of accurate thermochemical data like
standard heat of formation, activity coefficients etc. of the different
species that take part in these reactions [6]. Moreover, the works that
are published till date on HDS, HDN and HDA reaction equilibria per-
taining to diesel hydrotreating have reported mostly the gas-phase
equilibrium constants of these reactions at some particular temperature
and pressure conditions and thus, do not provide the holistic picture
from a thermodynamic perspective throughout the entire operating
range of a commercial diesel hydrotreating reactor. The different con-
ditions reported for reactions most relevant to diesel hydrotreating are
given in Table 1.
Here, we have used Aspen Plus (v9.0) to study HDS, HDN and HDA
reaction with representative compounds in order to gain insight into the
equilibrium conversion of these reactions around the entire range of
temperature, pressure and stoichiometric ratio used in operating an
industrial hydrotreating reactor. Temperature effects have been studied
between 200 °C and 600 °C which covers the optimum temperature
range of 300–400 °C. Similarly, pressure effects have been studied

⁎
Corresponding author.
E-mail address: sreedevi@chemical.iitd.ac.in (S. Upadhyayula).
1
Present address: Haldor Topsoe India Pvt. Ltd., Faridabad, Haryana 121003, India.

https://doi.org/10.1016/j.fuel.2017.10.015
Received 24 February 2017; Received in revised form 14 June 2017; Accepted 5 October 2017
0016-2361/ © 2017 Elsevier Ltd. All rights reserved.
R. Bandyopadhyay, S. Upadhyayula Fuel 214 (2018) 314–321

energy of the system. Thermodynamic analysis of dry reforming of CH4

Nomenclature
with CO2 has already been carried out in Aspen Plus and verified with
experimental results [16] thus signifying that this tool has adequate
xH2 mole fraction of hydrogen in liquid phase
capabilities to carry out thermodynamic analysis. We have also verified
yH2 mole fraction of hydrogen in vapor phase
the result of Aspen Plus with results obtained from first principle for a
pH2 partial pressure of hydrogen
typical hydrotreating reaction (HDA of toluene to methylcyclohexane)
log10Keq logarithm (base 10) of equilibrium constant
in order to ensure that the numerical algorithm of Aspen Plus is accu-
rate for a system like ours.
To carry out thermodynamic studies, an appropriate property
between 10 bar(a) and 80 bar(a), which also covers the optimum
method is required that can satisfactorily provide the rigorous ther-
pressure range of 40–70 bar(g). The effect of reactant concentration has
modynamic correlations. We have selected the Soave-Redlich-Kwong
also been studied by varying the ratio of H2 to representative compound
(SRK) equation of state for our purpose as it is capable of dealing with
from stoichiometric ratio to five times the stoichiometric ratio.
mixtures of hydrocarbon and hydrogen at elevated temperatures and
Although the equilibrium conversion is unique for each reaction yet,
pressures in both vapor and liquid phases with high accuracy [17–19].
earlier studies have shown that they follow similar trend for the same
The binary interaction parameters of Aspen Plus library have been used.
type of reaction [7–11]. Hence, study with representative compounds is
Experimental data of vapor liquid equilibrium is available for hy-
helpful in understanding the thermodynamic behavior of the same class
drogen/tetralin system at elevated temperatures and pressures [20].
of reaction. This study is also of particular importance from the point of
The experimentally obtained vapor and liquid mole fraction of hy-
view that the representative compounds considered are in liquid state
drogen at different temperatures and pressures were compared with
for a considerable range of the reaction conditions studied, thus, giving
those obtained from Aspen Plus using SRK equation of state and the
insight into the equilibrium conversion of two-phase (gas–liquid) sys-
values were found to match closely, with the maximum deviation being
tems, which have not been studied in detail before.
less than 5% (Table 2). This highlights the accuracy of the interaction
The equilibrium composition of a reaction system can be accurately
parameters present in Aspen Plus library. For other systems, the vali-
specified by Gibbs free energy minimization method [15]. The RGibbs
dation could not be done due to lack of published experimental data. Du
reactor model of Aspen Plus (v9.0) is based on this method and can be
et al. [21] has carried out simulation of diesel hydrotreating process
used to carry out the thermodynamic analysis of reactions. In RGibbs
with real components which included all the reactants that have been
reactor, the calculation varies the number of moles of each species in
used in this study, viz. dibenzothiophene, pyridine, naphthalene and
each phase subject to stoichiometric constraints, reaction pressure and
hydrogen. In the simulation, the thermodynamic properties were cal-
temperature to find a solution that minimizes the total Gibbs free
culated by SRK equation of state without any modification. The

Table 1
Temperature and pressure conditions at which thermodynamic equilibrium data of most relevant diesel hydrotreating reactions are reported.

Reference Reaction reported Temperature dependence reported Pressure dependence reported

HDS
Weisser and Landa [12] HDS of thiophene log10Keq values at T = 298–1000 K –
Speight [10] HDS of thiophene log10Keq values at T = 298.16 K, 500 K, 700 K, –
900 K
Devanneaux and Maurin HDS of thiophene and log10Keq values at T = 400 °C –
[13] benzothiophene

HDN
Cocchetto and Satterfield HDN of pyridine, pyrrole, log10Keq values at T = 298–1000 K Pyridine/piperidine equilibrium at T = 100–550 °C at
[8] quinoline pH2 = 11 atm, 100 atm

HDA
Wilson [14] HDA of naphthalene log10Keq values at T = 588–769 K –
Frye [7] HDA of naphthalene log10Keq values at 343 °C and 14.5 atm, 383 °C and 24.9 atm, 402 °C and 41.8 atm
HDA of phenanthrene log10Keq values at T = 588–769 K –
HDA of phenanthrene log10Keq values at 346 °C and 15.9 atm, 429 °C and 78.9 atm
Frye and Weitkamp [11] HDA of naphthalene log10Keq values at 301.5 °C and 6.1 atm, 301 °C and 6.97 atm, 351 °C and 20.35 atm, 396 °C and 10.66 atm,
396.5 °C and 47.1 atm, 457 °C and 68.8 atm

1 atm = 101325 Pa.

Table 2
Comparison of vapor–liquid equilibrium data for hydrogen/tetralin.

Temperature (°C) Pressure (atm) xH2 yH2

Aspen Plus Simnick et al. [20] % Deviation Aspen Plus Simnick et al. [20] % Deviation

268.7 20 0.0139 0.0143 -2.9 0.8121 0.8028 1.1

30 0.0212 0.0221 -4.2 0.8748 0.8676 0.8
50 0.0359 0.0373 -3.9 0.9255 0.9155 1.1
100 0.0722 0.0732 -1.4 0.9637 0.9538 1.0

348.6 50 0.0448 0.0452 -0.9 0.7107 0.7066 0.6

100 0.0935 0.0925 1.1 0.8532 0.8346 2.2

389.1 50 0.0488 0.0482 1.2 0.4905 0.4810 1.9

100 0.1161 0.1170 -0.8 0.7118 0.7008 1.5

1 atm = 101325 Pa.

315
R. Bandyopadhyay, S. Upadhyayula Fuel 214 (2018) 314–321

simulation results match closely with plant data and hence also validate 2.2. HDN reaction
the accuracy of the interaction parameters present in Aspen Plus library
corresponding to SRK equation of state. Nitrogen removal from heterocyclic organonitrogen compounds
involves hydrogenation of ring containing nitrogen atom before hy-
drogenolysis of the carbonenitrogen bond, at least with Ni-Mo/Al2O3
2. Details of reactions studied
catalyst [6]. Equilibrium constants for the gas-phase hydrogenation
and hydrogenolysis reactions of several organonitrogen compounds
2.1. HDS reaction
have been calculated by Cocchetto and Satterfield [8,9]. The data
suggests that hydrogenolysis equilibria are favorable even at tem-
The HDS reaction of organosulfur compounds is essentially irre-
peratures as high as 500 °C. Although the hydrogenation equilibria
versible under the reaction conditions applied industrially [10,22,23].
are unfavorable, higher hydrogen partial pressures which are usually
However, thermodynamics of organosulfur compounds present in
available in industrial hydrotreating reactors pushes the equilibria
higher boiling fractions, like dibenzothiophene (DBT) has not been
considerably towards the HDN products and the overall HDN reac-
studied in detail. Here we have studied the HDS reaction of di-
tion is practically irreversible under actual reactor operating condi-
benzothiophene (DBT) which has a normal boiling point of 332 °C. The
tions.
HDS of DBT takes place through two different pathways (Fig. 1) – the
As a representative HDN reaction, we have studied the HDN of
direct desulphurization route (DDS) with biphenyl (BP) as the product
pyridine to n-pentane (iso-pentane) and ammonia (reaction 3). Though
(reaction 1) and the hydrogenation (HYD) route with cyclohex-
the normal boiling point of pyridine is 115.6 °C, at a pressure of 40 bar
ylbenzene (CHB) as the product (reaction 2) [24].

(1)

(2)

The HYD route has a higher stoichiometric hydrogen consumption (a), the boiling point is ∼317 °C. So, this reaction is suitable enough for
than the DDS route. Here we have not only studied the overall HDS understanding the thermodynamic equilibria of HDN reaction in two-
reaction equilibria of DBT, but have also studied which pathway is phase too as the maximum pressure we have studied is 80 bar(a).
preferred over the other under what conditions.

Fig. 1. Reaction pathways for HDS of DBT.

S
HYD
DBT ROG
ENA
TION
ION
RIZAT
ULPHU
T DES

S
DIREC

BP CHB

316
R. Bandyopadhyay, S. Upadhyayula
2.3. HDA reaction
The studies that have been done on HDA reaction equilibria
[7,11,25,26] suggest that hydrogenation of aromatic hydrocarbons are
reversible, with the equilibrium conversion often less than 100% under
the operating conditions of industrial reactors. The equilibrium con-
versions are significantly low at high temperatures and low pressures.
However, equilibrium constants for liquid-phase HDA reactions are not
readily available and have not been studied in detail. As a re-
presentative HDA reaction, we have studied the hydrogenation of
naphthalene to tetralin (reaction 4). Naphthalene has a normal boiling
point of 218 °C and is thus present in both gas and liquid phases within
the operating range of our study.
3. Results and discussion
3.1. Model verification
In order to verify the accuracy of the numerical algorithm used in
the RGibbs reactor model of Aspen Plus, we have calculated the equi-
librium conversion of the HDA reaction of toluene to methylcyclo-
hexane (reaction 5) from first principles and cross-checked those values
with that calculated by the RGibbs reactor model of Aspen Plus.
(5)
Using the tables for the ideal gas thermodynamic functions provided
by Frenkel et al. [27], we developed a correlation (Eq. (A)) between
equilibrium constant, Keq and temperature, T in Kelvin. This equation is
valid throughout the entire temperature range of our study.
25954
ln K eq = −47.672
T (A)
Again, considering that reaction 5 takes place at atmospheric pres-
sure, the assumption of ideal gases is fully justified throughout the
entire temperature range of our study and with this assumption, Keq for
reaction 5 can be alternatively expressed by Eq. (B).
XTol (4−3XTol )3
K eq =
27(1−XTol ) 4 (B)
where, XTol is the fractional conversion of toluene at equilibrium. Keq
was calculated using Eq. (A) at intervals of 10 °C between 200 °C and
600 °C. These values were then substituted in Eq. (B) to get the equi-
librium conversion of toluene corresponding to that temperature.
Though Eq. (B) is a fourth order equation, it has only one real root less
than unity which is the only realistic solution. From Aspen Plus RGibbs
reactor model, the equilibrium conversion of toluene was calculated by
Eq. (C).
FTol,out ⎞
XTol = ⎜⎛1− ⎟
⎝ FTol,in ⎠ (C)
317
Fuel 214 (2018) 314–321
(3)
where, FTol,in and FTol,out are the molar flow rates of toluene in the input
and output streams of RGibbs reactor respectively. The RGibbs reactor
model was run at atmospheric pressure and the temperature sensitivity
analysis was carried out for 200–600 °C. The compressibility values of
hydrogen, toluene and methylcyclohexane are close to unity within this
temperature range corresponding to atmospheric pressure justifying the
fact that the reaction mixture can be considered as an ideal gas mixture.
Hence, the values obtained from first principle (Eqs. (A) and (B)) and
from Aspen Plus (Eq. (C)) are both under ideal gas conditions. As seen
in Fig. 2, the results obtained from RGibbs reactor and conventional
approach agrees very well throughout the entire range of temperature.
This comparison ensured the correctness and accuracy of RGibbs re-
actor model in Aspen Plus and was extended to study the equilibrium
(4)
conversions of other hydrotreating reactions.
3.2. Effect of temperature
The effect of temperature was studied by varying the temperature
between 200 °C–600 °C, at 55 bar (a) pressure and taking 3 times the
stoichiometric requirement of H2. 1 mol of the representative com-
pound (DBT/Pyridine/Naphthalene) was considered as the basis of the
reaction. For the HDS of DBT, 15 mol of H2 was considered which is 3
times the stoichiometric requirement of H2 if the reaction takes place
completely through the HYD route.
Within the given temperature range, the HDS of DBT is not equili-
brium limited. However, the relative preference for the two different
reaction routes changes across this temperature range. As seen from
Fig. 3, the HYD route dominates over the DDS route at lower tem-
perature and the entire reaction takes place through this route. Above
325 °C, the HYD route gets equilibrium limited and so the reaction
starts taking place through both routes. At 425 °C, both the reactions
Fig. 2. Comparison of equilibrium conversion of toluene.
R. Bandyopadhyay, S. Upadhyayula Fuel 214 (2018) 314–321

Fig. 3. Temperature effect on HDS of DBT.

Fig. 6. Effect of pressure on HDA of naphthalene.

are equally favored thermodynamically and above that the DDS route is
preferred. So, it is interesting to note that though DBT gets completely
converted to desulfurized product throughout this temperature range,
the proportion of the final products and the amount of hydrogen con-
sumed changes with the change in temperature.
HDA of naphthalene gets equilibrium limited at temperatures higher
than 350 °C. The decrease in equilibrium conversion is pretty sharp
with further increase in temperature and the fractional conversion be-
comes less than 0.1 above 480 °C. This is shown in Fig. 4(a).
As shown in Fig. 4(b), HDN of pyridine is not thermodynamically
limited in the temperature range studied.

3.3. Effect of pressure

In order to study the effect of pressure, the RGibbs reactor model

was run for a temperature range of 200–600 °C at pressures of 10 bar(a),
25 bar(a), 40 bar(a), 65 bar(a) and 80 bar(a). In line with the study of
temperature effect, 1 mol of the representative compound was con-
Fig. 4. Temperature effect on (a) HDA of naphthalene, (b) HDN of pyridine.
sidered as the basis of the reaction and 3 times the stoichiometric re-
quirement of H2 was specified in the reaction mixture.
The HDS of DBT is not thermodynamically limited in the entire
range of pressure studied. However, like the temperature effect, the

Fig. 5. Effect of pressure on HDS of DBT, (a) DDS route, (b) HYD route.

318
R. Bandyopadhyay, S. Upadhyayula Fuel 214 (2018) 314–321

Fig. 7. Effect of H2 concentration on HDS of DBT (a) DDS route, (b) HYD route.

preference for individual reaction pathways changes significantly with not so pronounced with further increase in concentration, as under-
the change in pressure. As seen in Fig. 5(a) and (b), the DDS path is standable from the graphs corresponding to H2/Naphthalene = 4 and
favored at lower pressures while the HYD path is favored at higher H2/Naphthalene = 6.
pressures. The HDN of pyridine is not thermodynamically affected and 100%
The thermodynamic limitation of HDA of naphthalene is very pro- equilibrium conversion is achieved at any H2 concentration above the
nounced at low pressures and reduces with increase in pressure as can stoichiometric requirement.
be seen from Fig. 6. At 400 °C, the fractional conversion of naphthalene
is only 0.18 at 10 bar(a), while it increases to 0.77 at 40 bar(a) and 3.5. Discussion on hydrotreating reactor operating conditions from
becomes as high as 0.93 at 80 bar(a). thermodynamic perspective
The HDN of pyridine undergoes complete conversion in the entire
range of pressure studied. The effects of temperature, pressure and H2 concentration on HDS,
HDA and HDN reactions with representative compounds can be gen-
eralized to understand the significance of the operating parameters
3.4. Effect of H2 concentration
selected for a commercial diesel hydrotreating reactor depending on the
type of feed and the product specifications required. For example, the
In order to understand the effect of hydrogen on the thermodynamic
main specifications for Euro-V diesel are [28]:
conversion, the RGibbs reactor model was run at 55 bar(a) for a tem-
perature range of 200–600 °C with H2 in the reactant mixture varying
Maximum sulfur content: 10 ppm
from stoichiometric ratio to 5 times the stoichiometric ratio. However,
Minimum cetane number: 51
changes in thermodynamic conversion are marginal at stoichiometric
Maximum density @ 15 °C: 845 kg/m3
ratios higher than 3 and have not been reported.
Among the two reaction pathways of HDS of DBT, the DDS path is
While the sulfur content in the diesel product is governed by the
favored at lower H2 concentration while the HYD path is favored at
extent of HDS reaction, the cetane number and density reduction lar-
higher H2 concentration. This is represented in Fig. 7(a) and (b) which
gely depend on the extent of HDA reactions [2].
plot the moles of BP and CHB respectively at equilibrium at different H2
Though finished diesel product does not have any specification for
concentration. At H2/DBT = 5, there is slight conversion through DDS
organic nitrogen, still the HDN reactions are important as presence of
pathway even at 200 °C. However, at H2/DBT = 10, the conversion
through DDS pathway starts only after 340 °C. This finding is in line
with the fact that the stoichiometric requirement of H2 for DDS route is
lower than that through HYD route. Another interesting observation is
that the effect of H2 concentration on reaction pathway decreases with
the increase in H2 concentration. The relative change between the DDS
and HYD pathways when the H2/DBT ratio is raised from 5 to 10 is
much higher than when it is increased from 10 to 15. Also, as seen in
the studies of temperature and pressure effect, the H2 concentration
does not affect the HDS of DBT as a whole and 100% equilibrium
conversion is achieved at any H2 concentration above the stoichio-
metric requirement.
As seen from the studies of temperature and pressure effect, the
HDA of naphthalene is thermodynamically limited in the operating
temperature and pressure range of commercial hydrotreating reactors.
This limitation is more pronounced at lower H2 concentration. For ex-
ample, as seen from Fig. 8, at 350 °C, the equilibrium conversion of
naphthalene is around 90% and 97% corresponding to H2/Naptha-
Fig. 8. Effect of H2 concentration on HDA of naphthalene.
lene = 2 and H2/Naphthalene = 4 respectively. However, the effect is

319
R. Bandyopadhyay, S. Upadhyayula
organic nitrogen compounds inhibit the HDS reactions [29]. As seen
from the studies in the previous sections, the overall HDS reaction and
HDN reaction is not equilibrium limited in the range of operating
conditions of a commercial hydrotreating reactor. The HDA reaction is,
however, equilibrium limited at lower pressures and higher tempera-
tures. From this study, we can say that temperatures as low as even
200 °C are suitable for carrying out hydrotreating reactions. Low tem-
peratures are, in fact, good for HDA reactions as up to 250 °C, the HDA
reactions are not equilibrium limited even at pressures as low as 10 bar
(a). So, it can be concluded that the temperature of 300 °C, which is the
typical SOR reactor temperature for most diesel hydrotreater units is a
value arrived upon from the kinetic point of view and not from ther-
modynamic requirement.
This temperature value obviously has a trade-off with the equili-
brium conversion of HDA reaction which is compensated by increasing
the reactor operating pressure. For hydrotreating units running with
straight run feed, the requirement for density reduction and cetane
number improvement is usually not very high. Such units can run at
relatively low pressures of 40–45 bar(a) and can still achieve the de-
sired HDA conversion, which is usually low. However, for units running
with significant amount of cracked feed like coker gas oil, the aromatic
content in the feed is quite high and consequently a large density re-
duction and cetane number improvement becomes necessary. In such
cases, it is not uncommon to operate the reactor at a pressure as high as
70 bar(a) in order to nullify any equilibrium limitation of HDA reac-
tions.
As seen in this study, the reactor operating conditions also de-
termines which HDS pathway (DDS or HYD) is preferred over the other.
This has direct impact on the selection of catalyst as studies have shown
that the Co-Mo catalysts desulfurize primarily by the DDS route while
Ni-Mo catalysts have a higher selectivity for desulfurization via the
HYD route [24]. As H2 consumption in HYD route is higher than that in
DDS route, proper selection of reactor operating parameters in con-
junction with catalyst selection is not only important for the optimal
reactor performance but also determines the chemical H2 consumption
of the plant. Hydrogen, being a very costly utility, significantly affects
the plant’s operating expenses and should be evaluated with utmost
care.
It is also seen that H2 availability up to twice the stoichiometric
requirement has some positive effect on the equilibrium conversion of
HDA reaction, but not very significant and it has no influence on the
equilibrium conversion of HDS and HDN reaction. So, it can be con-
cluded that the standard practice of maintaining three times the stoi-
chiometric requirement of H2 in hydrotreating reactors is mostly a ki-
netic requirement. This is also substantiated by the presence of partial
pressure of H2 as one of the parameters in the different rate expressions
for HDS [30][31], HDA [32]and HDN [33] reactions published in lit-
erature.
Moreover, it was observed that HDN reaction is not at all equili-
brium limited in the entire range of temperature, pressure and H2
concentration that we have studied. However, slip of organic nitrogen
compounds, especially carbazole, is a common phenomenon in hydro-
treating reactors [29]. From previous works [8,9], we can say that the
HDN equilibrium of carbazole will follow a similar trend as the HDN
equilibrium of pyridine. Thus, from our study, it can also be concluded
that the presence of carbazole in the diesel product after hydrotreat-
ment is purely because of kinetic limitation and not equilibrium lim-
itation.
4. Conclusion
Thermodynamic analysis of three major types of diesel hydro-
treating reactions – HDS, HDN and HDA – were carried out with re-
presentative compounds in the temperature, pressure and H2 con-
centration ranges of an operating commercial diesel hydrotreating
reactor. The simulations involve complete gas phase as well as two-
320
Fuel 214 (2018) 314–321
phase (gas–liquid) systems. The RGibbs reactor model of Aspen Plus
(v9.0) that works on the Gibbs free energy minimization technique was
used in the simulations. A numerical algorithm was used by RGibbs
reactor model and the Aspen Plus simulation results were validated by
calculating the equilibrium conversion of HDA of toluene to methyl-
cyclohexane from first principles. By extending the results obtained for
the representative compounds for all the reactions of the same class, the
following conclusions can be drawn for the diesel hydrotreating reac-
tions, in general, in the temperature, pressure and H2 concentration
ranges studied:
• Overall, HDS reaction is not equilibrium limited but the preference
for individual reaction pathways (HYD and DDS) for refractory
sulfur compounds like dibenzothiophenes and alkyl-substituted di-
benzothiophenes changes with change in operating conditions. High
temperature, low pressure and low H2 concentration favors the DDS
route while the opposite conditions favor the HYD route.
• The HDA reactions are equilibrium limited at low pressure and high
temperatures. H2 concentration higher than stoichiometric require-
ment favors the HDA equilibria but is not very significant.
• HDN reactions are not equilibrium limited.
Based on the above-mentioned conclusions, the following conclu-
sions can also be drawn about the standard operating conditions of a
commercial diesel hydrotreating reactor:
• The reactor SOR temperature is a value that is ascertained by re-
action kinetics and not by thermodynamics.
• The H consumption in a diesel hydrotreating reactor partially de-
2
pends on the HDS reaction pathway of refractory sulfur compounds.
Thus, the H2 consumption can be altered by altering the reactor
operating parameters.
• Reactor operating pressure is usually governed by the equilibria of
HDA reactions.
• H2 concentration higher than stoichiometric requirement is pro-
vided mainly to favor reaction kinetics and positively affects only
the HDA reaction equilibria to a small extent.
• The presence of organic nitrogen compounds at the product of diesel
hydrotreating reactors is solely because of kinetic limitation.
Funding
This research did not receive any specific grant from funding
agencies in public, commercial or not-for-profit sectors.
References
[1] Stanislaus A, Marafi A, Rana MS. Recent advances in the science and technology of
ultra low sulfur diesel (ULSD) production. Catal Today 2010;153:1–68.
[2] Topsoe H, Clausen BH. Anderson JR, Boudart M, editors. Hydrotreating catalysis –
science and technology, vol. 11. Berlin: Springer Verlag; 1996.
[3] Harwell L, Thakkar S, Polcar S, Palmer RE, Desai PH. ULSD hydrotreating – con-
clusion: study identifies optimum operating conditions for ULSD hydrotreater de-
sign. Oil Gas J 2003;30:58–62.
[4] Meyers RA. Handbook of petroleum refining processes. 4th ed. New York: McGraw-
Hill; 2016.
[5] Hoekstra G. The effects of gas-to-oil rate in ultra low sulfur diesel hydrotreating.
Catal Today 2007;127:99–102.
[6] Girgis MJ, Gates BC. Reactivities, reaction network, and kinetics in high-pressure
catalytic hydroprocessing. Ind Eng Chem Res 1991;30:2021–58.
[7] Frye CG. Equilibrium hydrogenation of polycyclic aromatics. J Chem Eng Data
1962;7:592–5.
[8] Cocchetto JF, Satterfield CN. Thermodynamic equilibria of selected heterocyclic
nitrogen compounds with their hydrogenated derivatives. Ind Eng Chem Process
Des Dev 1976;15:272–7.
[9] Cocchetto JF, Satterfield CN. Chemical equilibria among quinoline and its reaction
products in hydrodenitrogenation. Ind Eng Chem Process Des Dev 1981;20:49–53.
[10] Speight JG. The desulfurization of heavy oils and residua. New York: Marcel
Dekker; 1981.
[11] Frye CG, Weitkamp A. Equilibrium hydrogenation of multi-ring aromatics. J Chem
Eng Data 1969;14:372–6.
R. Bandyopadhyay, S. Upadhyayula
[12] Weisser O, Landa S. Sulphur catalysts, their properties and applications. Oxford:
Pergamon Press; 1973.
[13] Devanneaux J, Maurin J. Hydrogenolysis and hydrogenation of thiophenic com-
pounds on a Co-Mo/Al2O3 catalyst. J Catal 1981;205:202–5.
[14] Wilson TP, Caflisch EG, Hurley GF. The naphthalene–tetralin–hydrogen equilibrium
at elevated temperature and pressure. J Phys Chem 1958;62:1059–61.
[15] Smith JM, van Ness HC, Abbott MM. Introduction to chemical engineering ther-
modynamics. 7th ed. New York: McGraw-Hill; 2004.
[16] Chein RY, Chen YC, Yu CT, Chung JNG. Thermodynamic analysis of dry reforming
of CH4 with CO2 at high pressures. J Nat Gas Sci Eng 2015;26:617–9.
[17] Soave G. Equilibrium constants from a modified Redlich-Kwong equation of state.
Chem Eng Sci 1972;27:1197–203.
[18] Valderrama JO, Cisternas LA, Vergara ME, Bosse MA. Binary interaction parameters
in cubic equations of state for hydrogen hydrocarbon mixtures. Chem Eng Sci
1990;45:49–54.
[19] Park J, Robinson RLJ, Gasem KAM. Solubilities of hydrogen in heavy normal par-
affins at temperatures from 323.2 to 423.2 K and pressures to 17.4 MPa. J Chem Eng
Data 1995;40:241–4.
[20] Simnick JJ, Lawson CC, Lin HM, Chaco KC. Vapor-liquid equilibrium of hydrogen
tetralin system at elevated temperatures and pressures. AIChE J 1977;23:469–76.
[21] Du Z, Li C, Sun W, Wang J. A simulation of diesel hydrotreating process with real
component method. CJCHE 2015;23:780–8.
[22] Vrinat ML. The kinetics of the hydrodesulfurization process: a review. Appl Catal
1983;6:137–58.
[23] Gates BC, Katzer JR, Schuit GCA. Chemistry of catalytic processes. New York:
321
Fuel 214 (2018) 314–321
McGraw-Hill; 1979.
[24] Knudsen KG, Cooper BH, Topsøe H. Catalyst and process technologies for ultra low
sulfur diesel. Appl Catal A Gen 1999;189:205–15.
[25] Poling BE, Prausnitz JM, O’Connell JP. The properties of gases and liquids. 4th ed.
New York: McGraw-Hill; 2001.
[26] Stull DR, Westrum Jr EF, Sinke GC. The chemical thermodynamics of organic
compounds. New York: Wiley; 1969.
[27] Frenkel M, Kabo GJ, Marsh KN, Roganov GN, Wilhoit RC. Thermodynmics of or-
ganic compounds in the gas state. vols. 1 and 2. Florida: CRC Press; 1994.
[28] Union European. Directive 2009/30/EC of the European Parliament and of the
Council. Off J Eur Union 2009;L140:88–113.
[29] Zeuthen P, Knudsen KG, Whitehurst DD. Organic nitrogen compounds in gas oil
blends, their hydrotreated products and the importance to hydrotreatment. Catal
Today 2001;65:307–14.
[30] Van Parjis IA, Froment GF. Kinetics of hydrodesulfurization on a Co-Mo/γ-Al2O3
catalyst. 1. Kinetics of hydrogenolysis of thiophene. Ind Eng Chem Prod Res Dev
1986;25:431–6.
[31] Broderick DH, Gates GC. Hydrogenolysis and hydrogenation of dibenzothiophene
catalyzed by sulfided CoO-MoO3/γ-Al2O3: the reaction kinetics. AIChE J
1981;27:663–73.
[32] Sapre AV, Gates BC. Hydrogenation of biphenyl catalyzed by sulfided CoO-MoO3/γ-
Al2O3. Ind Eng Chem Process Des Dev 1982;21:86–94.
[33] Shih SS, Mathur KN, Katzer JR, Kwart H, Stiles AB. Quinoline hydrodenitrogena-
tion: reaction network and kinetics. Prepr Am Chem Soc Div Petrol Chem
1977;22:919–40.