Journal of Petroleum Science Research (JPSR)

Two Dimensional Dynamic Modeling of

Hydrodesulphurization Reactor
Peyman Mahinsa1,2, Mohammad Taghi Sadeghi2, Hamid Ganji1*, Saeed Shokri1
1

Process Development Division, Research Institute of Petroleum Industry, Tehran, IRAN

Faculty of Chemical Engineering, Iran University of Science and Technology, Tehran, IRAN
ganjih@ripi.ir

Abstract- In this research a pseudo-homogeneous twodimensional model was proposed to describe the dynamic
behavior of a fixed-bed pilot-plant hydrodesulphurization
reactor. The catalyst pellet used in this reactor was Co-Mo/Al2O3.
At first, using the experimental data, a power law kinetic model
was developed for hydrodesulphurization reaction. Then a
pseudo-homogeneous two-dimensional dynamic model was
proposed to describe the concentration profile in the reactor bed.
The simulation obtained with the proposed dynamic model
showed good agreement with experimental data and the sulfur
concentration error in the reactor outlet was 3.8 percent
compared to the experimental data. Two dimension modeling
revealed that the radial variation of sulfur concentration is more
in the reactor inlet than the outlet, but in general the
concentration profile can be considered in one dimension.
Unsteady reactor modeling showed that the transition time was
higher in the reactor outlet and estimated to be 11667 seconds.
Keywords- Hydrotreating reactor, Dynamic Modeling, Two
Dimension

I. INTRODUCTION
Hydrodesulphurization (HDS) is an important process in
oil industry. The HDS process is essential to obtain fuels with
improved quality and low polluting compounds and it is
usually conducted in a fixed-bed catalytic reactor either in
single gas-phase flow or two phases (gas and liquid) [1-3] and
usually a trickle-flow regime may occur. In the trickle flow
regime, the liquid reactant flows downward through the
reactor in the form of thin laminar film droplets around the
solid catalyst [4-6]. Modeling and simulation are commonly
applied in the design, performance analysis, optimization, and
scale-up of HDS reactors. More papers in modeling and
simulation are in steady state but reliable three-phase reactor
modeling and simulation should be based on dynamic
heterogeneous models, which can be used not only for scaleup, start-up and operability studies, but also to obtain a
meaningful continuity path to the steady state of the reactor,
since dynamic models provide a realistic description of the
transient states of three-phase reactors [7, 8]. The study of the
dynamic behavior of the three phase reactor also helps
designing the best control system in order to obtain a safe,
efficient and profitable operation. Although the dynamic
models are more complicated to formulate and to solve, they
should be preferred over steady-state models because the
numerical solution strategy of dynamic models is more robust
than the solution of steady-state models [9-12].
Numerous papers have been published on steady state
modeling of hydrotreating reactors. However, studies on
dynamic modeling of such reactors are reported less in the
open literature among which the two dimensional models are
rarely reported. Julcour et. al.[13] investigated the dynamic of
the three phase up-flow fixed bed reactor using a nonisothermal heterogeneous model. They compared a simplified

model with an extended one and concluded that the diffusion

of hydrocarbons is not limiting agent so that the simplified
model predicts accurately the reactor transient behavior.
Hastaoglue and Jibril[12] modeled the transient gas-solid
reactions in a fixed-bed reactor and applied to HDS reaction
with the Langmuir-Hinshelwood mechanism. They validated
the model through a comparison of experimental data from
naphtha HDS pilot plant. Mederos et. al[14] investigated a
dynamic heterogeneous one-dimensional model of trickle-bed
hydrotreating
reactor.
They
considered
the
hydrodesulphurization,
hydrodenitrogenation
and
hydrodearomatization reactions and axial changes in
concentration, partial pressure and temperature profiles were
obtained with time. Chen and Ring[15] considered a pseudo
homogeneous two dimensional reactor model of a fixed-bed
hydrotreater and studied concentration and temperature
profiles. Their model considered the heat conduction in the
thermowell to predict the temperature difference between the
thermowell and the catalytic bed. They concluded that if this
difference is too high and ignored, could cause errors in the
interpretation of pilot plant data. Mederos and Ancheyta [2]
developed a dynamic heterogeneous one-dimensional model
to predict the behavior of trickle-bed hydrotreating reactor
with co-current and counter-current operation. They
concluded that counter-current mode can have great potential
to be used for deep hydrodesulfurization of oil fractions since
it minimizes the inhibiting effect of some products in reactor
zones where these species tend to concentrate in concurrent
operation.
This paper reports experimental work and simulation
results on the steady-state and dynamic behavior of a fixedbed pilot-plant HDS reactor.
II. EXPERIMENTAL SECTION
The experiments were done in an isothermal pilot-plant
reactor. The schematic diagram of the pilot plant setup is
shown in fig. 1.

Fig 1. Schematic diagram of the pilot plant setup

JPSR Vol. 1, Iss. 2, Jul. 2012 PP. 32-35 http://www.jpsr.org American V-King Scientific Publishing
- 32 -

Journal of Petroleum Science Research (JPSR)

The unit has been operated at temperature range of 340380 C and pressure range of 48-60 bar. Feed volumetric flow
rate was 60-260
. The reactor length and internal diameter
were 200 and 1.9 cm respectively and the catalytic bed length
catalyst.
was 35 cm which was filled with CoMo/
Above and below catalyst bed were packed with glass beads
to provide a uniform gas and liquid flow.
Hydrodesulfurization unit needs two different kinds of feeds.
One for activation that consists of Dimethyl disulfide,
hydrogen sulfide and carbon disulfide and other is used for
separation of sulfur.
III. MATHEMATICAL MODEL
The reactor which has been used in pilot plant was
operated isothermally. In this research we considered a
pseudo-homogeneous two-dimensional model with axial
convective and radial dispersion of mass. The dynamic mass
balance equation in the catalyst bed is
C

(1)

a type of weight residual method (WRM) for elliptic PDE

with two spatial domains. The orthogonal collocation is
applied on both domains to yield a set of algebraic equations
[16-18].
IV. RESULTS
A. Kinetics Evaluation
The following power law rate equations were considered
for sulfur conversion although other expressions are available:
HDS

(2)

Where i=S, H2S, H2

1.
2.

Isothermal reactor
Plug flow regime

Equation 4 can be written as:

C .T

(3)

.T

It was assumed that all holes of catalyst pellet were filled

with liquid.
A. Model initial and boundary conditions and solution
For steady state and dynamic modeling it is necessary to
define initial and boundary conditions. The initial and
boundary conditions for liquid and solid phases are
summarized in tables 1 and 2 respectively.
TABLE 1 INITIAL CONDITIONS

Condition
z=0
0rR
0<z<
0rR
z= B
0rR

Liquid phase

Solid phase

64402.1 exp

CS .

(7)

B. Steady State Results

The reactor governing equations have to be solved
simultaneously to obtain sulfur axial concentration profile. Fig.
2 shows how sulfur concentration decreases in the liquid
phase through the reactor.
120

100

80

60

40

20

0.05

0.1

0.15

0.2

Length of reactor (m)

0.25

0.3

0.35

Fig 2. Sulfur axial concentration profile in liquid phase (feed rate: 140 cc/hr ,
P=50 bars inlet temperature=632.15 K)

Liquid phase

Solid phase

z=0
0rR
z= B
0rR
r=0
0<z< B
r=R
0<z< B

HDS

0
0

TABLE 2 BOUNDRY CONDITIONS

condition

(6)

LHSV

Kinetics parameters were obtained using Eq. 6 and fitting

experimental data where the following kinetics equation was
obtained with linear regression of 98.7 percent:

Concentration of sulfur (Mol/m3)

Were Cf and CP are sulfur concentration at reactor inlet

and outlet respectively.

Using Eq. 2, the effectiveness factor can be calculated with

the following equation:

(5)

In which three kinetics parameters ( , E, n) were

unknown and to calculate them two basic assumptions were
considered:

(4)

exp

Where i=S, N, A
For catalyst pellets:

kCS

The numerical method chosen to solve the developed

partial differential equations is orthogonal collocation method,

The results reveal that there exists good agreement with

experimental data and the sulfur concentration error in the
reactor outlet was 3.8 percent compared to the experimental
data and conversion was 99 percent. The partial pressure of
H2S in gas phase increases smoothly along the reactor bed as
shown in Fig 3. But the H2S concentration at the liquid phase
increases and then decreases (Fig. 4). This phenomenon can
be interpreted by mass transfer resistances. When the H2S
concentration in the liquid phase increases, the driving force
of mass transfer from liquid to gas phase increases
accordingly. This decreases the H2S concentration in liquid
phase.

JPSR Vol. 1, Iss. 2, Jul. 2012 PP. 32-35 http://www.jpsr.org American V-King Scientific Publishing
- 33 -

Journal of Petroleum Science Research (JPSR)

-7

x 10

Sulfur Concentration (mol/m3)

Partial Pressure H2S (MPa)

3.5
3
2.5
2
1.5
1
0.5

0
0

0.05

0.1

0.15

0.2

Length of reactor (m)

0.25

0.3

0.35

1.12
1.1
1.08
0

0.2

80

0.8

C. Dynamic Simulation
Figure 6 shows the variation of sulfur concentration at
initial part of the catalytic bed (8.75 cm) and also at the
reactor outlet with time. It was observed that the transition
time was higher in the reactor outlet and estimated to be
11667 seconds

60
50
40
30
20
10
0
0

0.05

0.1

0.15

0.2

Length of reactor (m)

0.25

0.3

0.35

0.16

17.4

0.14

0.1

0.06
0.04
0.02

0.2

0.4

0.6

0.8

0.4

0.6

0.8

Dimensionless Radius (r/R)

A

8.2
8.15
8.1
8.05
0.2

Dimensionless Radius (r/R)

B

6.2
6.15

20

30

40

50

60

Dimensionless Time

70

80

90

100

0.08)

V. CONCLUSIONS
A mathematical reactor model has been developed to
numerically simulate the steady-state and dynamic behavior of
a pilot-plant hydrotreater. The pilot plant reactor has been
operated isothermally. Simulation results revealed suitable
agreement with the pilot-plant experimental data. Simulation
results indicated that the axial sulfur concentration decreases
in the liquid phase while the hydrogen sulfide partial pressure
increases smoothly in the gas phase. Hydrogen sulfide
concentration in the liquid phase increases first and then
decreases. The reason of this behavior can be interpreted by
mass transfer driving force from liquid to gas phase which
increases when the H2S Concentration in liquid phase
increases. Two dimension modeling result indicated that there
is not much radial concentration variation and one
dimensional approach can be considered. Dynamic simulation
results proved that the transition time to steady state increases
when reach to the reactor outlet.
Nomenclature

6.1

= Molar concentration of component i in the liquid phase,

6.05
6
0

10

time is

17.1

8
0

reactor outlet

0.08

Fig 6. Sulfur concentration at 8.75 cm and the reactor outlet as a function of

time (Feed rate: 140 cc/hr, P=50 bars inlet temperature=632.15 K, Dimension

17.2

17
0

Z=8.75 CM

0.12

0
0

17.3

Sulfur Concentration (mol/m3)

0.6

70

Radial sulfur concentration profiles were obtained at

various axial sections of the reactor (Fig.5). Note that the
radial variation of sulfur concentration decrease when
approach the end of reactor.

Sulfur Concentration (mol/m3)

0.4

Dimensionless Radius (r/R)

D

Fig 5. Sulfur radial concentration profile in liquid phase at various bed

sections (feed rate: 140 cc/hr , P=50 bars, inlet temperature=632.15 K ,
A=0.25 Z, B=0.5 Z, C=0.75 Z, D=Z)

Fig 4. H2S concentration profile in liquid phase

Sulfur Cincentration(mol/m3)

1.14

Sulfur Concentration (Mol/m3)

Concentration of H2S (Mol/m3)

Fig 3. Partial pressure of H2S in gas phase

1.16

0.2

0.4

0.6

Dimensionless Radius (r/R)

C

0.8

= Molar concentration of component i inside the solid

filled with liquid phase,

= Catalyst particle diameter, m

JPSR Vol. 1, Iss. 2, Jul. 2012 PP. 32-35 http://www.jpsr.org American V-King Scientific Publishing
- 34 -

Journal of Petroleum Science Research (JPSR)

,

= Effective diffusivity of component i inside a porous

catalyst,

= Molecular hydrogen

= Mass axial dispersion coefficient of liquid phase,

= Mass radial dispersion coefficient of liquid phase,

LHSV= Liquid hourly space velocity, (

n= Reaction Order
P= Reactor Pressure, (MPa)
r= Radial reactor coordinate, (m)
= Radius of particle, (m)
=Rate of reaction j per unit of catalyst mass in the liquid

phase,

t= Time, s

= Dimensionless Time

z= Axial reactor coordinate, m

Greek Letters
= Catalyst bed void fraction or catalyst bed porosity
= Catalyst particle porosity
= Catalyst bulk (or bed) density,
= Density at process conditions of f phase,
= Catalyst effectiveness factor of reaction j in the liquid
phase
Subscripts
App= Apparent

= Hydrogen sulfide
L= Liquid phase
S= solid phase, inside catalyst pellet

= Apparent reaction rate constant

f= Phase (gas, liquid or solid)

REFERENCES
[1] Vishwakarma, S. K.(2007) Sonochemical and Impregnated Co-W/Al O Catalysts:Performances and Kinetic Studies on Hydrotreatment of
2 3 Gas Oil.
Light
[2] Mederos, F.S., Ancheyta, J. (2007) Mathematical modeling and
simulation of hydrotreating reactors: Cocurrent versus countercurrent
operationsApplied Catalysis A: General 332 (1) ,821.
[3] Ramachandran, P. A., Chaudhari, R. V. Three pPhase Catalytic Reactors;
Gordon and Breach Science Publishers Inc.: London, 1983.
[4] Doraiswamy, L.K., Sharma, M.M. (1984) Heterogeneous Reactions:
Analysis, Examples, and Reactor Design; John Wiley & Sons: New York,
1984; Vol. 2, 374 pp.
[5] Bhaskar, M., Valavarasu, G. (2002) Meenakshisundaram, A., Balaraman,
K. S. Application of a Three Phase Heterogeneous Model to Analyse the
Performance of a Pilot Plant Trickle Bed Reactor. Pet. Sci. Technol., 20
(3&4): 251268.
[6] Froment, G.F., Depauw, G.A., Vanrysselberghe, V. (1994) Kinetic
Modeling and Reactor Simulation in Hydrodesulfurization of Oil
Fractions. Ind. Eng. Chem. Res., 33 (12): 29752988.
[7] Fabin S. Mederosa Ignacio Elizalde, Jorge Ancheyta, Steady-State and
Dynamic Reactor odels for Hydrotreatment of Oil Fractions: A Review,
Mxico, 2009.
[8] Tarhan, M.O. (1983) Catalytic Reactor Design. McGraw-Hill: New York;
372 pp.
[9] Chen, J., Ring, Z., Dabros, T. (2001) Modeling and Simulation of a
Fixed-Bed Pilot-Plant Hydrotreater. Ind. Eng. Chem. Res., 40 (15):
32943300.
[10] Warna , J., Salmi, T. (1996) Dynamic Modelling of Catalytic Three
Phase Reactors. Comput. Chem. Eng., 20 (1): 3947.
[11] Salmi, T., Warna , J., Toppinen, S., Ronnholm, M., Mikkola, J.P.
(2000) Dynamic
Modelling of Catalytic Three-Phase Reactors for
Hydrogenation and Oxidation Processes. Braz. J. Chem. Eng., 17 (47):
10231035.
[12] Hastaoglu, M.A., Jibril, B.E. (2003) Transient Modeling of
Hydrodesulfurization in a Fixed- Bed Reactor. Chem. Eng. Comm., 190
(2): 151170.
[13] Julcour C., Chaudhari R. V., Lelann J M., Wilhelm A. M., Delmas H. ,
Dynamic modeling of three phase upflow fixed-bed reactor including
pore diffusion, Chem. Eng. And Proc., 41(2002)311-320.
[14] Mederos F. S., Rodriguoz J., Ancheyta J., and Arce E., Dynamic
modeling and simulation of catalytic hydrotreating reactor, Energy &
Fuels, 20(2006)936-945.
[15] Chen, J., Ring, Z., Dabros, T. (2001) Modeling and Simulation of a
Fixed-Bed Pilot-Plant Hydrotreater. Ind. Eng. Chem. Res., 40 (15):
32943300.
[16] Villadsen, J., Michelsen, M.L., (1978) Solutions of Differential Equations
Models by Polynomial Approximation. Prentice-Hall, Inc., New Jersey.
[17] Hilfebrand, F., (1965) Method of Applied Mathematics, second ed.,
Prentice-Hall, Inc., New Jersey.
[18] Mickley, H.S., Sherwood, T.K., Reed, C.E. (1990) Applied Mathematics
in Chemical Engineering, second ed., McGraw-Hill, Delhi.

B= Referred to reactor catalytic bed

JPSR Vol. 1, Iss. 2, Jul. 2012 PP. 32-35 http://www.jpsr.org American V-King Scientific Publishing
- 35 -