Chemical Papers 68 (12) 1716–1724 (2014)

DOI: 10.2478/s11696-014-0620-0

ORIGINAL PAPER

Modelling of catalytic hydrocracking and fractionation

of refinery vacuum residue

Eduard Manek, Juma Haydary*

Department of Chemical and Biochemical Engineering, Institute of Chemical and Environmental Engineering,
Faculty of Chemical and Food Technology, Slovak University of Technology in Bratislava, Radlinského 9, 812 37 Bratislava

Received 3 December 2013; Revised 12 May 2014; Accepted 6 June 2014

Dedicated to the memory of professor Elemír Kossaczký

The main objective of this work was to create a kinetic model of refinery vacuum residue hydro-
cracking and to monitor the impact of the operating conditions on the product yields. Data and
yield measurements were gathered from a residual hydrocracking unit (RHC). Reaction tempera-
ture ranged from 401 ◦C to 412 ◦C at the pressure of 18–20 MPa. A simplified kinetic yield model
was applied; where the feed and each product fraction are represented by one lump (reactant or
product of cracking) represented by the number of pseudo-components. The product fractions were
determined by fractional distillation of the output mixture from the reactor. The kinetic model
includes eight reaction steps and the following six fractions: vacuum residue, vacuum distillate, gas
oil, kerosene, naphtha, and gas. In addition, a model for hydrodesulphurisation has been proposed.
The average relative deviation between model and experimental yields was 5.36 %, and that for the
sulphur conversion model was 1.04 %. An Excel file with the kinetic model was implemented in the
Aspen Plus program using a user-defined model of the reactor. This model allows to input/output
data between the Aspen Plus and Excel programs. The Excel subroutine calculates the reaction
kinetics of cracking from the set temperature and residence time, and distributes the products into
30 pseudo-components created in the Aspen Plus program. The remaining part of the RHC unit was
simulated in the Aspen Plus environment. The effects of the reaction conditions such as temperature
and residence time on the conversion of the feed and on the distillation curves of the output mixture
from the reactor were investigated. The model was verified by comparison of the distillation curves
of simulated and real products.

c 2014 Institute of Chemistry, Slovak Academy of Sciences

Keywords: vacuum residue, catalytic hydrocracking, mathematical model, chemical kinetics

Introduction valuable products meeting actual requirements have

been gradually added. The use of the lowest boiling
The petroleum industry, like many other indus- point fractions (vacuum tower products) has always
tries, has to be dynamically adapted to actual market been complicated. Therefore, cracking units, which
demands, which is the impulse for the development increase the conversion of crude oil into lighter and
and implementation of new technologies and innova- more valuable products, play a crucial role in every
tions. Since the launch of the first oil refinery, more refinery processing heavier feedstocks. Currently, var-
than 150 years ago, the capacity and product range of ious types of commercial scale cracking units differ-
modern refineries have significantly expanded. From ing in the equipment layout, operating temperature
a simple idea to distil crude oil into a few fractions, and pressure, the use of hydrogen and catalysts, and
more ideas on how to process these fractions into more enabling the processing of different heavy fractions

*Corresponding author, e-mail: juma.haydary@stuba.sk

 E. Manek, J. Haydary/Chemical Papers 68 (12) 1716–1724 (2014) 1717

like atmospheric residue, vacuum distillate, or vac- five-lumped model with ten reaction pathways for at-
uum residue, are used. The economic result of a crack- mospheric residue hydrocracking. Experiments were
ing unit is mainly dependant on the feedstock conver- performed with a commercial size tetra lobular cata-
sion into desired products; therefore, conversion is the lyst. The model also incorporates the effectiveness fac-
most prominent indicator of the unit efficiency. Hydro- tor, and a time-dependant deactivation function for
cracking consists of two groups of processes: endother- obtaining the intrinsic kinetic parameters. Galiasso
mic hydrocarbon cracking and exothermic hydrogena- Tailleur (2005) studied mild hydrocracking catalyst
tion of unsaturated hydrocarbons, sulphur and nitro- deactivation with hydrodesulphurised vacuum gas oil
gen organo-compounds. The overall enthalpy balance as the feed and WNiPd/TiO2 Al2 O3 as the catalyst.
shows that the process is exothermic. The increasing The results confirmed that the deactivation changes
temperature leads to higher vacuum residue conver- the relative rate of the hydrogenation/hydrogenolysis,
sion into lighter components but also to operational cracking, and dealkylation reactions. Stochastic recon-
complications and unwanted effects such as coke and struction, reconstruction by enthropy maximisation
sediment generation and to suppression of the exother- and the kinetic Monte Carlo method for the two-step
mic hydrodesulphurisation and hydrodemetalisation. kinetic modelling strategy of the hydroconversion pro-
Therefore, the operation temperature must be care- cesses were used by Pereira de Oliveira et al. (2014).
fully chosen based on experience and a kinetic model This novel method was applied to the Ural vacuum
if available. residue hydroconversion and, given that this method-
Kinetic model is a useful tool for the reactor de- ology is molecule-based, a vast array of details was
sign, yield and conversion prediction as well as for unit provided by the reactor simulation. Molecular mod-
simulations. Various papers on different approaches to elling and the FTIR (Fourier transform infrared spec-
the kinetic study of heavy oil fractions hydrocracking troscopy) technique were used by Rodrigues Coelho et
have been published. Puron et al. (2014) developed a al. (2006) to determine the characteristics of aliphatic
lumped kinetic model for vacuum residue hydrocrack- chains in vacuum residues.
ing, testing two catalysts: Ni–Mo/Al2 O3 (aluminum– Most of the current papers deal with hydrocrack-
nickel–molybdenum) and Ni–Mo/Al2 O3 –Cr (alumina ing of vacuum distillates. Commercial scale resid-
doped with chromium). All proposed models consist of ual hydrocrackers processing vacuum tower bottoms
four lumps (products with the boiling point > 450 ◦C, (vacuum residues) are a relatively unique technology.
products with the boiling point < 450 ◦C, gas, and There are only a few RHC units in oil refineries all
coke) at various complexity levels. Kinetic data were over the word, mostly in the USA, Canada, and Eu-
obtained in a microbomb batch reactor. Morawski rope (Martínez et al., 2010).
and Mosio-Mosiewski (2006) observed the effect of
temperature, space velocity of feed, partial hydro- Experimental
gen pressure, and of the Ni–Mo/Al2 O3 (aluminum–
nickel–molybdenum) catalyst content during the vac- Data were collected using a commercial scale resid-
uum residue hydrocracking on the product composi- ual hydrocracker unit (RHC), which is a part of an oil
tion and quality. The experiment was run in an ex- refinery complex. A commercial catalyst based on Ni–
perimental scale continuously stirred fixed bed reac- Mo/Al2 O3 pellets was applied for this process. The
tor and the yields of the reactor output products purpose of this unit is to process vacuum residue
were determined by atmospheric and vacuum distil- from the vacuum tower bottom into lighter hydro-
lation. Liu et al. (1994) chose a dispersed catalyst carbon fractions and concurrently remove sulphur, ni-
phase and their experiment results suggest that the trogen, and oxygen from heterocyclic molecules via
dispersed phase of the Mo catalyst considerably in- hydrogenation. Metal content is removed via chemi-
hibits thepolycondensation of resins and asphaltenes, cal adsorption on a catalyst surface. Vacuum residue
and delays the coke formation by promoting hydro- is the heaviest fraction from primary crude oil dis-
genation and hydrodesulphurisation. Lababidi et al. tillation containing substances with high molecular
(2011) developed a model of the hydrocracking pro- mass (about 600 kg kmol−1 ), distillation curve is in
cess predicting the product quality of the vacuum the range of 440–750 ◦C, and density is approximately
gas oil (VGO). Model validation results showed that 1020 kg m−3 at 20 ◦C. High content of sulphur (about
the proposed model is capable of predicting distilla- 3.00 mass %), nitrogen (0.50 mass %), and metals
tion curves of the hydrocracked products. Kinetics of (vanadium: 0.021 mass %, nickel: 0.006 mass %) makes
non-catalytic hydrocracking of heavy residue oils was this fraction complicated to be processed in other
studied by Ramírez et al. (2012) using the lumping units. RHC consists of three reactor series with an
approach. The results confirmed that hydrocracking ebullated catalyst bed. The ebullated bed reactor sys-
reactions proceed by a cascade mechanism from the tem has been comprehensively described by Martínez
heavier to the lighter fractions. Naphtha does not hy- et al. (2010). Hydrocarbon cracking is an endother-
drocrack to form gases under the operating conditions mic and hydrogenation is an exothermic process, the
studied. Martínez and Ancheyta (2012) developed a overall effect is exothermic. Effluent from the first to