Numerical
Modelling of the FCC Regenerator Reactor Based on Shrinkage Reaction Rate Model

Salar Azizi, Markus Schubert

Helmholtz-Zentrum Dresden-Rossendorf, Institute of Fluid
Dynamics, Bautzner Landstraße 400, 01328 Dresden. Germany; tel. +49 351 260 3765, fax: +49 351 260 2383, e-mail: s.azizi@hzdr.de.

Introduction

Fluid catalytic
cracking (FCC) reactors are applied to convert gas oils and residues to
lighter, higher-value products. Circulating fluidized bed technologies are used
in modern FCC units, where cracking reactions take place in the riser with co-current
upflow of the vaporized gasoil feed and the catalyst. After the disengagement
from the product gas, the catalyst needs to be regenerated to compensate the
deactivation due to coke deposition. The deactivation of the FCC catalyst can occur
already in a few seconds and regeneration of the deactivated catalyst plays an
important role for the yield of the FCC unit (Han, 2004). For an effective
regeneration, burning of deposited coke in the fluidized bed solves the early decay
time of the catalysts. The burning of the coke has two essential functions: a)
the regeneration of the catalyst, restoring its catalytic activity and, b) the
increase of the catalyst temperature to provide energy to the cracking, when it
returns to the riser (Penteado
et al., 2003). During the process the heavy hydrocarbons deposit on the
catalyst, which subsequently block the active sites. To restore the activity,
deposited hydrocarbons are oxidized with air in the regenerator reactor. The
performance of the regenerator as well as its coupling with the cracking
reactor are important to reach a high overall yield in the FCC unit. Usually,
Geldart-A particles are used as catalyst in the FCC reactors. As the aeration
rate of the FCC catalyst is low and the particles are cohesive, big bubbles are
formed, channeling occurs and fluidization is often nonhomogeneous, which makes
the hydrodynamic modeling a challenge. The clustering behavior of the
fluidizing gas also increases the complexity of the reactor design. The
complicated hydrodynamic behavior of gas phase and solid particles is a critical
point to model coupled heat and mass transfer phenomena inside the fluidized
reactors.

Aims

The aim of the
work is to develop an Eulerian-Eulerian numerical model for the FCC regenerator
reactor operated as a fluidized bed based on the kinetic theory of granular
flows to consider the clustering effect of the FCC catalysts in the
regeneration process. In addition, the model shall consider the impact of the shrinkage
of the deposited coke on FCC catalyst regeneration time, product yield, and
temperature history of the regenerator reactor.

Modeling
approach

The shrinking
core model is used to predict the regeneration reaction rate of the catalyst, in
which deposited coke is being consumed by the oxidation reaction. As a result,
the deposited hydrocarbons are consumed and the deactivated catalyst core is
shrinking. Figure 1 shows schematically the progress of the FCC catalyst
regeneration, i.e. the removal of the coke from a single porous catalyst pellet
as a function of time.

Figure 1:
Consumption and shrinking of deposited coke with the diffusion of oxygen
through a catalyst particle.

The
Eulerian-Eulerian two-phase flow approach is applied to a batch fluidized bed
reactor for the modeling of the gas-solid flow hydrodynamic. Here, the two phases are mathematically treated as
interpenetrating continua with applying Granular Temperature Energy theorem to
the solid phase. The interfacial forces were used to describe the momentum
transfer between the two phases, which has the primary effect on the
hydrodynamic behavior.

A modified gas-solid drag coefficient was used to consider the cohesive behavior of the Geldart-A
particles in order better describe their clustering behavior, and in turn to increase the prediction
accuracy of the numerical reactor model as proposed by Shuai et al. (2011). The
mass and heat transfer conservation equations were coupled to the momentum
equation with considering the source terms due to the regeneration reaction of
the catalyst particles. The proposed numerical model was carried out using the open
source code MFIX with the modified drag model and the regeneration reaction
rate proposed by Syamlal et al. (1993). The length and the height of the computational
domain are 0.286 m and 4.5 m, respectively, and the simulations were performed
for 1000 seconds with a time step of 10^-4 seconds. Because of the good
agreement between the two-dimensional (2D) and the three-dimensional (3D)
simulations of the fluidized bed, the 2D domain was applied to reduce the computational
time for the simulations (Lettieri et al., 2002).

Results

The hydrodynamics
of the FCC regenerator reactor were modeled and large bubbles were detected in
the centerline section of the fluidized bed due to the high bubble coalescence
in presence of the cohesive catalyst particles. The results of the upgraded
hydrodynamic model were validated with experimental hold-up profiles (Ellis
(2003)).

Figure 2: Hold-up
distribution vs. regeneration time of FCC catalyst.

The
thermochemical behavior of the regenerator reactor was studied with the burn-up
reaction of the heavy deposited hydrocarbon. The effects of the deposited
hydrocarbon composition (molar hydrogen-to-carbon ratio) on the FCC catalyst as
well as the influence of the inlet air flow rate condition in the regenerator
operation were studied. For the different case studies, temperature and
chemical species distributions in the fluidized bed reactor were predicted.
Furthermore, the required regeneration time for the FCC catalyst was illustrated
to set desired operation condition of reactor with respect to the best
performance of the FCC reactors. It was found that for lighter deposited coke
(higher value of the molar hydrogen-to-carbon ratio) the temperature peak of
the regenerator reactor increases but the regeneration time reduces slightly. Furthermore,
it was found that the clustering behavior of the FCC catalysts results in
temperature hotspots accompanying with higher vapor compositions, which favor
irreversible disturbance (poisoning) of the catalyst matrix.

References

Ellis, N., Hydrodynamics of gas solid
turbulent fluidized beds, Ph.D. Thesis, University of British Columbia,
Vancouver, British Columbia, Canada, (2003).

Han, I.S.,
Riggs, J.B., Chung, C.B., Modeling and optimization of a fluidized catalytic
cracking process under full and partial combustion modes, Chemical
Engineering and Processing 43 (2004) 1063-1084.

Lettieri, P., Micale, G., Cammarata, L.,
Colman, D., Computational fluid-dynamics simulations of gas-fluidized beds:
A preliminary investigation of different modeling approaches, Proceedings
of the10th workshop on two-phase flow predictions, vol. 1 Merseburg, Germany
(2002, April), pp. 300-309.

Penteado, J.C., Rossi, L.F.S., Negrão, C.O.R., Numerical
Modeling of a FCC Regenerator, 17^th International Congress of
Mechanical Engineering, São Paulo, SP, (2003) COBEM2003-1724.

Shuai, W.,
Huilin, L., Guodong, L., Zhiheng, S., Pengfei, X., Gidaspow, D., Modeling of
cluster structure-dependent drag with Eulerian approach for circulating
fluidized beds, Powder Technology 208 (2011) 98-110.

Syamlal, M., Rogers,
W., O?Brien, T. J., MFIX documentation: Theory guide, Tech. Rep.
DOE/METC-94/1004 (DE9400087), Morgantown Energy Technology Center, Morgantown,
West Virginia (1993).