Two-component model for catalyst deactivation
Introduction
Running out of conventional light oils make petroleum industries turn to heavy oils with high viscosity, density and content of impurities to cover our needs in liquid fuel. Among catalytic technologies for heavy oil upgrading, hydroprocessing in a fixed-bed reactor still remains the most widely used [1], [2].
One of the main problems in this approach is a rapid catalyst deactivation by coke and metal deposits. Asphaltenes – high-molecular aromatic Ni, V, S, N-containing nanoparticles with a size from 1.5 to 10 nm [3], [4] – are found in heavy oils in large quantities (5–15 wt%) responsible for poisoning of acidic sites, formation of coke species and, finally, pore plugging [5]. To prolong catalyst lifetime, texture of the catalyst should be optimized. Enlarging the pore size in the range 30–200 nm reduces diffusional limitations and prevents rapid plugging of pore cavities with coke species. Nevertheless, too high pore size (>200 nm) is not preferred as well, since it leads to significant decrease in surface area, making them ill-suitable for good catalytic performance.
Deactivation of hydroprocessing catalysts has been a subject of numerous theoretical works. Hydroprocessing reactions in such multicomponent systems as heavy oils were usually described by lumped kinetics with the reaction orders from 0.55 to 2.0 [6]. Hydrodemetallization (HDM) and hydrodeasphaltenization (HDAsph) reactions were often represented by simplified first-order kinetics assuming Freundlich or Langmur-Hinshelwood adsorption [7]. The most widely used theory to calculate diffusion of asphaltene in porous structure was quasi homogeneous model [7], [8], which is based on continuity equations: the transport of solute – spheres with Brownian and hydrodynamic properties – is described by Fick law. Given a steady state case, in which the reagent flow to the pellet is compensated by the reagent consumption according to the reaction, the following expression could be formulated:
Catalyst porous structure is defined through porosity ε and tortuosity τ, which are composed into the effective diffusion coefficient Deff according to equation:where Dbulk is a combination of molecular and Knudsen diffusitivity in a straight pore, f(Rmol/Rpore) is a power or exponential dependence originated from the Renkin’s equation, which is valid at Rmol/Rpore < 0.5 [9]. This function reflects both partition of the solute concentration inside and outside porous structure and hydrodynamic drag experienced by the molecule as it moves through the pore. Using this equation, effective diffusion coefficient was estimated to be in the range 10−10–10−12 (m2/s) in various porous materials [7].
Among the reported deactivation mechanisms one could mention surface poisoning [10], competitive adsorption [11], uniform pore size reduction and plugging [12], pore plugging by lumps of metal and coke deposits [13]. Percolation theory [14], [15] explicitly incorporates porous structure into description of deactivation and takes into account pore connectivity, unlike the pseudo-homogeneous model.
The present paper aims at modeling catalyst deactivation and calculating the concentration profiles of coke depositions within the catalyst pellet. The model includes construction a porous material using Poisson distributed grains, solving Fick’s diffusion equation to obtain the concentration profile for given porous structure. Deactivation was modeled by reduction of pore size through monotonic increase of alumina grains radii. In the theoretical model it was proposed that coke depositions appear uniformly on the surface of the spherical alumina grains. The suggested model is not homogeneous: the geometric characteristics of the porous media depend on the position inside the catalyst pellet. The evolution of the asphaltene and coke concentration profiles in the cylindrical catalyst pellet during deactivation was simulated.
Section snippets
Model description
The model is two-component and describes the diffusion of asphaltene molecules inside the cylindrical pellet and deposition of the coke particles from heavy oil on the surface of the catalyst. The deactivation process is described via two distribution functions of the asphaltenes and the coke depositions , depending on the position inside the pellet and the stage of deactivation. At each stage of deactivation the following system is considered:
Here
Results and discussion
Deactivation of the catalyst was defined through evolution of the textural parameters. A consecutive increase in the alumina grain size leaded to the following trends in porosity, tortuosity and specific surface area (Fig. 1). The textural parameters obtained are consistent with those of the real catalysts. System of alumina grains was modeled as a random packing of the Poisson distributed spheres of diameter 10 nm, the packing for mesoporous catalyst in a cube with 1 µc edge contains
Conclusion
A novel two-component deactivation model was designed based on the diffusion of asphaltene molecules inside the cylindrical pellet and deposition of the coke particles from heavy oil on the surface of the catalyst. The modeling of porous catalyst was carried out by a set of Poisson distributed grains. Texture evolution during deactivation was specified through the increase of the catalyst grain diameters. Model describes the evolution of asphaltene and coke concentration profiles during the
Acknowledgements
The present work was supported by RFBR and Government of the Novosibirsk region according to the research project N17-43-543303. Ya. V. Bazaikin was supported by the Program of fundamental scientific researches of the SB RAS No. I.1.2., project No. 0314-2019-0006.
References (18)
- et al.
Modeling residue hydroprocessing in a multi-fixed-bed reactor system
Appl. Catal. A: General
(2008) - et al.
Models for catalytic pore plugging: application to hydrodemetallation
Chem. Eng. Sci.
(1989) A percolation model of catalyst deactivation by site coverage and pore blockage
J. Catal.
(1985)Modeling of catalyst deactivation
Appl. Catal. A: General
(2001)- et al.
A review of methods for the demetallization of residual fuel oils
Fuel Process. Technol.
(2006) - et al.
On the shape of a set of points in the plane
IEEE Trans. Inform. Theory
(1983) Deactivation of Heavy Oil Hydroprocessing Catalysts: Fundamentals and Modeling
(2016)- et al.
Asphaltene diffusion and adsorption in modified NiMo alumina catalysts followed by ultraviolet (UV) spectroscopy
Energy Fuels
(2010) - et al.
Diffusivity of asphaltene molecules by fluorescence correlation spectroscopy
J. Phys. Chem. A
(2006)
Cited by (7)
Enhancing low-temperature thermal remediation of petroleum sludge by solvent deasphalting
2022, ChemosphereCitation Excerpt :The appearance of the MPS was black and sticky (Figure S2). TGA and DTG are effective technologies to investigate the kinetic, thermal degradation and thermochemical conversions of oily wastes and fossil fuels (Malkovich et al., 2019). Previous reports investigated the pyrolysis process of PS and oil sand using TGA (Malkovich et al., 2019; Stanislaus et al., 2005; Gautam et al., 2020; Hao et al., 2017).
Photoinduction of palladium single atoms supported on defect-containing γ-AlOOH nanoleaf for efficient trans-stilbene epoxidation
2022, Chemical Engineering JournalCitation Excerpt :To date, a variety of approaches to constructing SACs have been reported, but the search for general methods that can facilely construct SACs with high activity and stability is still challenging [21–24]. Moreover, detachment and aggregation of the metal species, burying of the metal species by the support or cocking, and blocking of active sites by strong binding of foreign species are the potential causes for catalyst life reduction [25–28]. Therefore, strong metal-support interactions are key to achieving exceptional catalytic activity and stability [11,21,29,30].
4-Scale model for macromolecule conversion over mesoporous and hierarchical alumina catalysts
2021, Chemical Engineering JournalCitation Excerpt :Previously, we have presented the first steps of theoretical model, which was aimed to determine the optimal textural parameters of the catalyst, i.e. the required macropore size, compared with our experimental results on extra-heavy oil hydroprocessing [18]. The model included construction of a set of porous materials using Poisson distributed grains, calculation of surface area by Monte-Carlo method and tortuosity by Dijkstra algorithm for the percolation graph [19]. Here we present the next steps of the model development which is divided into four interconnected numerical calculations of different scales for the following phenomena: 1) effect of increasing the concentration of coke grains and their size (nanometer scale) on the porosity, tortuosity, and specific area of the catalyst; 2) percolation graphs of the mesoporous and hierarchically porous catalysts (dozens of nanometers); 3) kinetic patterns for asphaltene conversion and catalyst deactivation in the mesoporous and hierarchically porous pellets (millimeter scale); 4) macrokinetic model for reactor operation filled with mesoporous and hierarchically porous pellets (centimeter scale).
Evaluation of Time-on-Stream Deactivation Models with Effectiveness Factors for Industrial Feedstock Hydrotreating in Bench-Scale Tests
2023, Industrial and Engineering Chemistry Research