Riser Reactor Research Articles

Mathematical modelling of catalytic cracking riser reactor

The quality and the yield of gasoline and light olefins from the catalytic cracking unit depend on a broad range of operation indicators including the feedstock composition, the process conditions, the type and activity of the catalyst. The aim of research is to develop the mathematical model of catalytic cracking reactor on the basis of the formalized mechanism of hydrocarbon conversion taking into account the catalyst deactivation by coke. The experimental research of the feedstock and the product of catalytic cracking using a liquid-adsorption chromatography, gas chromatography–mass spectrometry, gas-liquid chromatography and the structural-group composition methods allowed determining the list of the catalytic cracking reactions. According to the discovered reactions, the thermodynamic analysis was performed using the methods of quantum chemistry. Thermo-gravimetric analysis of the coked catalyst allowed estimating the coke structure formed on the catalyst surface. The developed mathematical model allows to predict the product yields including the content of propane-propylene (PPF) and butane-butylene (BBF) fractions, the group composition and octane number of the gasoline depending on the feedstock composition, the process conditions of reactor-regenerator unit and the catalyst activity. As a result, the high theoretical yield of the gasoline (60.4 wt.%, RON 93.5) according to requirements for the content of olefins and benzene can be achieved at the process temperature of 533°C. This temperature is possible at the keeping the catalyst circulation ratio of 6.9 toncat/tonfeed at the catalyst flow temperature after regeneration (685.8°C) with the catalyst activity of 0.79 unit and the feedstock temperature (328.0°C). The yields of rich gas and coke are 24.0wt% and 4.5 wt%, the concentrations of PPF and BBF are 31.5 and 34.8 wt%.

Simulation of a Large-Scale FCC Riser Using a Combination of MP-PIC and Four-Lump Oil-Cracking Kinetic Models

Fluidized catalytic cracking (FCC) is the most used process for converting heavy oil into more valuable fuels and chemical products such as gasoline and propylene. A three-dimensional reactive gas–particle CFD model was built to study the hydrodynamics, heat transfer, and cracking reaction behaviors within an industrial FCC riser reactor. The multiphase particle-in-cell methodology (MP-PIC) was used to simulate the riser hydrodynamics. A four-lump kinetic model was selected to represent the cracking reaction network in the CFD model. Hydrodynamics, heat, and cracking reactions interplay was discussed.The numerical results of this investigation show a good agreement with the process real data on the yield distribution. The effects of injection (nozzle angle, position, and direction) as well as the riser operating pressure on the yield distribution were quantified and discussed. The methodology employed and the results obtained should serve as guidelines for possible process re-design and optimization.

Predictive Model for Particle Residence Time Distributions in Riser Reactors. Part 1: Model Development and Validation

In this computational study, we model the mixing of biomass pyrolysis vapor with a solid catalyst in circulating riser reactors with a focus on the determination of solid catalyst residence time distributions (RTDs). A comprehensive set of 2D and 3D simulations were conducted for a pilot-scale riser using the Eulerian–Eulerian two-fluid modeling framework with and without subgrid-scale models for the gas–solid interaction. A validation test case was also simulated and compared to experiments, showing agreement in the pressure gradient and RTD mean and spread. For simulation cases, it was found that for accurate RTD prediction, the Johnson and Jackson partial slip solid boundary condition was required for all models, and a subgrid model is useful so that ultra high resolutions grids, which are very computationally intensive, are not required. We discovered a 2/3 scaling relation for the RTD mean and spread when comparing resolved 2D simulations to validated unresolved 3D subgrid-scale model simulations.

Modelling of co-processing of HDO-oil with VGO in a FCC unit

As environmental policies drive to increase the renewables participation in the transportation sector, biomass-derived fuels are becoming an attractive alternative to replace, totally or partially, fossil ones. Taking into account the sustainability concerns about first generation biofuels, attention is paid to second generation ones. Their main drawbacks are the high CAPEX and OPEX required if high quality fuels are required as final product due to the current deployment of its technology. In this way, co-processing of biomass-derived feedstocks in oil refineries can be a good solution and can pave the way for the introduction of biofuels in the market, since it reduces environmental impacts from fossil fuels, decreases the oil dependency of countries and, the most important one, takes advantage of existing facilities avoiding the erection of new plants. In this work, the co-processing of hydrodeoxigenated pyrolysis oil (HDO-oil, from lignocellulosic biomass) with vacuum gas oil (VGO) in a FCC unit is modelled and implemented in a calculation block in MS Excel inside Aspen Plus®. 40 pseudocomponents and 10 real components are defined to represent both feedstocks and their possible cracking products. These components are used in a pseudoreaction mechanism of more than 12,000 single reactions, developed from a probabilistic reaction system. Such a large number of reactions is handled by considering key parameters for the estimation of each reaction rate constant and integrated along the riser which is modelled using a finite volume method. Tuning parameters are used to fit the model depending on the feedstock and the riser conditions, and have to be estimated using experimental data. Thereby, the kinetic reactor model allows to work with different feedstocks and conditions. Material, heat and hydrodynamic balances are performed in all volume elements of the modelled riser reactor, which results in a composition and temperature profiles along the riser. The model is validated with experimental data, showing a low relative error. Therefore, this model can be used to predict product yields, according to different feedstocks and experimental conditions.

Uneven distribution of particle flow in RFCC reactor riser

The uneven distribution of particle flow, i.e., a different particle mass flow rate in each outlet of the riser in residue fluidized catalytic cracking (RFCC) processes, is one major problem associated with commercial RFCC processes. This problem affects the formation of carbonaceous deposits in the secondary reactor cyclone, which incurs serious catalyst carryover in the fractionators. This study analyzes particle-fluid flow patterns in the riser, and diagnoses the uneven distribution of particle flow using a computational particle fluid dynamics (CPFD) method to solve this real industrial problem. Through this analysis, the effect of the number of feed injectors is investigated. The CPFD method, which has been developed to complement the Eulerian-Eulerian and Eulerian-Lagrangian methods, applies the Navier-Stokes equation for fluid phase and multi-phase-particle-in-cell (MP-PIC) models for particle phase. The particle flow distribution was found to vary by 15.5–18.7% at different outlets in the 1 injector case, which implies that the solid loading ratio in each cyclone is different, thereby affecting the separation efficiency of the cyclone and the formation of carbonaceous deposits. The uneven distribution of particle flow phenomena was identified, and the standard deviations of particle mass flow rates were evaluated for the cases of 1, 2, 4, 6, 8 and 12 injectors, and were found to be 7.52, 4.07, 2.66, 1.78, 2.85 and 3.82, respectively. From these results, the 6 injectors case was found to have a largely even particle flow distribution.

Effect on the Flow Behaviors by Adding Internals in a Riser Reactor

Riser reactor is a key unit in the Fluid Catalytic Cracking (FCC), and it has important influences on increasing the yield coefficient of gas and oil. In this paper, the behaviors of gas-solid two-phase flow in the traditional y-type riser reactor are investigated by numerical simulation. The calculated particle concentration distribution is in good agreement with the experimental data, which verified the advanced models and calculating methods. The non-uniform distribution, such as core-annulus flow, may result in the unreasonable matching relationship of catalyst-to-oil ratio. An optimized riser with cuneal internals is proposed and the comparison of two different structures of riser reactor is presented. The comparison results show that the cuneal internals in the riser both can block effectively the slip down of the particles near wall region and weaken core-annulus flow structure due to the redistribution of particles. The results also prove that the particle concentration distribution becomes uniform along the axial and radial direction in the optimized riser by adding cuneal internals, which would be benefits for the catalytic cracking reactions.

FCC riser quick separation system: a review

The riser reactor is the key unit in the fluid catalytic cracking (FCC) process. As the FCC feedstocks become heavier, the product mixture of oil, gas and catalysts must be separated immediately at the outlet of the riser to avoid excessive coking. The quick separation system is the core equipment in the FCC unit. China University of Petroleum (Beijing) has developed many kinds of separation system including the fender-stripping cyclone and circulating-stripping cyclone systems, which can increase the separation efficiency and reduce the pressure drop remarkably. For the inner riser system, a vortex quick separation system has been developed. It contains a vortex quick separator and an isolated shell. In order to reduce the separation time, a new type of separator called the short residence time separator system was developed. It can further reduce the separation time to less than 1 s. In this paper, the corresponding design principles, structure and industrial application of these different kinds of separation systems are reviewed. A system that can simultaneously realize quick oil gas separation, quick oil gas extraction and quick pre-stripping of catalysts at the end of the riser is the trend in the future.

The micro-explosion strength of emulsified heavy oil droplets in catalytic cracking process

The emulsified feedstock technique applies the secondary atomization mechanism to heavy oil catalytic cracking process, which can significantly improve the atomization of feedstock and result in a more desirable product distribution. A model of micro-explosion strength of emulsion droplets in riser reactor is proposed in this paper. Model calculation shows that the superheat limit of emulsified heavy oil is 307.7°C at atmospheric pressure, and the micro-explosion strength increases with the size of dispersed water sub-droplets. The theoretical prediction was validated experimentally through feedstock emulsification tests and catalytic cracking experiments. Results indicate that with an increase of the water content in feedstock emulsification, the size of dispersed water sub-droplets in emulsified oil rises, and thus the micro-explosion strength of emulsion droplets in riser reactor increases, which results in a higher light oil yield and lower coke and dry gas yields of catalytic cracking. Compared with pure heavy oil feedstock, the light oil yield rises by 2%, and the yields of coke and dry gas decline by 1.23% and 0.17% respectively for emulsified feedstock at 5% water content.

Oxidative dehydrogenation of ethane to ethylene in a system with circulating microspherical metal oxide oxygen carrier: 2. Ethylene production in a pilot unit with a riser reactor

The features of interaction of ethane with a circulating microspherical solid oxygen-carrier contact in a flow-through riser reactor have been revealed. The effects of temperature and residence time of hydrocarbons (HCs) in the reactor on the performance characteristics of the process have been examined. Under optimum conditions, the conversion of ethane can be as high as 12%, and the selectivity for ethylene is higher than 90%. The results of determination by X-ray photoelectron spectroscopy of relative atomic ratios of elements in the surface layer of the fresh catalyst and catalyst samples discharged from the unit after the set of experiments are presented.

Radial solids flow structure in high flux gas-solids circulating fluidized bed downers

High flux gas-solids circulating fluidized bed downer reactors have unique characteristics especially suitable for very fast reactions. Detailed hydrodynamics of such reactors are experimentally studied using FCC particles at various superficial gas velocities (1–7m/s) under high flux conditions up to 700kg/m2s for the first time. Results show that although the radial distribution of solids holdup is somewhat less uniform under very high flux conditions in the downers, it is still much more uniform compared to riser reactors. Radial profiles of solids holdup, particle velocity and solids flux are significantly affected by the operating conditions. Particle velocity distribution is characterized by a relatively flat core and an annulus, where the particle velocity slightly increases towards the wall under low flux. In very high flux downer reactors, the shape of the local particle velocity becomes parabolic. It is also found that relationships between local solids holdup and particle velocity are different in the downers compared to the risers due to their different flow characteristics. Compared to the riser reactor, downer has a self-adjusting mechanism to perform with the nearly “ideal” plug flow nature.

CFD optimization of feedstock injection angle in a FCC riser

Feedstock injection zone is a key section for fluid catalytic cracking (FCC) riser reactor. In conventional design of commercial FCC risers, the injection angle of the feedstock is 30° upward with the riser axis, which is found to easily cause nonuniform oil-catalyst contact and severe back-mixing in the feedstock injection zone. In this work, the effect of feedstock injection angle is investigated by performing 3-D simulations with seven injection configurations including three inclined upward, one horizontally and three inclined downward. The two-fluid model (TFM) and the energy-minimization multi-scale (EMMS) drag is combined in simulations. Comparison with experimental data is conducted in terms of the profiles of solids and feed volume fractions. Several variables including the backflow ratio, the radial nonuniformity index and the mean residence time of feed, are further introduced to quantify the hydrodynamic dependence of flow and mixing on the injection settings. It is found that the downward injection schemes are better to realize uniform mixing and matching between catalysts and feed than the upward injection schemes. A 30° downward injection angle is desired to improve the distributions of catalysts and feed, reduce the catalyst-feed contacting time and eliminate the effect of secondary flow on severe back-mixing near the riser wall.

CFD investigation of hydrodynamics, heat transfer and cracking reactions in a large-scale fluidized catalytic cracking riser

A three-dimensional reactive gas-particle CFD model was built to study the hydrodynamics, heat transfer and cracking reaction behaviors within an industrial Fluid Catalytic Cracking (FCC) riser reactor designed to maximize propylene production. The two-fluid methodology (TFM) was used to simulate the riser hydrodynamics with solid phase properties derived from the kinetic theory of granular flows (KTGF). An 11-lump kinetic model was selected to represent the cracking reaction network in the CFD model. The selection of the kinetic model is dictated by the properties of the feedstock processed and the aim of the process which is maximizing propylene. A novel treatment of the coke component was conducted by incorporating coke into the secondary granular phase which is more realistic since carbon deposition occurs on catalyst phase. Momentum transfer, heat transfer and reaction behavior inside the riser were discussed in detail and inhomogeneity in these aspects were observed especially above the high speed injection nozzles. The numerical results of this investigation show a good agreement with the process real data on the yield distribution despite the use of a coarse grid to mesh such an industrial scale FCC riser. The methodology employed used and the results obtained should serve as guidelines for possible process redesign and optimization.

The model prediction of solid phase hydrodynamics in FCC riser feed injection zones

The research regarding particle concentration and particle velocity in Fluid Catalytic Cracking (FCC) riser feed injection zones is of great importance for FCC riser reactor design. In this study, the result of experiment data showed that the particle phase energy density was almost kept constant along radial direction in riser feed injection zone. Based on that, a basic model was proposed to predict particle concentration and particle velocity in an FCC riser feed injection zone. The average prediction deviations of the basic model were 12.00% and 5.71% for the particle concentration and particle velocity, respectively, under the experimental operating conditions. The basic model was simplified for conciseness purposes in the radial position r/R=0–0.7. The average deviations after simplification were 11.64% and 11.41% for the particle concentration and particle velocity, respectively, in the radial position r/R=0–0.7, under experimental operating conditions.

Effects of asymmetric feeding on gas–solid–liquid transport and catalytic cracking reaction in the feed zone of riser

Asymmetric feeding of liquid reactants and catalytic solids are common in the operation of fluid catalytic cracking (FCC) riser reactors. The asymmetric feeding of solids can be caused by the designed-geometry of J-bend right before the riser inlet or unbalanced operation of wall-steam feeding. In order to ensure uniform distributions in local catalytic to oil ratio (CTO) and temperature, a possible measure of matching the asymmetric feed of catalysts is to introduce liquid jets in an accordingly asymmetric feeding. The asymmetric feeding will cause complicated three-phase interactions, in terms of the transport and phase-change characteristics of each interacting phase, in the feed zone of a riser reactor. This study has developed a theoretical framework with an inertia-based multi-spray interaction, which considers not only a complex geometric intervene of mutual-penetration sprays but also the individual spray properties and feeding operations. Based on the fact that the traverse gradient of pressure is insignificant compared to that of axial gradient in a riser flow, a sub-model is also established to redistribute the gas and solid phases within the cross-section of concern. Due to the lack of experiment data or available information on the asymmetric feeding of catalysts from J-bend, we have also conducted a CFD study of a J-bend dense gas–solid flow to obtain the asymmetric distribution catalysts feeding to the feed-zone model. Our modeling examples suggest that the reaction and flow characteristics can be altered noticeably by adjusting asymmetric feedings of liquid reactants, such as changes in droplet penetration, spray coverage and local CTO. In addition, the result of over-penetration of droplets onto the feed zone wall shows the need for a various-droplet size/flowrate model rather than mono-sized/constant-flowrate model of liquid phase. This can also assist in better design or selection of the characteristics for the spray nozzle.

A lumped kinetic model for heavy oil catalytic cracking FDFCC process

ABSTRACTBased on the reaction mechanisms and the characteristics of flexible dual-riser fluid catalytic cracking (FDFCC) process, the reaction networks of 10 lumps for heavy oil and seven lumps for gasoline were put forward. The cracking reactions of resin and asphaltene into gasoline and gases in heavy oil riser reactor were considered as second-order ones while the other reactions were described as first-order ones. The kinetic parameters indicate good consistence with the reaction rules of heavy oil catalytic cracking. The relative errors between calculated values and actual values of products are all less than 4%, which shows the reliability of the models.

A lumped kinetic model for heavy oil catalytic cracking FDFCC process

ABSTRACTBased on the reaction mechanisms and the characteristics of flexible dual-riser fluid catalytic cracking (FDFCC) process, the reaction networks of 10 lumps for heavy oil and seven lumps for gasoline were put forward. The cracking reactions of resin and asphaltene into gasoline and gases in heavy oil riser reactor were considered as second-order ones while the other reactions were described as first-order ones. The kinetic parameters indicate good consistence with the reaction rules of heavy oil catalytic cracking. The relative errors between calculated values and actual values of products are all less than 4%, which shows the reliability of the models.

Experimental evaluation and modeling of liquid jet penetration to estimate droplet size in a three-phase riser reactor

In this work, the effects of injecting an evaporating liquid jet into solid–gas flow are experimentally investigated. A new model (SHED model) and a supplementary model (spray model) have also been proposed to investigate some flow-field characteristics in three-phase fluidized bed with the mean relative error 4.3% between model and measured results. Some experiments were conducted to study the influences of flow-field parameters such as liquid volumetric flow rate, injection velocity, jet angle and gas superficial velocity as well as solid mass flux on the jet penetration depth (JPD). In addition, independent variables were experimentally employed to propose two empirical correlations for JPD by using multiple regression method and spray cone angle (SCA) by using dimensional analysis technique. The mean relative errors between the JPD and SCA correlations versus experimental data were 7.5% and 3.9%, respectively. In addition, in order to identify the variable effect, a parametric study was carried out. Applying the proposed model can avoid direct use of expensive devices to measure JPD and to predict droplet size.

Oxidative dehydrogenation of ethane to ethylene in a system with circulating microspherical metal oxide oxygen carrier: 1. Synthesis and study of the catalytic system

By screening a large series of oxide systems, the most effective modified Mo catalysts for oxidative dehydrogenation of ethane to ethylene with selectivity for the desired product of more than 90% have been found. Using TG–DTA and XRD techniques, it has been shown that the state of the support (Al2O3) determines the formation of active sites. The procedure for preparing microspherical Mo catalysts is characterized by simplicity and complete reproducibility of their properties. Further investigation in this line is associated with testing a pilot batch of the Mo catalyst on a pilot unit with a riser reactor.

CFD simulation of gas–solid two-phase flow and mixing in a FCC riser with feedstock injection

The feedstock injection zone is a key section for Fluid Catalytic Cracking riser reactor. A 3-D Computational Fluid Dynamics simulation is performed by coupling the two-fluid model and the energy-minimization multi-scale (EMMS) drag. The prediction agrees well with the available data and the secondary flow phenomenon is captured. The simulation reveals that the M-shape distributed feed sprays interact with the W-shape distributed solid particles and then cause the undesirable W-shape distribution of the matching ratio of catalysts to feed. The transformation of such uneven distribution to the finally core-annular distribution heavily depends on the secondary flow, whereas its mechanism can be explained by using the Kutta–Joukowski lift theorem. The increase of the feed spray velocity facilitates the feed diffusion and reduces the transition region. However, excessively high jet velocity may intensify the back-mixing near the wall and causes intense attrition or breakage of catalysts and thereby an increase of costs.

Equivalent Reactor Network Model for the Modeling of Fluid Catalytic Cracking Riser Reactor

Modeling description of riser reactors is a highly interesting issue in design and development of fluid catalytic cracking (FCC) processes. However, one of the challenging problems in the modeling of FCC riser reactors is that sophisticated flow-reaction models with high accuracy require time-consuming computation, while simple flow-reaction models with fast computation result in low-accuracy predictions. This dilemma requires new types of coupled flow-reaction models, which should own time-efficient computation and acceptable model accuracy. In this investigation, an Equivalent Reactor Network (ERN) model was developed for a pilot FCC riser reactor. The construction procedure of the ERN model contains two main steps: hydrodynamic simulations under reactive condition and determination of the equivalent reactor network structure. Numerical results demonstrate that with the ERN model the predicted averaged error of the product yields at the riser outlet is 4.69% and the computation time is ∼5 s. Contrast to...