Analysis of design sensitivity of flow-reversal reactors: Simulations, approximations and oxidation experiments

The analysis of a flow-reversal reactor, constructed from a catalytic bed imbedded within two inert beds and catalysing an instantaneous reaction, shows that the main parameters that determine the maximal reactor temperature are the thermodynamic parameters, heat loss through the walls and the conductivity and length of the inert zones. We show how these parameters can be easily extracted from the experimental data of very fast reactions, and demonstrate it for ethylene and for propane oxidation. We also derive an approximation for the maximal temperature for fast or slow reactions. These approximations are compared with experimental results obtained during propane (with low feed concentration) or methane oxidation.     

Comparison of flow-reversal, internal-recirculation and loop reactors

In this work, we compare the performance of flow-reversal, internal-recirculation and loop reactors. In the absence of analytical results we use asymptotic, approximate and simulated solutions and present some experimental results. As criteria for comparison we use the maximal temperature achieved and the robustness of solution.Experiments and simulations of ethylene oxidation in the flow-reversal and internal-recirculation reactor, showed that the technically simpler inner-outer internal-recycle reactor may operate better at low flow rates than that with flow reversal, but the conclusion is reversed at high flow rates. Using approximate solutions, we show the dependence of the maximal temperature on the inner-outer heat-transfer coefficient.Loop reactor can generate rotating pulse solution: we simulate such solutions for two asymptotic cases where the ratio of switching velocity (i.e., unit length/switching time) to pattern velocity is either around unity or very large. We compare them with solutions of 4-8 units reactors. The slow-switching regimes require a delicate control. The fast-switching solution is robust but its peak temperature depends on the kinetic parameters and reactor length, compared with that of the flow-reversal reactor where it depends mainly on bed conductivity.     

Mathematical simulations of the performance of trickle bed and slurry reactors for methanol synthesis

Three- and two-phase reactor models were developed to simulate the performance of trickle bed and slurry reactors for methanol synthesis. The combination of orthogonal collocation and quasi-linearization was used to solve the trickle bed reactor model incorporating resistance to interparticle and intraparticle diffusion and resistance to mass transfer between gas and liquid phases. Model parameters were estimated independently from either published correlations or literature data. The model predicts significant resistance to intraparticle diffusion on the performance of trickle bed reactors. However, comparisons between pilot size trickle bed and slurry reactors illustrate the superior performance of trickle bed reactors over the slurry reactors for methanol synthesis even with diffusion limitations.     

Combustion of toluene-hexane binary mixtures in a reverse flow catalytic reactor

This work is focused on the application of reverse flow reactors to the combustion of lean mixtures of aliphatic and aromatic hydrocarbons in air. For this purpose, hexane and toluene were chosen as model compounds. The combustion of binary mixtures of these compounds (up to 500 ppmV total hydrocarbon concentration) over a commercial Pt/Al₂O₃ catalyst in reverse flow reactors has been studied both experimentally, in a bench-scale unit, and by simulations, using a heterogeneous mono-dimensional dynamic model, good correspondence being observed between both approaches.As general trend, it was observed that the behaviour of the reactor is determined mainly by the combustion enthalpies and reactivities of toluene and hexane. Hence, increasing total concentration and increasing fraction of toluene (the most reactive compound) lead to more stable operation. Regarding the kinetic inhibition effects, in the conditions studied no influence on the reactor performance was observed, probably because the hydrocarbons combust in different reactor zones. This behaviour can be extended to the combustion of aromatic and C₅-C₈ alkanes, characterised by their relatively low concentrations (determined by their vapour pressure) and high reaction rates.     

Optimum operation of oxidative coupling of methane in porous ceramic membrane reactors

This paper presents analysis of oxidative coupling of methane on Li/MgO packed porous membrane reactor (PMR) by the fixed-bed reactor (FBR) model with reliable reaction kinetic equations. PMR can improve the selectivity and yield by controlling the oxygen feed to the catalyst bed through manipulating the feed pressure. At a fixed methane feed rate there is an optimal oxygen feed pressure that will achieve the highest yield. With a commercial ultrafiltration ceramic membrane, theoretical analysis shows that PMR can achieve, by operating with both side pressures at 1 bar at 750 °C, a maximal 30% yield at 53% selectivity. The maximal yield achieved in the FBR of identical dimension and temperature is 20.7% at 52.5% selectivity. Parametric study shows that lowering the membrane permeability improves the performance. Higher oxygen feed pressure will reduce the yield as well as the selectivity. Homogeneous reactions at high shell-side pressure can have adverse effect on the performance due to the fact that homogeneous reaction rates are strongly pressure dependent. The shell (oxygen feed) side volume must be minimized to reduce the homogeneous reactions. The results of PMR model calculation fit the published experimental result unexpectedly well.     

Modelling of unsteady-state hydrodynamics in periodically operated trickle-bed reactors:Influence of the liquid-phase physical properties

Process intensification using periodic operation of trickle bed reactors (TBRs) is still a long way from replacing conventional steady-state operation in industrial use, despite the numerous benefits described in the literature. Complex interactions between hydrodynamics, mass transfer and reaction phenomena make the design of periodically operated TBRs an almost insurmountable challenge. The development of hydrodynamic models able to provide reliable quantitative predictions of flow behaviour and possessing a sound physical basis, is an essential prerequisite for obtaining the necessary insights into this complexity. In this work, the two-phase pressure drop and dynamic liquid hold-up during max/min and on/off periodical operation were predicted using a model based on the relative permeability concept. In order to demonstrate the utility of this approach, a systematic investigation of the quantitative influence of the liquid-phase physical properties was carried out. The results obtained show that the modelling of the hydrodynamics in periodically operated TBRs using the relative permeability concept is feasible. By selecting suitable permeability parameters, unsteady-state hydrodynamics for different periodic operating modes can be predicted successfully.     

CFD and experimental studies of reactive pulsing flow in environmentally-based trickle-bed reactors

Pulsing flow in trickle-bed reactors (TBRs) has been claimed to promote the averaged heat and mass transfer rates, thereby enhancing the overall conversion and productivity. In this work, we focused on the mineralization of organic matter by catalytic wet oxidation at different liquid flow modulations. The convective nature of the disturbances that lead to pulsing was simulated by an Eulerian Computational Fluid Dynamics (CFD) model and validated with experimental data. In order to evaluate the predicted effects of pulsing on reaction outcome, first the multiphase flow governing equations were detailed with the computational methodology used in the simulation procedure. Prominent numerical parameters were optimized in terms of mesh aperture and time step. Second, to enable objective assessments of the merits of the different cyclic strategies, several computational runs were performed addressing the effect of nominal gas and liquid flow rates as well as the oxidation temperature. Here, we found that the concentration profile computed by the CFD model for pulsing flow conditions demonstrated superior oxidation performance over the trickling flow regime, which has been further corroborated by experimental evidences. Afterwards, the normalized concentration series close to the top of the TBR exhibited sharp fronts and gradually become less intense as they travel downward that reflected the nonisolated nature of the traveling liquid pulsations. Finally, these computational and experimental findings enabled us to intensify the detoxification of high-strength wastewaters and can be further exploited due to advantageous dispersive and convective heat/mass transfer phenomena under pulsing flow conditions.     

Flow and temperature patterns in an inductively coupled plasma reactor: Experimental measurements and CFD simulations

Measurements of temperature patterns in an inductively coupled plasma (ICP) have been carried out experimentally. Plasma torch was operated at different RF powers in the range of 3–14 kW at near atmospheric pressure and over a wide range of sheath gas flow rate (3–25 lpm). Measurements were made at five different axial positions in ICP torch. The chordal intensities were converted into a radial intensity profile by Abel Inversion technique. Typical radial temperature profile shows an off‐axis temperature peak, which shifts toward the wall as the power increases. Temperatures in the range of 6000–14,000 K were recorded by this method. The temperature profiles in the plasma reactor were simulated using computational fluid dynamics (CFD). A good agreement was found between the CFD predictions of the flow and temperature pattern with those published in the literature as well as the temperature profiles measured in the present work. © 2014 American Institute of Chemical Engineers AIChE J, 60: 3647–3664, 2014     

Modelling of methanol synthesis in a network of forced unsteady-state ring reactors by artificial neural networks for control purposes

A numerical model based on artificial neural networks (ANN) was developed to simulate the dynamic behaviour of a three reactors network (or ring reactor), with periodic change of the feed position, when low-pressure methanol synthesis is carried out. A multilayer, feedforward, fully connected ANN was designed and the history stack adaptation algorithm was implemented and tested with quite good results both in terms of model identification and learning rates. The influence of the ANN parameters was addressed, leading to simple guidelines for the selection of their values. A detailed model was used to generate the patterns adopted for the learning and testing phases. The simplified model was finalised to develop a model predictive control scheme in order to maximise methanol yield and to fulfil process constraints.     

Transversal thermal patterns in packed‐bed reactors with simple kinetics: Bifurcation criterion and simulations

We derive a new criterion for transversal instability of planar fronts based on the bifurcation condition dV_f/dK|_K=0 = 0, where V_f and K are the front velocity and its curvature, respectively. This refines our previously obtained condition, which was formulated as α = (ΔT_adPe_T)/(ΔT_mPe_C) > 1 to α > 1 + |δ|, where ΔT_ad and ΔT_m are the adiabatic and maximal temperature rise, respectively, Pe_C and Pe_T are the axial mass and the heat Pe numbers, respectively, and δ is a small parameter. The criterion is based on approximate relations for ΔT_m and V_f, which account for the local curvature of a propagating front in a packed bed reactor with a first‐order activated kinetics. The obtained relations are verified by linear stability analysis of planar fronts. Simulations of a simplified 2D model in the form of a thin cylindrical shell are in good agreement with the critical parameters predicted by dispersion relations. Three types of patterns were detected in simulations: “frozen” multiwave patterns, spinning waves, and complex rotating–oscillating patterns. We map bifurcation diagrams showing domains of different modes using the shell radius as the bifurcation parameter. The possible translation of the 2D cylindrical shell model results to the 3D case is discussed. © 2010 American Institute of Chemical Engineers AIChE J, 2011     

Modeling for the catalytic coupling reaction of carbon monoxide to diethyl oxalate in fixed-bed reactors: Reactor model and its applications

A two-dimensional (2D) pseudo-homogeneous reactor model was developed to simulate the performance of fixed-bed reactors for catalytic coupling reaction of carbon monoxide to diethyl oxalate. Reactor modeling was performed using a comprehensive numerical model consisting of two-dimensional coupled material and energy balance equations. A power law kinetic model was applied for simulating the catalytic coupling reaction with considering one main-reaction and two side-reactions. The validity of the reactor model was tested against the measured data from different-scale demonstration processes and satisfactory agreements between the model prediction and measured results were obtained. Furthermore, detailed numerical simulations were performed to investigate the effect of major operation parameters on the reactor behavior of fixed bed for catalytic coupling reaction of carbon monoxide to diethyl oxalate, and the result shows that the coolant temperature is the most sensitive parameter.     

Application of recurrent neural networks in batch reactors: Part I. NARMA modelling of the dynamic behaviour of the heat transfer fluid temperature

This paper is focused on the development of nonlinear models, using artificial neural networks, able to provide appropriate predictions when acting as process simulators. The dynamic behaviour of the heat transfer fluid temperature in a jacketed chemical reactor has been selected as a case study. Different structures of NARMA (Non-linear ARMA) models have been studied. The experimental results have allowed to carry out a comparison between the different neural approaches and a first-principles model. The best neural results are obtained using a parallel model structure based on a recurrent neural network architecture, which guarantees better dynamic approximations than currently employed neural models. The results suggest that parallel models built up with recurrent networks can be seen as an alternative to phenomenological models for simulating the dynamic behaviour of the heating/cooling circuits which change from batch installation to installation.     

Phenylacetylene hydrogenation in a three-phase catalytic packed-bed reactor: experiments and model

A mathematical model was developed for the cocurrent operation of a three-phase catalytic packed-bed reactor under both trickling- and pulsing-flow regimes. The local fluctuations of liquid-solid mass transfer, liquid flow rate, and liquid holdup in unsteady pulsing-flow were simulated as periodic square-wave functions. The transport properties employed in the model were obtained using published correlations, while expressions for the intrinsic reaction kinetics were taken from our previous work. The model results were found to be in good agreement with experimental data obtained from a laboratory-scale reactor, and verified the advantage of pulsing-flow operation over trickling-flow.     

Investigation of the oxygen evolving electrode in pH-neutral electrolytes: Modelling and experiments of the RDE-cell

A model has been developed to illustrate the complex interplay between the acidifying electrode reactions for oxygen evolution, mass transport and homogeneous reactions in pH-neutral electrolytes. Modelled polarisation curves of the oxygen evolution reaction were verified by polarisation curves experimentally measured in 5 M NaClO₄ on a RDE of DSA material. The conditions in the simulations and in the experiments were similar to those in the chlorate process (high ionic strength, 70 °C, chromate-containing electrolyte, DSA electrode), in which the oxygen evolution reaction is one of the possible side reactions. The model predicted the concentration gradients of H⁺, OH⁻, CrO₄²⁻ and HCrO₄⁻ during oxygen evolution on the RDE. It was found that an important part of the chromate buffering effect at high current densities occurs in a thin (in the order of nanometers) reaction layer at the anode. From comparisons between the model and experiments, a buffering reaction has been proposed. The most likely reaction for the chromate buffering in the investigated system is CrO₄²⁻ reacting with water to HCrO₄⁻ and OH⁻. In the chlorate process, where chromate is a buffer and oxygen evolution is a side reaction, it is likely that chromate promotes oxygen evolution from OH⁻.     

What is the leanest stream to sustain a nonadiabatic loop reactor: Analysis and methane combustion experiments

While previous studies experimentally demonstrated that loop reactor (LR) can be sustained with a lean feed (using ethylene combustion) and have analyzed the single‐reaction adiabatic case, this work analyzes the effects of heat loss and of reactor size to determine the leanest stream (expressed in terms of adiabatic temperature rise ΔT_lim) that will sustain the operation. For an adiabatic infinitely long reactor ΔT_lim→0 while for a finite reactor ΔT_lim scales as (1 + Pe/4)^?1 where Pe = Luρc_pf/k, and heat loss increases this limit by (β/Pe)^1/2. Thus, a good design of a LR will aim to decrease conductivity (k) and radial heat‐transfer coefficient (β) while increasing throughput (u) and reactor length. This article is also the first experimental demonstration of auto‐thermal operation in a LR for catalytic abatement of low‐concentration of methane, showing the leanest stream to be of 8000 ppm vs. 33,000 ppm in a once‐through reactor. Experimental combustion results of methane and of ethylene are compared with model predictions. © 2016 American Institute of Chemical Engineers AIChE J, 63: 2030–2042, 2017     

Flow regimes in a spout-fluid bed: A combined experimental and simulation study

Spout-fluid beds find a widespread application in the process industry for efficient contacting of large particles with a gas. However, detailed understanding of the complex behavior of these systems is lacking, which leads to significant scale-up problems in industry. In this paper we report results of a combined experimental and simulation study on the various regimes, which can be encountered during spout-fluid bed operation.A regime map for a 3D spout-fluid bed was composed employing spectral analysis of pressure drop fluctuations and fast video recordings. In addition 3D Euler-Lagrange computations were performed to assess the capability of the model to reproduce the experimentally observed flow regimes.The influence of the drag closure on the model results was assessed and the influence of the computational grid was studied using a new method for the implementation of the two-way coupling, which is proposed in this paper.For most regimes our model is able to predict the appropriate regime. The frequency, at which the largest power is found, is overpredicted by the model. Contrary to the experimental observations, our model did not predict any large slugs in the slugging bed regime.The remaining differences between the simulated and experimentally observed bed behavior is most likely related to the representation of the effective fluid-particle interaction in our model, which relies on local spatial homogeneity.     

Comparison of fluidized bed flow regimes for steam methane reforming in membrane reactors: A simulation study

An important decision in the design of fluidized bed reactors is which of several flow regimes to choose. Almost all fluidized bed reactor models are restricted to a single flow regime, making comparison difficult, especially near the regime boundaries. This paper examines the performance of fluidized bed methane reformers with three models—a simple equilibrium model and two kinetic distributed models, based on different assumptions of varying sophistication. Membranes are incorporated to improve reactor performance. Eighteen cases are simulated for different flow regimes and membrane configurations. Predictions for the fast fluidization and turbulent flow regimes show that the rate-controlling step is permeation through the membranes. Bubbling regime simulations predict somewhat less hydrogen production than for turbulent and fast fluidization, due to the effects of interphase crossflow and mass transfer. Overall reactor performance is predicted to be best under turbulent fluidization operation. Practical considerations also affect the advantages, shortcomings and ultimate choice of flow regime.     

A comparative study on optimised and non‐optimised axial flow,spherical reactors in naphtha reforming process

In this study, the operating conditions of an axial flow spherical reactor have been optimised using a reliable optimisation technique and the results are compared with the results of non‐optimised conditions. The dynamic behaviour of the reactor has been considered in the optimisation process and orthogonal collocation method has been used in order to solve the obtained equations from mathematical modelling of the process. The goal of this study is to maximise the aromatics and hydrogen production rate. Therefore, the objective function is the combination of two terms which include the production rate of the mentioned components. The catalyst distribution for each reactor, the inlet pressure of the system, Length per radius for each reactor, the naphtha feed molar flow rate and the hydrogen mole fraction in the recycle stream as well as the inlet temperature of each reactor have been optimised in this study. © 2011 Canadian Society for Chemical Engineering     

Hydrodynamics of power-law fluids in trickle-flow reactors: Mechanistic model, experimental verification and simulations

A fully predictive one-dimensional mechanistic model was developed for describing the hydrodynamics of power-law fluids in trickle-bed reactors. The model is a generalization of the slit approach to the case of non-Newtonian fluids obeying Ostwald-deWaele rheological behavior. Without recourse to adjustable parameters, the proposed model enabled prediction of the experimental values of (i) total two-phase total pressure drop and total liquid holdup in the trickle flow regime, (ii) frictional pressure drop in single-phase flows through packed beds, and (iii) total liquid holdup in gravity driven liquid downflow and stagnant gas through packed beds. Parametric simulations guided by knowledge of the behavior of highly viscous Newtonian liquids in trickle beds highlighted the capability of the model in the simulation and design of trickle flow operation using power-law fluids.     

Yield Optimization in a Cycled Trickle‐Bed Reactor: Ethanol Catalytic Oxidation as a Case Study

The effect of slow ON‐OFF liquid flow modulation on the yield of consecutive reactions is investigated for oxidation of aqueous ethanol solutions using a 0.5 % Pd/Al₂O₃ commercial catalyst in a laboratory trickle‐bed reactor. Experiments with modulated liquid flow rate (MLFR) were performed under the same hydrodynamic conditions (degree of wetting, liquid holdup) as experiments with constant liquid flow rate (CLFR). Thus, the impact of the duration of wet and dry cycles as well as the period can be independently investigated. Depending on cycling conditions, acetaldehyde or acetic acid production is favored with MLFR compared to CLFR. Results suggest both the opportunity and challenge of finding a way to tune the cycling parameters for producing the most appropriate product.