Measuring gas–liquid distribution in a pilot scale monolith reactor via an Industrial Tomography Scanner (ITS)

An Industrial Tomography Scanner (ITS) was designed and developed to study and quantify the phase distribution in a two-phase flow pilot scale monolith reactor that was 24 in. (0.60 m) in diameter and 192 in. (4.9 m) in height. The monolith reactor was operated co-current up-flow in the Taylor flow regime with water as the liquid phase and air as the gas phase. The cross-sectional holdup distributions were measured at three axial elevations. The operating conditions were selected to bracket commercial operating conditions for fixed bed monolithic reactor systems. The results show that ITS can capture the flow features in a large diameter column. Also the findings suggest the need for careful design of the internals of the reactor. Spatial resolution down to 1.5 cm was obtained so that gross phase maldistribution could be reliably observed. However, improvement is needed for the ITS to be effectively utilized in industry.     

Liquid saturation and gas-liquid distribution in multiphase monolithic reactors

The monolith bed is one of the promising catalytic reactors for a number of chemical gas-liquid-solid processes. In the present work, liquid saturations for five different monoliths have been investigated experimentally in a cold-flow unit with a reactor diameter of 5.0 cm. The influences of gas and liquid flow rates and of the direction of two-phase flow on liquid saturation were examined. The results indicate that the direction of flow has no significant influence on liquid saturation for proper gas-liquid distribution. The experimental results are in good agreement with predictions of the drift flux model using the distribution parameter proposed by Ishii (ANL Report ANL-77-47, 1977) along with the assumption of zero drift velocity.In preliminary experiments, gamma-ray computed tomography (CT) has been successfully applied to measure time-averaged liquid distribution over the monolith cross-section in a selected condition. The employment of a nozzle-type distributor provides an almost uniform liquid distribution over the monolith substrate. It is demonstrated that CT is a viable technique for studying two-phase flow in laboratory-scale monolith reactors.     

Influence of channel geometry on hydrodynamics and mass transfer in the monolith film flow reactor

The hydrodynamics and the gas–liquid mass transfer as a function of the channel geometry have been investigated for the monolith film flow reactor. For the hydrodynamic studies, the liquid distribution and the flooding boundaries have been experimentally determined. The liquid distribution improved with increasing liquid flow rate. The flooding limits are in the range of other commercial structured packings and allow operation under industrially relevant conditions. Larger channel sizes and lower surface tension expand the operating window, while viscosity seems to have a minor impact. The gas–liquid mass transfer is a strong function of the surface to volume ratio defined by the channel dimensions. Co- and counter-current flow operation result in similar performance. Furthermore, shorter monoliths, with larger contribution of the inlet section have significant higher mass transfer due to the development of the concentration profile. The obtained k_GLa_V values of around 0.01 s⁻¹ are in the range of other commercial packings in counter-current flow operation. A three-dimensional single channel model describing the hydrodynamic and diffusion phenomena in the monolith is in good agreement with the experimental results. The flexibility in channel size and dimension allows tailoring the monolith reactor to the specific needs of the individual application.     

Modelling propane dehydrogenation in a rotating monolith reactor

The catalytic dehydrogenation of propane is equilibrium limited, strongly endothermic and normally carried out at high temperatures. The catalyst deactivates due to the laydown of carbonaceous species on the surface. This is conventionally countered by subjecting the catalyst to periodic regeneration. In commercially available processes, the catalyst time on line for a given cycle is in the order of 10–10,000 min.

In this study, the catalyst has been observed to exhibit very high activity and selectivity in the short period after regeneration. Conceptual and model development of a reactor with structured catalyst to capitalise on this beneficial early activity is presented.

The preferred reactor comprises a cylindrical block of honeycomb monolith that rotates past various feed zones, subjecting the catalyst successively to propane and regenerating gas. The exothermic nature of the regeneration reactions is used at least in part to provide heat to the endothermic dehydrogenation reaction via the regenerative heat transfer facilitated by the movement of the solid monolith. Specifically, it is noted that an oxidisible catalyst provides operating advantage due to the additional exotherms associated with the regeneration stage.

The process modelling shows the design to be feasible in terms of matching the heats of reactions and achieving high conversions, but questions are raised over its practicability from mechanical design and process stability viewpoints.     

Experimental investigation of gas-liquid distribution in monolith reactors

The present work investigates the influence of gas and liquid flow rates on inlet liquid distribution across monoliths operating in gas-liquid cocurrent downflow mode. Gas and liquid superficial velocities range from 0 to 68 and 1.4 to 8.5 cm/s, respectively. Gas-liquid distribution was studied using a packed bed liquid distributor and a pipe distributor for the aforementioned range of operating conditions. To determine the liquid distribution over the monolith, gravimetric, time-averaged liquid collection method was applied using a customized collector apparatus. Quantification of the distribution is reported using a suitably defined maldistribution factor. For each liquid velocity, gas velocities are varied and corresponding maldistribution factors are calculated. The results are reported in view of the varying operating conditions.     

A study of Nusselt and Sherwood numbers in a monolith reactor

A two-dimensional model of a single channel of a monolith reactor is used to evaluate the values of the Nusselt and Sherwood numbers under reaction conditions. The circular channel is assumed to have axisymmetry with a first-order reaction occurring at the wall. The values of the Nusselt and Sherwood numbers do not correlate uniquely with the Graetz number but rather depend on the reaction rate at the wall. Hence they depend on such variables as gas velocity, inlet temperature and reactant concentration.     

Effect of pulsing on reaction outcome in a gas–liquid catalytic packed-bed reactor

A novel experiment is described for studying the effect of flow regime on reaction outcome for a consecutive-parallel reaction. By taking advantage of the convective nature of disturbances that grow into pulses in gas–liquid packed-bed reactors, it is shown that it is possible to compare reaction behavior for pulsing and trickling at the same flow rates. This contrasts previous studies where effects of regime were found, but at different flow rates. This experiment is accomplished by packing the column with mostly inert particles and confining the catalytically active region either near the inlet, where pulses have not yet formed, or near the end where they have developed. It is found that for the reaction of phenylacetylene to styrene and ethylbenzene over a platinum/alumina catalyst, where pulses are present in the bottom of the reactor but not at the top, about a 15% increase in styrene concentration, as an intermediate, occurs under pulsing conditions.     

Liquid phase residence time distribution for gas–liquid flow in Kenics static mixer

The residence time distribution (RTD) of the liquid phase for co-current gas–liquid upflow in a Kenics static mixer (KSM) with air/water and air/non-Newtonian fluid systems was investigated. The effect of liquid and gas superficial velocities on liquid holdup and Peclet number was studied. Experiments were conducted in three KSMs of diameter 2.54 cm with 16 elements and 5.08 cm diameter with 8 and 16 elements, respectively, of constant L_e/D_e = 1.5 for different liquid and gas velocities. A correlation was developed for Peclet number, in terms of generalized liquid Reynolds number, gas Froude number and liquid Galileo number, where as for liquid holdup, a correlation was developed as a function of gas Reynolds number. The axial dispersion model was found to be in good agreement with the experimental data.     

Hydrodynamics of churn turbulent bubble columns: gas–liquid recirculation and mechanistic modeling

A phenomenological (mechanistic) model has been developed for describing the gas and liquid/slurry phase mixing in churn turbulent bubble columns. The gas and liquid phase recirculation rates in the reactor, which are needed as inputs to the mechanistic reactor model are estimated via a sub-model which uses the two-fluid approach in solving the Navier–Stokes equations. This sub-model estimates the effective bubble diameter in the reactor cross-section and provides a consistent basis for the estimation of the volumetric mass transfer coefficients. The strategy for the numerical solution of the sub-model equations is presented along with the simulation results for a few cases. The overall reactor model has been tested against experimental data from radioactive gas tracer experiments conducted at the Alternate Fuels Development Unit (AFDU), La Porte, TX under conditions of methanol synthesis.     

Phase distributions in a gas–liquid–solid circulating fluidized bed riser

The distributions of the three phases in gas–liquid–solid circulating fluidized beds (GLSCFB) were studied using a novel measurement technique that combines electrical resistance tomography (ERT) and optical fibre probe. The introduction of gas into a liquid–solid circulating fluidized bed (LSCFB), thus forming a GLSCFB, caused the increase of solids holdup due to the significantly decreased available buoyancy with the lower density of the gas, even with a somewhat increased liquid velocity due to the decreased liquid holdup giving space for the gas holdup. The gas passed through the riser in the form of bubbles, which tended to flow more through the central region of the riser, leading to more radial non‐uniformity in radial holdup of the phases. The gas velocity has the most significant effect on the gas phase holdup. While the gas velocity also has an obvious effect to the solids holdups, the liquid flow rate had a much more considerable effect on the phase holdups. The solids circulation rate also had a significant effect on the phase holdups, with increasing solids circulation rate causing much more increased solids holdup in the central region than close to the wall. A correlation was developed for the relative radial distributions of solids holdup in GLSCFB, as such radial profiles were found similar over a wide range of operating conditions, like those in a typical gas–solid circulating fluidized beds (GSCFB). Finally, the axial solids profiles in a GLSCFB was found to be much closer to those in an LSCFB which are very uniform, than those found in a GSCFB which are less uniform and sometime having a S shape. Water was used as the continuous and conductive phase, air was the gas phase and glass bead and lava rock particles were used as the solid and non‐conductive phase.     

Euler–Lagrange modeling of a gas–liquid stirred reactor with consideration of bubble breakage and coalescence

Simulations of a gas–liquid stirred reactor including bubble breakage and coalescence were performed. The filtered conservation equations for the liquid phase were discretized using a lattice‐Boltzmann scheme. A Lagrangian approach with a bubble parcel concept was used for the dispersed gas phase. Bubble breakage and coalescence were modeled as stochastic events. Additional assumptions for bubble breakup modeling in an Euler–Lagrange framework were proposed. The action of the reactor components on the liquid flow field was described using an immersed boundary condition. The predicted number‐based mean diameter and long‐term averaged liquid velocity components agree qualitatively and quantitatively well with experimental data for a laboratory‐scale gas–liquid stirred reactor with dilute dispersion. Effects of the presence of bubbles, as well as the increase in the gas flow rate, on the hydrodynamics were numerically studied. The modeling technique offers an alternative engineering tool to gain detailed insights into complex industrial‐scale gas–liquid stirred reactors. © 2011 American Institute of Chemical Engineers AIChE J, 2012     

Mass transfer characteristics in a gas–liquid reciprocating plate column. II. Interfacial area

This paper is the second part of a continuing study on mass transfer in a reciprocating plate column. The first part dealt with k_La. The bubble size distribution, the Sauter mean diameter and the interfacial area are the subject of this paper. The bubble size increases slightly with gas flow rate and decreases with agitation intensity above a “critical” level. The interfacial area increases with increasing agitation and aeration intensities, while the liquid flow rate and coalescing properties of the liquid have no significant effect. The specific interfacial area is correlated in terms of the superficial gas velocity and the maximum power consumption. The correlations obtained for k_La and a were used to calculate k_L. It was found that k_L depends on the agitation intensity and the bubble size.     

Modeling of a metal monolith catalytic reactor for methane steam reforming-combustion coupling

A novel metal monolith reactor for coupling methane steam reforming with catalytic combustion is proposed in this work, the metal monolith is used as a co-current heat exchanger and the catalysts are deposited on channel walls of the monolith. The transport and reaction performances of the reactor are numerically studied utilizing heterogeneous model based on the whole reactor. The influence of the operating conditions like feed gas velocity, temperature and composition are predicted to be significant and they must be carefully adjusted in order to avoid hot spots or insufficient methane conversion. To improve reactor performance, several different channel arrangements and catalyst distribution modes in the monolith are designed and simulated. It is demonstrated that reasonable reactor configuration, structure parameters and catalyst distribution can considerably enhance heat transfer and increase the methane conversion, resulting in a compact and intensified unit.     

Visual study on the characteristics of liquid and droplet in a novel rotor-stator reactor

Rotor–stator reactor (RSR), an efficient mass transfer enhancer, has been applied in many fields. However, the hydrodynamic characteristics of liquid flow in RSR are still a mystery despite they are fundamental for the mass transfer performance and processing capacity. In view of the above, this paper studies the liquid–liquid flow and liquid holdup in RSR under various conditions with a high-speed camera. The paper firstly demonstrates two flow patterns and liquid holdup patterns that we obtained from our experiment and then presents in succession a flow pattern and a liquid holdup criterion for the transition of film flow to filament flow and complete filling to incomplete filling. It is found that experimental parameters, including rotor–stator distance, rotational speed and volume flow rate exert great influence on the average droplet diameter and size distribution. Besides, by comparison and contrast, we also find that the experimental values match well with our previous predicted calculations of the average diameter, and the relation between the average diameter and the mean energy dissipation rate.     

Characterization of the performance of an industrial monolith reactor by accurate mapping of temperature differences

The performance of a liquid-phase hydrogenation carried out in a monolithic reactor on an industrial scale has been studied by measuring temperature differences accurately. The process studied is the hydrogenator in the autoxidation process for large-scale production of hydrogen peroxide by Eka Chemicals, Akzo Nobel.Temperature differences are primarily caused by the exothermic reaction and can be used to measure variations in local reaction rates. Non-uniformities in liquid flow distribution can also be detected. One important finding was the existence of multiple steady states that can be attributed to the stop/startup procedure.     

Gas–liquid mass transfer of aqueous Taylor flow in monoliths

The gas–liquid mass transfer of a monolith operating in the Taylor flow regime is presented. Mass transfer measurements are compared with a literature model derived for single capillaries. The comparison resulted in a prediction of the unit cell length (gasbubble+liquidslug). Independent measurements of the liquid slug length showed that the predicted unit cell length is close to the measured ones. This leads to the conclusion that mass transfer models for single capillaries may indeed be used for monoliths. Additionally, it is shown that the liquid slug length may also be estimated from pressure drop measurements.     

Inclination effects on wave characteristics in annular gas–liquid flows

Measurements of wave characteristics have been conducted in a 0.0762 m internal diameter (ID) pipe at inclinations of 0°, 10°, 20°, 45°, 60°, 75°, and 90° from horizontal. Wave celerity and frequency are very strongly dependent on modified Lockhart–Martinelli parameter, X*, and the inclination angle. Wave amplitude increases with increasing liquid film thickness at the bottom of the pipe. Wave amplitude depends on liquid film thickness for any pipe diameter, surface tension, and viscosity. Strouhal number (dimensionless wave frequency) decreases with increasing X*. Effect of pipe diameter, surface tension, and liquid viscosity on the liquid film Reynolds number, Re_LF, was studied. Re_LF variation with X* is not sensitive to the surface tension and less sensitive to the pipe diameter. However, Re_LF is very sensitive to the viscosity of the flowing liquid. Correlations for wave celerity, amplitude, frequency, and liquid film Reynolds number are proposed. © 2011 American Institute of Chemical Engineers AIChE J, 2012