Band 3, Heft 1

This review examines the key features and configurations of short residence time cracking processes from a diverse range of industries that have been developed over the past 25 years. These industries include: bitumen or heavy oil upgrading, biomass pyrolysis, olefin production, catalytic cracking, and coal gasification. Characterization of the gas, liquid, and solid products and feedstock is provided wherever possible. In addition, a description of the source and mechanism of heat transfer, and how the feedstock is brought into contact with and separated from – this source is also given.There is a strong economic incentive for considering short residence time cracking processes. Not only do such processes increase the yields of the more valuable liquid and gaseous products, but more compact designs would also decrease capital costs. Careful control of the vapour residence times appears to be crucial in order to prevent secondary cracking and yet allow for maximum cracking of the feedstock. Rapid and thorough mixing of the feedstock with the heat source, not just creating a uniform dispersion, is also a key design aspect to consider. Finally, rapid and complete separation must also be carefully considered; again, to help control product residence time and avoid secondary cracking but also from a heat balance point of view.

Recent research results on microchannel reactors are reviewed with particular reference to their applications and use as a cost effective tool during process development tasks. The high surface to volume ratio, efficient heat and mass transfer characteristics, vastly improved fluid mixing etc., allow precision control of reaction with improved conversions, selectivities and yields of desired products. The reaction times are shorter as compared to the conventional reactors with less degradation and side products. Scalability and optimization are also significantly easier. The article reviews variety of applications to catalysis, polymers, advanced materials, biosystems, organic reaction etc., with emphasis on the related engineering science issues.

Discovered in 1991, carbon nanotubes (CNTs) have reached the forefront of many industrial research projects. Carbon nanotubes are tubular carbon molecules with remarkable mechanical, chemical, thermal and electrical properties, which make them useful in various applications. This paper reviews three methods of synthesizing the nanomaterial, namely arc-discharge, laser-ablation and fixed bed/fluidized bed catalytic. These methods have generated a large interest in many industrial companies to date. At the moment, the most critical issue faced by industrial companies is determining the best synthesis method, which will give the most economical large-scale production of CNTs. Compared to the other two methods, the catalytic technique to synthesize CNTs is simple, inexpensive, energy-efficient and can produce large CNTs quantities of high yield and purity.

Packed bed columns are intensively used as multiphase contactors and reactors in chemical, biochemical, and petrochemical industries. Due to two-phase through the packed beds, the performance of multiphase contactors depends on the complex interaction of the interphase mass transfer, thermodynamics and hydrodynamics. Under reaction conditions, the intraparticle mass transfer and reaction kinetics have additional considerable impact on the reactor performance. The inherent complexity of multiphase systems leaves many issues unresolved and leads to many inconsistencies and questionable approximations, especially in the modelling of multiphase flow. Multiphase flow modelling in packed beds is a complicated task because of the difficulty incorporating the complex geometry into the flow equations, and the difficulty in accounting for the gas-liquid interactions in the presence of complex fluid-particle (e.g., partial wetting) contacting. This review focuses on recent advances made in the modelling of the hydrodynamics of two-phase downflow and counter-current flow in packed beds via two-fluid models coupled with slit and double slit phenomenological approximations relating the mutual three phase hydrodynamic interactions.

The effect of gas permeation through horizontally immersed membrane tubes on the heat transfer characteristics in a membrane assisted fluidized bed operated in the bubbling fluidization regime was investigated experimentally. Local time-averaged heat transfer coefficients from copper tubes arranged in a staggered formation with the membrane tubes to the fluidized bed were measured in a square bed (0.15 m x 0.15 m x 0.95 m). Glass particles (75-110 micrometer) were fluidized with air distributed via a porous plate, where the ratio of gas fed or removed through the membrane bundles and the porous plate distributor was varied. The experimental results revealed that high gas permeation rates through the membranes strongly decreased the heat transfer coefficient at high superficial gas velocities for tubes at the top of the tube bundle, which was attributed to the reduced mobility and increased bubble hold up and/or dilution of the emulsion phase, reducing overall heat capacity.In the design of membrane assisted fluidized beds care must be taken to include the effect of gas addition or withdrawal through the membranes on the required heat transfer surface area.

In order to decrease the manufacturing cost, the molten poly(ethylene terephthalate)(PET) produced by batch process is stored in a large tank named Pool Tank for after treatment to bring about a continuous manufacture. The alterations of quality happened during the storage should be precisely controlled. In this paper, a method for prediction of fluctuation of molten PET properties with computer simulation only was designed. First of all, a new reaction model which could present the mechanism of polycondensation occurred in Pool Tank was proposed. Then, reaction rate constants were determined by experiments according to the optimization procedure and subsequently, applied to the simulation of storage of molten PET. The fluctuation of intrinsic viscosity and concentration of carboxyl terminal groups of molten PET was predicted. Meanwhile, some useful information for the design of storing tank could be obtained.

For complex reactions the optimal reactor networks can involve several reactors operating at various temperature profiles. The often-reported strategy of optimizing parameters of a heuristically predetermined reactor system (Super-structure Approach) falls short of obtaining true solution due to the presence of multiple local optima. Attainable set method gives Global optimum but requires study of each reaction scheme in depth. Here one such study using phase-plane analysis (instead of convexity based analysis) is reported for finding globally optimal non-isothermal reactor network for van de Vusse reaction (A -> B -> C, 2A -> D, objective is to maximize yield of B). Compared to two-reactor networks proposed earlier, it is found that up to 5 reactors (CSTR with/without bypass of feed, Isothermal PFR, Non-isothermal PFR, CSTR, Isothermal PFR) may be required to get the highest yield of the desired intermediate. The proposed method involves only elementary calculus. The detailed solution algorithm has been described using analogy with highways. Three cases with the values of reaction constants reported in the literature have been solved.

The research outlined here includes a study of the production of carbon black (CB) in an inductive plasma reactor, using dodecane (C12H26) as the starting material. Thermodynamic and kinetic studies were carried out to predict the species concentrations likely to be obtained in the pyrolysis process at plasma temperatures. A thorough statistical experimental design was undertaken to investigate the influence of the different operational conditions such as: the feed rates of dodecane, the reactor pressure, the plate power applied to the plasma torch and the composition of the plasma gas, on the production of carbon black and gaseous acetylene-like compounds. Thermodynamic and kinetic models were compared with experimental results, a kinetic reaction model best representing the experimental results. Morphological analysis of the solid product using Transmission Electron Microscopy (TEM) indicates that a high plasma temperature is the most important factor affecting the final morphology of the carbon black formed during the reaction. Carbon black with average particle size of 10-30 nm and specific surface of 130 m2/g was obtained. Morphological analyses also demonstrated the presence of new structures ranging from carbon black to fullerenes and including certain “graphitized” carbon molecules.

In this study, catabolic repression of tannase in solid-state culture of Aspergillus niger Aa-20 on polyurethane foam was investigated considering the diffusion of a repressor molecule as the main responsible factor. The fundamental hypothesis considers that if repressor molecule uptake rate is higher than repressor molecule diffusion rate on solid matrix, the catabolic repression phenomenon will be minimized. Results showed that addition of low glucose concentrations enhanced the extracellular and intracellular tannase activities (from 5.8 to 6.6 and from 0.96 to 1.76 IU). Besides the glucose uptake rate was higher (1.23 x 10-5 cm2/seg) than glucose diffusion rate on solid matrix (2.8 x 10-6 cm2/seg). On the contrary results at high glucose concentrations, showed that when extracellular tannase activity decreased from 6.6 to 1.87 IU, while intracellular tannase activity was not affected; and under these conditions the glucose diffusion rate governed the process. A particular behaviour was obtained in the repression ratio and the glucose uptake rate/glucose diffusion rate ratio, which permitted to know the sensitivity degree of the culture system, according to the glucose amount added. Substrate diffusion in solid-state culture can be a key factor in its capacity to minimize the catabolic repression of hydrolytic fungal enzymes.

During the oxidation of toluene under semibatch conditions, the formation of benzyl alcohol is initially equal to the rate of formation of benzaldehyde. As the overall conversion increases the benzyl alcohol concentration at first decreases much faster than benzaldehyde, but this decrease slows down causing the benzyl alcohol concentration to reduce to zero only very slowly. To account for this phenomenon a new reaction pathway has been proposed where the formation of benzaldehyde out of benzylhydroperoxide is catalysed by benzoic acid. Incorporation of this new reaction in a model improves the description of benzyl alcohol concentration prophiles while maintaining good predictions for benzaldehyde and benzoic acid.

A numerical study has been undertaken to explore the influence of geometry and flow parameters on the entrainment of solid in an ESE nozzle system immersed in a fluidized riser. A fully three-dimensional computational model of the nozzle system has been developed and all appropriate approximations and simplifications are described. A multi-phase Eulerian-Eulerian model incorporating the kinetic theory for solid particles is used. Numerical results are obtained using the commercial Computational Fluid Dynamics software FLUENT. The results indicate that solid entrainment in the ESE system is a strong function of both geometry and flow. The optimal entrainment is seen to occur when the ratio of the draft tube diameter D to separation distance I is approximately unity. At this value, the jet of injected gas is seen to spread fully into the opening of the draft tube causing the highest transport of solid particles through the tube. The entrainment is shown to increase with increasing jet velocity across the full range of flows considered. The results are consistent with similar experimental results. The results of this study should find immediate application in the design and implementation of ESE nozzle systems.

In this work an Artificial Neural Networks (ANN) approach for estimation of catalyst deactivation during methanol synthesis has been proposed. The approach predicts deactivation of the catalyst at different operating conditions as a function of time. Experimental data of a typical differential reactor were pre-scaled and used for training. Among the various training algorithms, Exact Radial Basis (RBE) method had the best prediction performance and was used for simulation of the reactor. The proposed approach interprets deactivation data, while there are not enough data versus time at different inlet conditions. By using the model, sufficient data were generated which vary with time and agree very well with experimental data. Capability of the model in generating deactivation data at different temperatures, pressures and feed compositions was excellent. The proposed method has great potential as a means to compensate for lack of the phenomenological kinetic modeling techniques.

The dynamic behavior of a multistage agitated contactor has been numerically studied for its potential use for catalytic hydrogenation of nitrile butadiene rubber. A model consisting of stirred tanks in series with back flow is applied to describe the flow behavior in the presence of hydrogenation within a multistage agitated contactor of six stages. The coupled behavior between kinetics and mass transfer is analyzed by manipulating parameters involved in the model. The model predicted that the performance of the reactor is completely governed by the ability of the catalyst in activating hydrogen, the loading level of C=C double bonds, the relative capacity of the reaction to mass transfer, and the backmixing. Mass transfer and backmixing are important factors for the hydrogenation process and enhancing mass transfer and minimizing backmixing helps in increasing hydrogenation conversion. The effects of backmixing and the number of stages for the hydrogenation are discussed, and optimal operation conditions are suggested.

The hydrogenation of acrolein was investigated experimentally in a fixed-bed reactor (FBR) using several classical and a newly developed hydrogenation catalyst. The aim was to evaluate selectivity and yield with respect to the desired product allyl alcohol. The kinetics of the two main parallel reactions of acrolein hydrogenation were quantified for a supported silver catalyst which offered the highest performance. In a second part the reaction kinetics identified were used in a theoretical study applying a simplified isothermal 1D reactor model in order to analyse the hydrogenation of acrolein performed in single- and multi-stage packed bed membrane reactors (PBMR). The goal of the simulations was to evaluate the potential of dosing one reactant in a distributed manner using one or several membrane reactor stages. The results achieved indicate that the membrane reactor concept possesses the potential to provide improved yields of allyl alcohol compared to conventional co-feed fixed-bed operation.

With ever-growing environmental concerns, petrochemical and fine chemical industries face an omnipresent issue in recovering dilute acetic acid from its aqueous solutions. Catalytic distillation holds an ascendancy over conventional physical separation methods such as distillation and extraction. Distillation is associated with the high costs involved in vaporising the more volatile water that exists in high proportions and possesses a high latent heat of vaporisation. Extraction is limited in view of the distribution of the components in the reacting system. The implementation of catalytic distillation reduces capital and operating costs, and allows for a wider range of operating conditions. Catalytic distillation is receiving increasing attention and holds a huge potential for the recovery of acetic acid. Through the application of catalytic distillation via the reaction of acetic acid with iso-amyl alcohol, a useful ester in the form of iso-amyl acetate could be produced.In the present work towards further process development, the synthesis of iso-amyl acetate via reactive distillation is studied using Katamax® catalyst packing in the catalytic reactive section. The reactive distillation experiments were carried out at laboratory scale in a 50 mm diameter column with a catalytic packing section of 1 m and non-reactive packed enriching and stripping sections of about 1 m each. A cation exchange resin catalyst, Purolite® CT-175, was used. The experiments were conducted with the aim of achieving an optimum column configuration as well as process conditions for the synthesis of iso-amyl acetate in a reactive distillation column (RDC). Several variants of the RDC set-up e.g. total feed mole ratio, reflux ratio, location of feed points, reflux configuration and acid concentrations were explored for the recovery of dilute acetic acid and to achieve a high purity value added product, iso-amyl acetate.

In this paper experimental results are presented for a demonstration unit of a recently proposed novel integrated reactor concept (Smit et. al., 2005) for the partial oxidation of natural gas to syngas (POM), namely a Reverse Flow Catalytic Membrane Reactor (RFCMR). Natural gas has great potential as a feedstock for the production of liquid fuels via the Gas-To-Liquid (GTL) process, but this process has not found widespread application yet, mainly due to the large costs associated with cryogenic air separation and complex heat integration. In conventional GTL processes excess O2 (20-40 %) is used together with preheating of the feed (250-400 °C). The O2 consumption and heat integration cost can be reduced substantially by integrating the recuperative heat exchange inside the POM reactor using the reverse flow concept. The RFCMR concept basically consists of two fixed bed compartments (e.g. in a shell-and-tube configuration) separated by a porous membrane (or filter), through which the O2 is fed distributively to the syngas compartment, thereby avoiding possibly explosive feed mixtures and hot spots. Furthermore, the flow directions of the gas streams are periodically alternated. A small amount of CH4 is combusted in the O2 compartment to create the trapezoidal temperature profile. Also some steam is added to the O2 feed to keep the center of the reactor isothermal.To demonstrate the RFCMR concept an experimental set-up was constructed with a single shell-and-tube design, from which axial temperature profiles and the composition of the produced syngas could be measured. This set-up was first operated as a conventional reverse flow reactor and it was found that radial heat losses have a major influence on the axial temperature profiles as expected, but that suitable temperature profiles could be established. Subsequently, the demonstration unit was operated as a reverse flow catalytic membrane reactor. Syngas with high CO (93 %) and H2 (96 %) selectivities with high CH4 (85 %) conversions was produced from undiluted CH4 feed. These selectivities are higher than the typically encountered values of 90 % in industrial practice, because of the lower O2/CH4 ratio, and could be improved even further by going to higher temperatures, but this was not possible in this study due to mechanical constraints. The temperature plateau was flat in the center of the reactor and no hot spots were observed. The experiments have clearly demonstrated the potential of the RFCMR concept for energy efficient production of syngas.

A comparative study has been made on the hydrodynamics and gas-liquid mass transfer performances of upflow and downflow bubble columns. Gas-liquid distributions in the column are made by an ejector system using liquid as motive fluid and gas is used as secondary fluid for both the cases. The two-phase hydrodynamic studies have been done with water and aqueous solutions of carboxy methyl cellulose at different concentrations and air as gas phase. The volumetric mass transfer coefficients have been measured by absorbing air-carbon dioxide mixture in aqueous solutions of sodium carbonate and bicarbonate. For downflow system, experiments are carried out at natural suction of air through the ejector; whereas for upflow system, air is forced through the secondary entrance of the ejector from a high pressure tank at any desired rate. The effect of gas and liquid flowrates on gas holdup, frictional pressure drop and mass transfer coefficients have been investigated. All the experiments have been performed in the homogeneous flow regime for a stable operation.

The kinetics of acetylene hydrogenation has been studied in a fixed bed reactor of a commercial Pd/Al2O3 catalyst. The experiments were carried out at 30, 50 and 70 ºC with various feed compositions at atmospheric pressure. The experiments were repeated at 70 ºC in the presence of the used catalyst to determine the effect of the catalyst deactivation where the corresponding deactivation rate constant was determined in order to predict the activity of the catalyst during each run. Two well known kinetic models were used for a nearly similar catalyst to predict the experimental data of this work and none of them were found satisfactory. A new model was then proposed to fit the experimental data. The hydrogenation reactor was also simulated at industrial operating conditions with the proposed kinetics for both plug and dispersion flows. The results of these simulations were almost close to each other in most cases.

Water is an environmentally friendly chemical species that can be employed either as catalyst or as reactant. The use of water has attracted much interest for many acid-catalyzed organic syntheses. These reactions have been studied under hot water conditions with maximum formation of ions. Under these circumstances, long reaction times are required to obtain suitable yields however with poor selectivities given the influence of side reactions. The microreaction system of the present study can heat up a substrate solution very quickly from ambient conditions to the supercritical water state (scH2O) and can then quench rapidly the same solution to sufficiently low temperatures. This allows epsilon-caprolactam production with close to 100% selectivity and 80% yield in a acid free reaction system in less than 1-second reaction time. To justify these results, first principle molecular dynamics calculations and in situ IR measurements are employed as they can be applied to the structural and electronic properties of water in the near critical point region. Close to the critical point at low fluid densities, small clusters, mainly dimers and trimers, are the dominant features with no H-bonds expected. The significant reduction of H-bond strength in H2O, peculiar to near the critical point, leads to the activation of protons or H3O+. Hence, in the scH2O microreaction system, the acidic properties of H2O at the supercritical state are considered as the likely pathway promoter which prevents the hydrolysis and the pyrolysis of cyclohexanone-oxime as well as the pyrolysis of epsilon-caprolactam. It is also observed that dilute acids at supercritical conditions further enhance the epsilon-caprolactam process with close to 100% yield and 100% selectivity.

Solid mixing in fluidized bed reactors has a great impact on the transport phenomena in the reactor. Most studies, concerning solid behavior and hydrodynamic correlations of fluidized bed reactors, have been done at ambient temperature. Industrially, however, fluidized bed reactors operate at high temperatures. The lack of studies at higher temperatures is due to difficulties associated with measuring techniques under these conditions. In extrapolating hydrodynamic parameters derived at ambient temperature to higher temperatures, only the physical property changes of gas and solid phases, such as density and viscosity are taken into consideration. On a microscopic scale, however, change of temperature strongly affects the interaction between particles, which in turn has a substantial impact on the hydrodynamics of a fluidized bed.In this study and for the first time, the Radioactive Particle Tracking (RPT) is used to investigate the effect of temperature on the fluidization of silica sand particles (Geldat-B) in a bubbling fluidized bed reactor. Experiments have been carried out at different temperatures (25-400oC) and superficial gas velocities (0.17-0.75 m/s). The effect of temperature on the global mixing is studied in conjunction with the changes found in the dynamic of the ascending and descending phases. A two-phase countercurrent back-mixing model (CCBM) was used to investigate global solid mixing at different temperatures. The wake exchange coefficient, in the CCBM model, is calculated and compared with the values obtained from different correlations. For various experiments, the exchange coefficient is found to be in the range of 0.6-1.7 sec-1. The correlations can predict the trend of the wake exchange coefficient change with temperature, but they all overestimate it. The correlations developed by Hoffmann et al. (1993) and Lim et al. (1993) were found to give a better agreement with the results at high temperatures.

In a large number of industrial semi-batch reactor applications, a single reaction takes place in the reactor and the operational objective is to compute the optimal feed rate of reactants and the optimal batch temperature that optimizes an objective function at the end of a fixed batch time. In this work, it is shown that the optimal operating policy can be split into several time intervals. Analytical expressions for the optimal input are developed for each interval. Parametric uncertainty is handled by implementing this strategy in closed-loop. This methodology is illustrated via simulation of two semi-batch reactor applications.

Industrial history is full of events related to scale-up challenges. Failure at the scale-up stage is no longer surprising. Engineers and scientists have been trying to gather all the key parameters for decades, but, unfortunately, there are still no exact and well-established rules ensuring an accurate transition from one scale to another. Even the minimum lab-scale fluidized bed is still undefined.In this work, the effect of bed diameter on gas-solid behavior is investigated in three fluidized beds, 152mm, 78mm, and 50mm in diameter, in which the conventional scale-up rules are respected. The experiments were carried out using sand and alumina particles.The results were obtained and confirmed using fiber optic techniques and radioactive particle tracking, respectively. The results show that radial solid hold-up behavior on a small bed diameter scale is completely different from that on the intermediate bed diameter scale. The radial profile of solid hold-up on a small and very small bed diameter indicates an increase from a low value near the wall to a high value at the center at different heights from the distributor. By contrast, the opposite profile is observed on the intermediate bed diameter, similar to what is usually reported in the literature.

The rate of gas-liquid mass transfer is very important in several industrial chemical engineering applications. In many multi-phase reaction systems, however, the mechanism of mass transfer is not well understood. This is for instance the case in Gas-Liquid-Solid (G-L-S) and Gas-Liquid-Liquid (G-L-L) systems. To obtain more knowledge of the mechanism of mass transfer, the mass transfer coefficient, kL, and the interfacial area, a, should be studied separately. In this work an ultrasonic measurement technique is used to study the local interfacial area in a standard sized vessel, equipped with a Rushton type impeller. This is done in combination with experimental determination of the volumetric mass transfer coefficient, kLa, using the dynamic oxygen method, to obtain values for kL. The gas hold-up is determined additionally to obtain values for the Sauter mean bubble diameter at different positions in the vessel. In a coalescing air-water system the bubble size was non-uniform throughout the vessel and increased from small bubbles at the impeller along with the flow pattern to larger sizes in the bulk of the vessel. In a non-coalescing electrolyte system the vessel was much more uniform and the bubbles were smaller when compared to the air-water system. To obtain overall values of the mass transfer parameters the local values were integrated according to their volume fraction in the reactor. In both coalescing and non-coalescing systems the overall values for the mass transfer parameters were in good agreement with literature correlations. The addition of small volume-fractions of toluene to an air-water system caused a strong decrease in both the volumetric mass transfer coefficient and in the gas hold-up. The interfacial area increased, however, but it was shown that this was due to the presence of microbubbles in the solution, which do not take part in the mass transfer process. The enhancing effect on gas-liquid mass transfer due to the addition of larger volume-fractions of toluene could be described reasonably well by a homogeneous model of the shuttle mechanism.

The equilibrium gas solubility (C*), gas-holdup (eG), Sauter mean bubble diameter (dS), volumetric mass transfer coefficient (kLa), gas-liquid interfacial area (a) and mass transfer coefficient (kL) of N2, O2 and air were measured in an agitated reactor operating in surface-aeration (SAR), gas-inducing (GIR) or gas-sparging (GSR) modes in pure toluene and three mixtures of organic liquids (toluene-benzoic acid-benzaldehyde) aimed at simulating the continuous liquid phase toluene oxidation (LPTO) under wide ranges of temperatures (300-453K), pressures (1-15 bar), mixing speeds (13.3-20.0 Hz), superficial gas velocities (0.000-0.004 m/s in the GSR) and liquid heights (0.171-0.268m in the SAR and GIR).C* values of the gases in the organic liquids were calculated using a modified Peng-Robinson Equation-of-State and kLa data were determined using the Transient Physical Absorption technique. The bubble size distributions as well as dS were obtained from the Photographic method, and eG values were measured through the Dispersion Height technique using the reactor’s Jerguson windows. From eG, dS and kLa experimental values, a and kL were calculated under various operating conditions. The Central Composite Statistical Design and analysis technique was used to study the effect of operating conditions on the hydrodynamic and mass transfer parameters.At constant temperature, the equilibrium solubilities (C*) of the three gases in all liquids used appeared to increase linearly with pressure and obey Henry’s Law, however, the values exhibited minima with increasing temperature. The C* values were found to increase with increasing gas molecular weight, and decrease with the addition of benzaldehyde and benzoic acid to pure toluene. A dimensionless form of Arrhenius-type equation, in which the activation energy was dependent of temperature, was developed to predict Henry’s law constant for the three gases in toluene and mixtures with a regression coefficient > 99%.In the SAR, increasing N, T or decreasing H increased eG, a, kL and kLa, and decreased dS, whereas increasing P, decreased eG, a, kL and kLa. In the GIR, increasing N or decreasing H increased eG, a, kL, kLa and dS. Also, increasing T increased and then decreased eG and a; increased kL and kLa; and decreased dS. In addition, increasing P did not affect these hydrodynamic and mass transfer parameters under the operating conditions used. In the GSR, increasing N, T and UG increased eG, a, kL and kLa. Also, increasing N and T, or decreasing UG decreased dS.The addition of benzaldehyde and benzoic acid to pure toluene was found to significantly affect the hydrodynamic parameters (dS and eG), in the GSR and GIR, especially at low temperature due to formation of froth, which led to the enhancement of kLa. The hydrodynamic and mass transfer parameters obtained indicated that the behavior of the SAR was mainly dependent on kL, whereas those of the GSR and GIR were strongly affected not only by kL, but also by a. Statistical correlations were also developed to predict the hydrodynamic and mass transfer parameters obtained in this study with confidence levels > 95%. These correlations could be used to model, design and scale-up the LPTO process in agitated reactors.

Numerical and experimental studies of the dynamics of bubbles coalescence is reported. The numerical method based on the PLIC-VOF method and our advection and surface tension algorithms has been applied to analyse the effect of surface tension and convective terms of the Navier-Stokes equations on air bubbles' coalescence in a quiescent water column. Experimental and numerical results show a mechanical attraction between both bubbles' interface before the coalescence. We analyse the effect of the type of discretization of the convective terms of the Navier-Stokes equations on the shape of bubble after the coalescence phenomenon. Numerical results are in agreement with the images of bubbles' coalescence of our experimental device recorded by a camera. Somes images show obviousness bubbles' coalescence caracterized by an intense liquid jet within the following bubble.

Sectionalized bubble columns are finding a wide use in the case of FT synthesis, petroleum refining, wastewater treatments, extraction, absorption, leaching, ion exchange etc., due to their superiority in terms of reduced liquid phase back-mixing. The present work covers a mixing aspect in a 0.41 m id. sectionalized bubble column over a wide range of superficial gas velocity, liquid height to column diameter ratio, percent free area of the sectionalizing plates and electrolyte concentration for Air-water system. The acquired conductivity data was optimized using a compartmental and axial dispersion models and the model parameters have been estimated. For a given Hc/D ratio; mixing time was found to decrease with an increase in gas superficial velocity and free area of the sectionalizing plate. The compartmental model successfully predicts the tracer concentration profile and the longitudinal dispersion coefficient. The mixing time in a sectionalized bubble column was also compared with a non-sectionalized bubble column and was found to be order of magnitude higher.

The alkylation of bromobenzene with allyl acetate on zeolite H-USY (Si/Al 30) was studied in liquid phase conditions, at 150 °C. Reactions were performed at different space velocities (LHSV = 60, 120, 240 and 480 h-1) in tubular plug flow reactors using a high-throughput frontal analysis setup. Besides the desired alkylation products, cis-propenyl bromobenzene (cis-PBB) and trans-propenyl bromobenzene (trans-PBB) and a dialkylated product, light and heavy side products were formed. Polymerization reactions starting from allyl acetate lead to blocking of the zeolite pores, and result in catalyst deactivation. Based on the experimental data, several molecular kinetic models were proposed, and implemented in a reactor model, accounting for axial dispersion, mass transfer into the catalyst crystals, competitive adsorption effects and catalyst deactivation. Model discrimination revealed the underlying reaction mechanisms.

Consumers demand for portable audio/video/ICT devices has driven the development of advanced power technologies in recent years. Fuel cells are a clean technology with low emissions levels, able to work with renewable fuels and capable, in a next future, to replace conventional power systems meeting the targets of the Kyoto Protocol for a society based on sustainable energy systems. Within such a perspective, the objective of the European project MOREPOWER (Compact direct methanol fuel cells for portable applications) is to develop a low cost, low temperature, portable Direct Methanol Fuel Cell (DMFC; nominal power 250 W) of compact construction and modular design for the potential market area of weather stations, medical devices, signal units, gas sensors and security cameras. This investigation is focused on a conceptual study of the DMFC system carried out in the Matlab/Simulink platform. Two different DMFC configurations were devised in which alternative components arrangements along the methanol recycle line are considered. Based on a number of simulations, the system configuration characterized by a gas-liquid separator for carbon dioxide removal placed upstream the radiator for heat removal shows the most promising results, thanks to an easier design and operability of the radiator itself.

Citral was hydrogenated over Ni on alumina wash-coated cordierite monoliths in a screw impeller batch reactor. The prepared catalyst was characterized by means of nitrogen adsorption (BET), X-ray diffraction (XRD) and scanning electron microscopy (SEM). The performance of the monolith reactor configuration at different temperatures and pressures was compared. Reasonable reaction rate was obtained even after 18 consecutive experiments without regeneration of the catalyst. At temperatures up to 80°C and 5 bar pressure the selectivity to citronellal was ca. 93 %. Increasing the temperature to 100°C decreased the selectivity of citronellal to 65 %. At 100°C and 40 bar pressure the selectivity of citronellal was diminished to 1.4 % due to the formation of the fully hydrogenated product 3,7-dimethyloctanol with a selectivity of 93 %.

The adsorption of C5H8 on Pt(111) has been studied using a semiempirical theoretical method implemented with the YAeHMOP package. The energetic calculations show that the adsorption is weak. The adsorbate-surface interaction requires that both C-H and Pt-Pt bonds be weaken. The main interaction comes from the C and H nearest to the metal surface. The tilted geometrical configuration adopted by the molecule facilitates this interaction. The Pt-C and Pt-H interactions are mainly due to overlaps between the dz2 and pz metal orbitals and the C pz and H s orbitals, respectively.

The effect of different solvents for model hydrodesulfurization (HDS) reactions on both the kinetics and the mass transfer resistances on a wide range of hydrogen partial pressures has been investigated. Special attention was directed to study the effect of low hydrogen partial pressures. While some researchers associate solvent effects to competitive adsorption of the solvent molecules on the catalytic surface, there is evidence that other effects may also play an important role in the modification of the kinetics. In this work, HDS experiments were performed in a slurry stirred tank reactor using the model molecule 4,6-dimethyldibenzothiophene (4,6-DMDBT) over a commercial NiMoP/alumina catalyst. The effect of hydrogen pressure on the transformation of 4,6 DMDBT in different solvents was studied. Thermodynamic calculations for hydrogen solubility under the whole set of conditions was considered. From our experiments, modification of the catalyst activity was observed to occur only at low hydrogen pressures (less than 34.7 bar) when n-dodecane was the used solvent. On the other hand, when using decaline there seems to be practically no effect in the catalytic activity of the hydrogen pressure over the entire range of pressures that were studied (18.0-55.5 bar). Additionally, the selectivity to direct desulfurization (DDS) products increased, while simultaneously decreasing the selectivity to the hydrogenation (HYD) pathway products, as the hydrogen partial pressure diminished below 34.7 bar when both solvents were used. It was concluded that hydrogen solubility effects become important and modify the activity at pressures below 34.7 bar.

Benzothiophene, Dibenzothiophene (DBT), 4-methyldibenzothiophene, and 4,6-dimethyldibenzothiophene are typical thiophenic compounds prevailing in diesel fuels. Using a hexadecane solution of these model compounds, experiments were carried out to compare the reactivity of the different benzothiophenes in oxidesulfurization (ODS) with several polar solvents. V2O5/Al2O3 catalyst, hydrogen peroxide as oxidant and the solvents: N,N-dimethylformamide (DMF), 2-ethoxyethanol (EEOH), acetonitrile (MeCN) and gamma-butyrolactone (BuL) were used. The experimental results showed that removal of benzothiophenes contained in diesel can be carried out efficiently in conditions here studied, with a yield to sulfones higher than 80%, depending on the reactivity of each S-compound and solvent used. DBT was completely removed but the yield to DBT sulfone with MeCN was only 95%. According to the solvent used, benzothiophenes reactivities decrease in the order: MeCN > BuL > EEOH > DMF. It was found that an important fraction of the eliminated compounds in the diesel phase, is not totally transformed to its corresponding sulfone, it is only removed as sulfur compound by extraction, without undergoing ODS reaction. Depending on the thiophenic compound, it was possible to remove about 70-85% of benzothiophenes by physical extraction process, using the best solvent (DMF) and only close to 30-60% with MeCN.

Heterogeneous kinetics of catalytic wet air oxidation (CWAO) of phenol was studied using copper supported on activated carbon as the catalyst and oxygen as the oxidant. Both mass transfer effect and deactivation of catalyst can affect the kinetics measurement. To eliminate the mass transfer barriers, the study was conducted in a stirred tank batch reactor using catalysts of very small particle size. In order to prevent the catalyst deactivation caused by copper leaching from the support, saturated pH buffer solution was added into the reaction system and the initial phenol and catalyst concentrations were reduced. A mathematic model was proposed following the Langmuir-Hinshelwood mechanism to simulate the reaction kinetics. It is believed through simulation that the dissolved oxygen is adsorbed by the catalyst in molecular form which subsequently starts to degrade phenol on the catalyst surface. Both reaction rate constant and activation energy were determined for a temperature range of 140 - 160 oC and an oxygen partial pressure range of 0.4 - 1.6 MPa. The proposed mechanism and kinetic model were found to be in good agreement with experimental results.

Esterification reactions are typically equilibrium limited, and face challenges with product purification. They are carried out commercially using either large excess of one of the reactants, or by removing one of the products continuously in a reactive distillation column (RDC). In the present work, we studied the kinetics of esterification reactions of dilute acetic acid with iso-amyl alcohol to produce a value added ester (iso-amyl acetate) using ion-exchange resins, Purolite® CT-175 and CT-275 as catalysts in a batch reactor. The effect of agitation speed, catalyst particle size, reaction temperature, acetic acid concentration, acetic acid to iso-amyl alcohol molar ratio and catalyst loadings were investigated to optimise the reaction conditions. The non-ideality of each species in the reacting mixture was accounted for by using the activity coefficient via the use of the UNIFAC group contribution method. The kinetic data were correlated with Langmuir-Hinshelwood-Hougen-Watson (LHHW) and Eley-Rideal (E-R) models. The detailed kinetic data and a reliable rate expression for the esterification of acetic acid with an aliphatic alcohol (iso-amyl alcohol) would be useful for the simulation and design of an RDC for removing dilute acetic acid from aqueous streams. Another objective of this work was to evaluate the techno-feasibility of this operation through a systematic procedure of residue curve map (RCM). RCM gives a design engineer an idea of the existence of any separation boundaries imposed by the singular points corresponding to the reactive or kinetic azeotropes and thereby provide an insight into the feasibility of the desired operation. In the present study, RCM was generated for the quaternary system (acetic acid - iso-amyl alcohol - iso-amyl acetate - water) under different conditions to work out the feasibility of the operation. This map also suggests possible column sequencing or configurations to achieve a desired duty in an RDC.

Ni/SiO2, NiMo/Al2O3 and CoMo/Al2O3 were prepared by microemulsion method. Precipitation and sol-gel of the support precursor allowed the synthesis of catalysts with differences in textural, structural and catalytic properties. The main difference between the precipitation and sol-gel microemulsion routes is the role of surfactant (Cetyltrimethyl Ammonium Bromide) in each synthesis method. The aggregation mechanism of support is modified by the synthesis route. There are differences in the catalytic performance of Ni/SiO2 microemulsion catalysts by the precipitation or sol-gel of support precursor. The hydrogenation of benzene, a low structure-sensitive reaction, showed that the precipitation 9%Ni/SiO2 microemulsion catalyst is more active than the sol-gel 9%Ni/SiO2 microemulsion catalyst. On other hand, these catalysts showed opposite result in the hydrogenation of C=C double bond of crotonaldehyde molecule. These results are indicative of the effects of nanostructured support and nanostructured active-phase in the hydrogenation activity. NiMo/Al2O3 and CoMo/Al2O3 catalysts prepared by sol-gel-microemulsion method were used in the hydrodesulphurization (HDS) of dibenzothiophene and 4,6-dimethyldibenzothiophene. The HDS activity was improved as a function of metal content and it motivates the study of NiMo and CoMo microemulsion catalysts with higher contents of metals to enhance the HDS of 4,6-dimethyldibenzothiophene.

The effect of Al2O3-ZrO2, Al2O3-TiO2 and Al2O3-La2O3 mixed oxides on the deactivation of bifunctional Pt-Sn/ Al2O3 reforming catalysts has been investigated. The n-heptane reforming at 500°C was used as a test reaction. Changes in the catalytic behavior due to differences of the acidity and the support-metal interaction were observed. Levenspiel´s and Beltramini´s deactivation models were developed assuming a series fouling for the carbonaceous deposits. Through the Beltramini’s model it was possible to distinguish the amount of acidity that participates in the deactivation processes. Both models successfully correlated with the experimental profiles of n-heptane activity decay. The following deactivation decreasing order, Al2O3-TiO2-1>Al2O3-ZrO2-25>Al2O3-TiO2-2> Al2O3>Al2O3-La2O3-10 was found with both deactivation models. A quasi-linear correlation between the deactivation order and the coke formation kinetic constant (Levenspiel’s parameters) was observed. The catalyst acidity and the n-heptane conversion were correlated with Beltramini’s model. It was found that a high acidity (12 X 10-17 acid sites/g cat.) or metal dispersion (83%) increases the catalyst deactivation and it is necessary to have a balance of active sites in order to have a catalyst working as a bifunctional catalyst. On the other hand, the auto-regeneration Beltramini’s parameter suggests that the lowest deactivation of the Pt-Sn/ Al2O3-La2O3-10 catalyst is attributed to the cleaning capacity of the active sites. It was observed that the highest deactivation (80-92%) of the platinum-tin catalysts supported in alumina-titania mixed oxides were a result of the strong metal-support interaction (SMSI) effect. The Pt-Sn/Al2O3-La2O3-10 showed the best catalytic behavior with high initial and residual conversions (70 and 48%, respectively) and low deactivation (17 %) at a 50-minute reaction time. Furthermore, in the Pt-Sn/Al2O3 catalyst, the benzene yield was 1%, while the Pt-Sn/Al2O3-La2O3-10 showed a total inhibition of benzene production yield at residual conversions.

Novel proposed processes for H2 production and energy generation such as partial oxidation of hydrocarbons (POX-MeO) and chemical looping process (CLP), respectively require the use of solid oxides as oxygen carriers. In POX-MeO the required oxygen for the partial oxidation of methane is provided by a transition metal oxide (MeO). First, H2 is produced through CH4+MeO = CO+H2+Me. Secondly, Me is re-oxidized through Me+O2 = MeO to regenerate the oxygen carrier. In the CL process, CH4 is being completely oxidized through CH2 + MeO = CO2 + H2O + Me producing heat and CO2 ready for sequestration. Finally, Me is re-oxidized using air to regenerate the Me back to MeO. In both processes the regenerated MeO is sent back to the initial step to result in a cyclic operation. Continuous exposure of MeO to Redox cycles frequently produces sinterization and MeO stabilization is needed to avoid loss of activity. The objective of this study is to investigate the stabilization effect of TiO2 in Co3O4 during Redox cycles to be used as an oxygen carrier using CoxTiOy type spinnels. Characterization of the synthesized samples included XRD, TPR, and SEM. Co2TiO4 and CoTiO3 spinnels were synthesized by solid state reaction. TGA and TPR Redox performance cycles of Co3O4 produced sintering, while results using a Co2TiO4 spinnel structure suggest a strong stabilization effect of TiO2 on Co. Ten Redox cycles using H2 and CH4 as reducing agents and a mixture of O2/N2 as oxidizer resulted in fixation of Co to TiO2 avoiding sintering.

The catalytic hydrogenation of adiponitrile is generally carried out industrially at high temperature and pressure over Ni-Raney catalysts, which have the disadvantages of possessing low mechanical resistance and of being pyrophoric. In this work adiponitrile was hydrogenated in a three phase stirred slurry reactor using rhodium over alumina powder as a catalyst instead. Catalysts were prepared with a ion exchange technique by contacting 100 micron alumina powder with a solution of hydrate rhodium cloride for 36 h. The reaction runs were conducted in a 100 cc vessel at a pressure of 3 MPa and temperatures of 72 – 92 °C varying the initial concentration of adiponitrile. A variety of series-parallel reactions took place when adiponitrile was contacted with hydrogen in our reactor. Attention was concentrated on hexamethylenediamine, the final product, and on aminocapronitrile, a very interesting intermediate which could be used subsequently for caprolactam production (and from this the direct synthesis of nylon). Samples were withdrawn from the reactor at fixed interval of time and analysed with a gas-chromatograph so that time dependency of reactant conversion and product selectivity could be determined. The results showed the expected qualitative behaviour: adiponitrile conversion increased with time, as did the hexamethylenediamine selectivity, whereas aminocapronitrile selectivity showed a clear maximum and then decreased to negligible values. The “Chemical reaction engineering” approach was used and proved sufficient to describe quantitatively the observed kinetic behaviour of the reactor.

In the metallurgical industry, various types of rotors are used for the injection and distribution of gas and for homogenizing molten metal. In the present work, the liquid-gas two-phase flow around an axial type impeller is studied in a water model, in order to analyze the bubble break-up and coalescence and metal mixing. Details like primary and secondary vortex structure, gas flooding between the blades and gas dispersion are recorded by using high speed photography.A mathematical model that takes into account the combined effect of bubble break-up and coalescence is implemented in the commercial computational fluid dynamics (CFD) software FLUENT. In the proposed work, the impeller is explicitly described in three dimensions using Multiple Reference Frame Model. Dispersed gas and bubbles dynamics in the turbulent water are modeled using an Eulerian-Eulerian approach with dispersed k-epsilon turbulent model. The model predicts spatial distribution of gas hold-up, average bubble size and flow structure. Good qualitative agreement between physical model and simulation is achieved when comparing the bubble size distribution, flow structure and mixing.

The numerical investigation of interphase heat transfer in a circular channel under laminar flow regime at different values of the Pe number is performed by means of a 2D mathematical axisymmetric model in which energy equations are coupled to the Navier-Stokes equations. The constant wall heat flux and constant wall temperature cases are simulated at different values of the Pe number. Results show that the heat transfer efficiency in both cases is strongly influenced by the back-diffusion even at very high values of Pe (Pe > 700). This behaviour is shown to be the result of the coupling between the interphase heat transport phenomena and the fluid flow. New correlations are hence derived to take into account the effect of back-diffusion on heat transfer.

An approximate solution for the problem of diffusion-reaction in an isothermal catalyst pellet when coke deposition occurs was developed. The approximation is based on the linearization of reaction terms, producing analytical expressions to evaluate concentration profiles inside the pellet and the effectiveness factor. The approximate solution can be extended to any form of reaction rate functionalities for the main and coking reactions. Coke formation from reactants (parallel coking) or products (consecutive coking) and deactivation functions dependent on the coke content are considered in pellets with slab, cylindrical and spherical geometry.In this work, first-order kinetics for the main and coking reactions and linear and exponential deactivation functions were evaluated. The approximate predictions of the concentration profiles and the effectiveness factors were compared with results obtained by the complete numerical solution. A good agreement was found for both parallel and consecutive coking, including the cases when low and high intra pellet diffusion resistances are present.

Gasification of biomass is an environmentally important technology that may contribute to fulfill requirements set by the Kyoto protocol. Biomass can be converted into a vast array of chemical products and fuel, and can be utilized to produce power/electricity. However, a major limitation of biomass gasification is the resultant tars and particulate matter that potentially destroy downstream process equipment, harm the environment, and hinder economic efficiency. This study considers the catalytic steam gasification of waste-wood, with the catalyst being fluidizable nickel/a-alumina. Experiments are conducted in the CREC Riser Simulator Reactor, at temperatures between 800°C - 850°C, near atmospheric pressures, and reaction times over 10 seconds. Experimental results suggest that catalytic steam gasification of biomass is a versatile process; with a considerable amount of H2 being produced at steam/biomass feed ratios above 0.3kg/kg daf. It is also found that the catalyst effectively converts tars to permanent gases at temperatures ranging from 800°C - 850°C and for steam/biomass feed ratio ranging from 0.17 – 0.58kg/kg daf, with the carbon conversion being above 90% in all cases. An equilibrium model from the literature is developed and modified, and then considered to analyze the experimental data. At 800°C and steam/biomass ratios from 0.17 – 0.42kg/kg daf, the H2/CO product ratio is estimated accurately by the equilibrium model. However, product lump compositions are estimated with less accuracy, which provides support to the need of a non-equilibrium model to fully explain the inter-conversion of gaseous species following biomass catalytic gasification.

A common form of chemical reactor is the bubbling fluidised bed where gas is introduced into a free bed of particles at a rate exceeding that necessary to support their weight. The behaviour of the reactor is dependent on the distribution and characteristics of the bubbles within the bed. Information about these can be inferred from pressure measurements within the bed, which can also be used to characterise the dynamics of a bed and therefore to control it and the reaction.It is necessary to understand what the pressure signals are measuring, how many of them are necessary, and where they should be placed for them to be used to characterise the state of the bed. Pressure measurements are made at single points within a bed, and it is not immediately clear whether the measurements are local (owing to some sort of wall effect, for example), characterise the region close to the sensor, or the whole bed. It is shown how the dynamics of the bed can be characterised from the pressure signals, and that these signals are sensitive to the state of the bed. It is also necessary to evaluate where the sensors should be placed and how many of them should be used. Important considerations here are the extent to which the measurements of pressure are axisymmetrical, and whether in some circumstances and positions the local dynamics mask the overall bed dynamics. Axisymmetry is shown to exist under many conditions and that a single transducer can be sufficient. Furthermore, the depth of the bed is also shown to have a significant effect on its dynamics.It is also shown how the characterised pressure measurements can be used to control the state of the fluidised bed and used to enhance the performance of a fluidised bed reactor.

In this paper the applicability of using activities instead of concentrations in kinetic expressions has been investigated using the reaction of CO2 with various hydroxide solutions as a model system. For this system it is known that when the reaction rate constant is based on the use of concentrations in the kinetic expression, this "constant" depends both on the counter ion in the solution and the ionic strength. Using the data of Pohorecki, it could be revealed that the use of activities reduces the influence of the ionic strength on the derived rate constant to a very large extent. Hence the kinetic constant using the activity based approach seems much more like a real constant than the "constant" using the traditional kinetic expression using concentrations. Besides this, the absolute values of the rate constants for the three different counter-ions also moved closely together. This points out that applying one fixed rate constant, independent of the counter-ion or the ionic strength, might be sufficient for describing the reaction rate in the studied system. A large advantage of using activities in kinetic expressions is that the fundamental description of kinetics and thermodynamics is consistent which each other so that at equilibrium the net reaction rate (based on the kinetic expression) is indeed zero.

Fluidized Bed Membrane Reactors (FBMR) offer significant advantages for steam reforming and the production of hydrogen. Potential advantages include higher yields by reducing thermodynamic equilibrium limitations, process intensification by combining three vessels into one, reduced temperatures of operation, countering the adverse effects of pressure, virtually eliminating catalyst diffusional limitations, high productivity per unit volume of reformer, and flexibility in using alternative feedstocks. Realization of the FBMR process for hydrogen production requires that a number of unusual challenges in reactor design be met. This paper discusses the technical challenges and outlines key factors which are being addressed in providing the membranes, reactor configuration and integrity, catalyst, energy integration and operating conditions needed to establish an economically viable FBMR process.

The current solid acid catalyst based alkylation processes suffer from rapid deactivation of the catalysts which is affected by the shape and size of the catalysts, reactor configuration and the reactants’ feeding pattern. In this work, the deactivation of the solid acid catalysts in alkylation of iso-butane with butene and the reactor performance are studied using the multi-scale mathematical models. A pseudo-steady state catalyst pellet level model, incorporating the alkylation and deactivation kinetics, is coupled to the transient model of back mixed reactor configurations like a single CSTR and CSTR-in-series. The effects of catalyst shape, initial acid strength of the catalyst, paraffin-to-olefin feed ratio on the conversion of the olefin, on the yield of alkylates and on the time required for complete deactivation of the catalyst are investigated. The role of the size of the catalyst particles on the types of deactivation mechanism is elucidated. The effect of contacting pattern of iso-butane and butene in the reactor on the catalyst life and on the alkylate yield is also presented.

The purpose of this work is to verify the possibility of using a simple physical adsorption technique to improve enzyme efficiency in performing lactose hydrolysis. Immobilization experiments of ß -galactosidase on zeolite pellets showed that this process is fairly slow and can be accomplished with contact times shorter than few hours. Comparison between homogeneous and heterogeneous hydrolysis experiments in mixed batch reactors indicates that the enzyme distribution within the zeolite pellets is concentrated in a thin shell under the pellet surface. Continuous conversion experiments were carried out on a purposely-built fluidization column. Results were fitted with a mathematical model for fluidized bed conversion without the use of any further adjustable parameter than those used in the heterogeneous batch conversion. Furthermore, no change of the parameter values was necessary. Scale-up predictions given by the model showed that fluidized columns few meters high of pellets supporting b-galactosidase are sufficient for an industrially suitable process of lactose hydrolysis. which can be performed on columns of few meters and, therefore, these are applicable to relatively small process scales as those found in traditional Italian cheese making factories.

The effects of feed composition on the conversion of CO and CO2, selectivity to DME, yield of DME and product distribution are studied in the synthesis of DME from syngas and from (H2+CO2) in a single reaction step using the bifunctional catalysts CuO-ZnO-Al2O3/gamma-Al2O3 and CuO-ZnO-Al2O3/NaHZSM-5. For feeds made up of H2 and CO, the maximum conversion levels of CO, yield of DME and selectivity to DME obtained with both catalysts are very similar, close to 95 %, 75 % and 80 %, respectively, although the catalyst CuO-ZnO-Al2O3/gamma-Al2O3 requires a higher H2/CO ratio. For the feeds made up of H2 and CO2, the catalyst based on NaHZSM-5 zeolite (moderately acid function) performs better due to the fact that higher water content in the reaction medium leads to its adsorption on the Lewis acid sites of gamma-Al2O3 and limits the transformation capacity of methanol to DME.

Indoor air pollution by microbial contaminants is increasingly receiving attention as a public health problem. Under a suitable environment, such as in heating, ventilation and air conditioning (HVAC) system, airborne bacteria are able to proliferate and grow causing various allergies and illnesses. This can be particularly serious in tropical regions due to high relative humidity and warm temperatures all round the year. Application of photocatalysis using UV-A and TiO2 to inactivate air-borne bacteria is relatively new and systematic parametric study is required for the engineering design of a process based on this technology. This study investigates the effects of TiO2 mediated inactivation of various bacterial species in batch and continuous systems using different TiO2 loadings and radiation intensities. Gram-negative bacteria, E. coli and two Gram-positive bacteria, Microbacterium sp. and Bacillus subtilis were used for the inactivation studies. In both systems, inactivation rates of Gram-negative E. coli are higher than the Gram-positive Bacillus subtilis and Microbacterium sp. and the inactivation rates increased in presence of TiO2 for all bacteria. Depending on the type of bacteria, TiO2 loading and light intensity, an increase of 1.3-5.8 times in the inactivation rates was obtained from those in the absence of TiO2. The inactivation rates in the batch and continuous systems were reasonably comparable. Inactivation rates in the continuous system are somewhat higher than those in the batch system due to the unaccounted loss of bacteria via adsorption and settling on the reactor walls in the flow system. The study demonstrates an approach that can be used for the designing of large scale systems for the treatment of bioaerosol.

Historically, the priority in disinfection reactor design has been microbial inactivation with a high factor of safety, neglecting other processes such as disinfection byproduct (DBP) formation and chemical use minimization. With heightened awareness and regulation of DBPs, advanced chemical engineering principles and practices are being employed in disinfection process analysis and design in order to minimize DBP production while ensuring sufficient microbial inactivation. Computational fluid dynamics (CFD) is increasingly popular as a tool to evaluate reactor performance and assess design concepts. This study presents the results of CFD analyses of two disinfection reactors with very different aspect ratios and design, but being characterized by the same theoretical hydraulic residence time (i.e., 30 minutes). Both reactors were used to inactivate E. coli with the novel disinfectant peroxyacetic acid (PAA). CFD calculations were performed for inactivation of E. coli (10000 CFU/100 mL) with PAA concentrations of 4 and 8 mg/L. Calculations were carried out for two serpentine reactors with aspect ratios (channel width to water depth) of 2:1 and 1:3. In the second reactor, perforated baffle plates were placed at the end of each channel in an attempt to promote uniform flow. Both reactors achieved adequate removal of organisms, but displayed non-ideal hydrodynamics, with significant recirculation regions that reduced the volume available for reaction and promoted short-circuiting. CFD simulations emphasized that reactor hydrodynamics impacts on the specific disinfection performances inducing different ratio of reactor volume active during the disinfection process.

With over 300 units in operation, Fluid Catalytic Cracking (FCC) is currently the dominant refinery conversion process. Over the last years, rapid separation of hydrocarbon vapors and catalyst at the riser outlet has been put forward to limit non-selective decomposition of reaction products, thus enhancing product selectivities.A new riser separator system, called RS², was developed to minimize post-riser cracking, while maintaining flexibility of operation. The development included cold flow testing, CFD and pilot plant testing, and aimed at minimizing both thermal degradation and catalytic overcracking. Cold flow testing was conducted at a representative scale (scale 1/10th) on a large cold flow mock-up. Testing various riser separator configurations provided a thorough understanding, both through visual observations, pressure drop measurements and tracer studies. Hot pilot plant testing was conducted to investigate the effect on yields of the riser separation system in the reactor-stripper under adiabatic operation as a function of operating conditions. The separator had a significant impact on the FCC heat balance and a large reduction of dry gas yield was observed.The resulting technology was successfully implemented on an industrial resid catalytic cracking unit. Operation of the separator was proven smooth and flexible with enhanced performance, while the industrial observations confirm the R&D conclusions. This paper therefore summarizes the R&D study and the industrial operation trends. In particular, the impact of stripper bed level on operation is discussed and explained based on the flow phenomena and design aspects.

In the present investigation, the kinetics governing wet air oxidation (WAO) of organic pollutants is modeled by two distinct parallel kinetic terms corresponding to a readily oxidizable and a recalcitrant structural parts of the pollutant molecule involving the concept of thermodynamic severity factor. The developed model was successfully applied to several classes of organic water pollutants including phenolic compounds as well as mono- and di-carboxylic acids. In the offshoot of the model, a recalcitrance index was proposed classifying the pollutant molecules according to their resistance to oxidation. For the carboxylic acids, it was found that the recalcitrance towards WAO was inversely correlated to the molecular size and weight. Acetic acid proved to be the more recalcitrant component. As for the phenolic compounds, the oxidation efficiency was shown to be affected by the nature of the substitutents present on the phenolic ring. The electron donating substitutents such as alkyl groups (CH3 or –C2H5) confer to the molecule an enhanced reactivity towards WAO. On the contrary, electron-accepting substitutents, such as methoxy groups, lower the oxidation and consequently enhance the recalcitrance.

Recombinant human interleukin-11(rhIL-11) is still the only therapy drug for thrombocytopenia. Although the biological properties and clinical behaviors of rhIL-11 have been studied extensively, its bioprocess development data have rarely been reported. In this work, fermentation conditions for recombinant E. coli, which expressed interleukin-11 as a fusion protein, were investigated. Batch and fed-batch fermentation experiments were performed in shake flasks and in a BioFlo fermenter with different medium and carbon sources under different operational parameters. Based on the results of the study, M9Ca medium was shown as the better medium and glycerol was identified to be a better carbon source than glucose for the expression of rhIL-11. When glycerol was used as the carbon source in fed-batch fermentation, a dry cell weight of 9.2 g/L and a rhIL-11 expression of 22%, were achieved. The results could be used as references in the optimization of the fermentation process.

Earlier investigations (Cartland Glover et al., 2004) into the use of computational fluid dynamics (CFD) for the modelling of gas-liquid and gas-liquid-solid flow allowed a simple biochemical reaction model to be implemented. A single plane mesh was used to represent the transport and reaction of molasses, the mould Aspergillus niger and citric acid in a bubble column with a height to diameter aspect ratio of 20:1. Two specific growth rates were used to examine the impact that biomass growth had on the local solids concentration and the effect this had on the local hydrodynamics of the bubble column.

For highly exothermic catalytic reactions particular measures have to be taken for obtaining reliable kinetic and mechanistic information, which has been preferably determined under the prevalence of isothermicity, and which is not disguised by transport processes. If this is true such kinetics can be reliably extrapolated to conditions outside the frame of experimental conditions. To this end various reactor set-ups are presented and their operation is illustrated by some case studies. The temperature-scanning reactor, which is not inherently isothermal, is included since non-isothermicity is not an obstacle to easy kinetic data assessment. Its performance is clarified by estimating kinetic parameters for the ammonia synthesis at elevated pressure using a commercial catalyst. Two types of novel micro-structured reactors are presented which have been applied for two highly exothermic catalytic reactions, i.e., the oxidative dehydrogenation of propane (ODP) and the ammonia oxidation to N2, N2O and NO. These studies served the purpose of getting mechanistic insights into the reaction mechanisms of both reactions, to validate a kinetic scheme for the ODP reaction and to derive the estimates of the respective kinetic parameters. Finally, a reactor for transient experiments in the peak pressure range of 1 to 10 Pa has been used for studying reaction steps related to the catalytic ammonia oxidation under isothermal conditions. Isothermal studies are possible due to the low concentrations of reactants by which any undesired hot spots on the catalyst surface were suppressed. The instrumental arsenal described is an asset for both fundamental as well as applied work in studying exothermic catalytic reactions.

Downer reactors have many advantages over risers for reactions requiring very short residence time. However, its application for reactions where a high solid/gas ratio is required has been restricted by the low solids holdup in the fully developed region (typically less than one percent). In this paper, we present the performance of a 0.078 m diameter and 3.2 m long high-density downer reactor equipped with a recently developed novel solids feeding system.Experiments were carried out using fluid coke particles of particle density 1600 kg/m3 and mean diameter 133 mm, with solids fluxes up to 1400 kg/m2s and superficial gas velocities ranging from 0 to 6 m/s. The downer was operated under batch mode, with solids fed from a solids feeder installed inside a fluidized bed at the top of the downer and the bottom end of the downer connected to a receiving hopper. The solids flux was controlled by a butterfly valve installed 0.4 m below the downer entrance. Air was injected into the downer through four 45-degree-angled nozzles located right below the butterfly valve. Solids were lifted up from the bottom receiving-hopper to the upper fluidized bed through an external riser after each run.The cross-sectional average solids hold-up calculated from the integration of local measurements obtained from a capacitance probe was found to be a function of both the solids flux and superficial gas velocity. The solids hold-up increased with increasing solids flux at a given superficial gas velocity, but decreased with increasing gas velocity at a fixed solids flux. A cross-sectional average solids holdup of 16.5% was achieved at an axial position of 3.0 m below the air injection point under the operating condition of Gs=1400 kg/m2s and Ug=2.0 m/s. The shape of radial solids distribution varied not only with the superficial gas velocity but also with the solids flux in high-density downers. The radial solids distribution became more uniform under higher solids fluxes and lower gas velocities. The dense ring at the near wall region disappeared under the high-flux operating condition (Gs=1400 kg/m2s, Ug=2.0 m/s). The radial solids holdup profiles became less uniform, denser near the wall and more dilute and non-uniformly distributed in the central region, as the gas velocity increased from 3.0 to 6.0 m/s at a fixed solids flux of about 400 kg/m2s, which is significantly different from flow patterns reported in the literature under lower solids flux conditions.

The total gas holdup and the holdup of large gas bubbles were predicted in bubble column reactors (BCRs) and slurry bubble column rectors (SBCRs) using two Back-Propagation Neural Networks (BPNNs). Over 3880 and 1425 data points for gas holdup and Large gas bubble holdup respectively, covering wide ranges of gas-liquid-solid physical properties, operating variables, reactor geometry, and gas sparger type/size, were employed to develop, train and validate the two neural networks. The developed BPNN for gas holdup has a topology of [14,9-7,1] and was able to predict the trained and untrained data with an average absolute relative error (AARE), standard deviation, and regression coefficient (R2) of 16, 19 and 90%, respectively. The developed BPNN for large gas bubble holdup has a topology of [14,8,1] and was capable of predicting the trained and untrained data with AARE, standard deviation, and R2 of 10, 14 and 93%, respectively. The BPNNs were then used to predict the effects of pressure, superficial gas velocity, temperature and catalyst loading on the total syngas holdup for Low-Temperature Fischer-Tropsch (LTFT) synthesis carried out in a 5 m ID SBCR. The predicted total syngas holdup appeared to increase with increasing reactor pressure, superficial gas velocity and the number of orifices in the gas sparger. The predicted syngas holdup, however, was found to decrease with increasing catalyst loading and reactor temperature. Also, under similar LTFT operating conditions (P = 3 MPa, T = 513 K, CW = 30 and 50 wt%), the total syngas holdup values predicted for H2/CO ratio of 2:1 and cobalt-based catalyst are consistently lower than those obtained for H2/CO ratio of 1:1 and iron oxide catalyst in the superficial gas velocity range from 0.005 to 0.4 m/s. These predictions are in perfect agreement with reported literature trends, which underscore the reliability and validity of the developed BPNNs in predicting the total syngas holdup and the holdup of large gas bubbles in large-scale bubble columns and SBCRs operating under industrial conditions.

Homogeneous flows of air-water in two bubble columns, 0.130 m and 0.292 m in diameter, were investigated using RPT (radioactive particle tracking) and fibre optics. The homogeneity of bubble flows at U g < 0.04 m/s was revealed not only in the hydrodynamics of the gas/liquid phase but also in the transport phenomena (turbulence intensity…). RPT Lagrangian data derived that turbulence was about 0.15 m/s for the radial direction and 0.26~0.35 m/s for axial direction in the 0.292 m column, 0.05~0.16 m/s for the radial direction and 0.25~0.35 m/s for the axial direction in the 0.13 column. Average bubble diameter was 4~5.3 mm and was almost constant for the radial locations. The result was twice the sizes of the predicted bubble sizes in the CFD modeling. The slip velocity was 0.3 m/s at U g=0.015~0.033m/s as it was assumed in many reactor designs, and only a small variation was found at the radial locations. In flow patterns, stable circulation cells were revealed using velocity vector plots if a single nozzle sparger was used in the 0.13m column, which confirmed Joshi’s assumptions (Joshi and Sharma, 1979) and Lin’s assumption (Lin et al., 1996) based on 2-D columns. Using a perforated sparger was helpful to form a stable mean velocity field. Mean bubble diameters and bubble size distribution varied only slightly with radial/axial locations, which was consistent with the homogeneity of radial/axial turbulence. Bubble-size evolution from their formation to equilibrium mean diameter was discussed. Mean bubble size correlation was approached. The homogeneity of investigated flows was also substantiated by the criteria of stability for bubble flows.

Fast chemical reactions in turbulent liquid flows are studied experimentally and computationally. Experiments are carried out for reactive mixing in order to obtain local mean and fluctuating concentrations. The experimental findings are used to test the performance of numerical models for turbulence, mixing and chemical reactions.The experiments are carried out in a mixing channel with two inlets separated by a block. A reactive Planar Laser Induced Fluorescence (PLIF) technique is used to obtain instantaneous concentration fields of product from an acid-base reaction. Two different flow cases are studied. For the first flow case, there are equal flow rates of acid and base, whereas the flow rate of acid is reduced to half the flow rate of base in the second flow case. The amount of product formed is found to be strongly dependent on how well the two inlet streams are mixed.The turbulent flow is modelled in two-dimensions by solving the Reynolds Averaged Navier-Stokes equations in combination with a standard k-? turbulence model. Two different micro-mixing models are adopted, namely the multiple-time-scale turbulent mixer model (MTS) and the multi-peak presumed PDF-model.The simple reaction model, the eddy dissipation concept (EDC), developed for combustion reactions is found to be inadequate for liquid phase reactions. Modifying the model with a multiple-timescale turbulent mixer model (EDC-MTS) gives a substantial improvement.The multi-peak presumed PDF model and the two EDC-models are tested against reactive mixing experiments. The multi-peak presumed PDF model is found to give the best agreement between measured and predicted product concentration. The modified version of the EDC model, the EDC-MTS model, gives results which are close to the multi-peak presumed PDF model.

Photochemical and photocatalytic reactors are being widely used in the wastewater treatment and water/air purification. One of the popular reactor configurations for water purification is the multiple lamp tubular reactor. The efficiency of purification in a photoreactor is determined by the distribution of UV radiation within the tubular space. In this study, the light intensity distribution in a two-lamp photoreactor has been evaluated using a finite volume scheme (which is essentially a variant of the discrete ordinate model). The irradiated space was divided into a number of finite volumes and angles (directional discretisation). Simulations were performed for both homogenous and heterogenous (with titania catalyst) reactors. Contours of incident radiation and local volumetric rate of energy absorption were plotted and analysed. Model predictions were used to assess the effect of catalyst loading on the light intensity distribution inside various reactor configurations. The optimum value of the distance between lamps in a 2-lamp photoreactor was found to be dependent on the optical properties of the photocatalyst used. Model predictions were found to match-well qualitatively with the published data from the literature.

For better understanding the process of particle coating by chemical vapor deposition (CVD) in the fluidized bed, the simulation of the deposition process was combined with a discrete particle model (DPM). Based on the experimental results of the thermal decomposition of tri-isobutyl-aluminum (TIBA) to produce aluminum onto glass beads, mechanisms on the micro-scale were investigated by single particle tracking. Zones of excessive growth as well as zones of insufficient mixing were identified. In particular, the take-up of aluminum was traced for selected particles that exhibited a large mass of deposited aluminum what in turn provides insight into the homogeneity and quantity of the coating throughout the bed material.

A novel fluidized bed membrane reactor has been developed for the production of high-purity hydrogen based on steam methane reforming (SMR). The reactor incorporates perm-selective membranes for in-situ removal of hydrogen from the reactor, thus shifting the thermodynamic equilibrium of the SMR reaction. The membranes also eliminate the need for downstream hydrogen purification.The endothermic reaction duty is provided either by external heating of the vessel wall or through direct air injection into the fluidized catalyst bed (autothermal reforming). The gas flow pattern within the fluidized bed induces internal circulation of catalyst particles between the central reaction (permeation) zone and outer heating zones. The circulating hot catalyst particles from the oxidation zone carry the required endothermic heat of reaction for the reforming while ensuring that the palladium membranes are not exposed to high temperatures or to oxygen. Another characteristic of the reactor configuration is that very little of the nitrogen present in the oxidation air reaches the reaction zone, thus maintaining the hydrogen driving force for the perm-selective membranes.The reactor concept was proven in a pilot reactor (0.13 m diameter, 2.3 m tall). A number of variables were studied, including steam-to-carbon ratio, temperature and pressure. The pilot reactor was operated with both external heating and direct air addition. Pure hydrogen (99.999+%) was obtained from the reactor and an equilibrium shift was demonstrated. The maximum pure hydrogen recovery obtained from the pilot reactor was 0.96 mol H2/mol CH4, limited by the installed membrane surface area for these tests.

Hydrogenation of sunflower oil over novel structured catalysts with pore size ranging from 3 to 20 nm, BET-specific surface area of 710-1200 m²/g, and catalyst metal loadings ranging from 0.7 to 5.0 % w/w, was investigated and compared to a commercial Ni-catalyst. The activity and selectivity of the catalysts as well as profiles of the reaction products such as trans fatty acids (TFA) of the hydrogenated oils were investigated. Surface characteristics of the support and the metal loading significantly affected the activity and selectivity of the Pd-catalysts. Catalyst supports with the pore diameter between 7-8 nm were more active than supports with lower pore diameters. The activity and selectivity of hydrogenation depended on Pd content, with maxima in the concentration range of 0.8 to 1.2 % w/w.The catalyst with Pd-loading of 1% w/w, supported on structured silica material was active and selective for the hydrogenation of sunflower and canola oils under mild process conditions. For same iodine value (IV) reduction, this catalyst produced about the same level of TFA, but produced less stearic acid and was more selective towards cis monoenes formation than Ni-catalyst. More importantly, this catalyst produced a reduced level of stearic acid which causes waxy mouth feel of the hydrogenated fat at increased levels.The lumped kinetic model of the hydrogenation of vegetable oils over Pd as well as Ni catalysts was found useful in the prediction of the fate of reaction and product lumps and was statistically robust. The uncertainty and confidence joint regions were estimated using the Bootstrap “Monte Carlo” technique.

Bio-desulfurization of model organosulfur compounds, namely alkylated dibenzothiophene (DBT) and diesel oil was conducted in a laboratory scale fermenter using pure and isolated bacterial strains namely, Rhodococcus sp. (NCIM-2891) and Rhodococcus sp. JUBDS1 respectively. The diesel used as the feedstock was previously hydrodesulfurized up to a total sulfur content of 300-500 ppm. The Sulphur concentration of product stream of diesel from the process lies in the range of 10-15 ppm. Alkylated dibenzothiophenes were used in pure forms. During the study of chemostat in batch mode the initial substrate concentration of organosulfur compounds, the sulfur concentration in diesel, the stirring speed, temperature, oil to aqueous phase ratio, etc., were varied as parameters. The intrinsic kinetic parameters of bacterial growth like maximum specific growth, mmax, half saturation constant, and KS were determined by carrying out experiments in Erlenmeyer flasks using suitable medium and subsequent non-linear regression analysis. A mathematical model showing the transient behavior of chemostat was developed. Simulated data were compared with the experimental findings to validate the model developed.

This study proposes a comparison between CFD (Computational Fluid Dynamics) and experimental results for a cylindrical stirred tank reactor. The stirred tank reactor considered here is 0.7 m high, 0.4 m in diameter, baffled and has four impeller blades on a single shaft, placed at the symmetry axis. The tank is filled with water and serves as a model for molten aluminium purifying reactors.The experimental part of the study compares the use of PIV and LDV techniques to determine the flow field inside the tank. We use PIV to get a complete picture of the flow field in two vertical sections of the tank (one radial and one eccentric). LDV is then used to check and compare the velocities obtained by PIV along horizontal lines belonging to the two vertical sections considered under the PIV-measurements. Good overall agreement between the two methods is obtained.The experimental results are compared to simulation of the tank using an alternative model that we have implemented in the commercial CFD software Fluent. The modeling approach we propose here avoids the classical sliding mesh technique and its tedious use (especially for pre- and post-processing). Our model does not physically represent the impeller blades. It rather introduces them as a time-dependent source term in the momentum equation. This term is only accounted for at the cells that would have been swept by a blade if the blades were present. The modeling of the tank and the set-up of the case are faster and easier. The computation of the flow field using this modeling approach is compared to the sliding mesh approach and to the experimental results discussed above. Results show good agreement between experimental data and both modeling approaches (sliding mesh technique and no impeller model). The model we propose does not bring more accuracy but more convenience in the modeling work.

The Mexican petrochemical industry, Morelos S.A. de C.V. is one of the biggest and most important petroleum industries in Mexico and in Latin America. It has an activated sludge system to treat its wastewater flow, which is about 7000m3/d. The wastewater contains volatile organic carbon substances classified as toxic. The old surface aeration system was changed for fine bubble diffusers; however, one major drawback of the new aeration system is that the temperature in the bioreactor has increased due to the compression of the air, which at the compressor exit reaches up to 85oC. This effect produces a temperature in the aerated bioreactor of 32oC during the fall, whereas in the spring and summer, the bioreactor temperature reaches higher values than 40oC. The high temperatures reduce the microorganisms’ activity and cause higher volatilization rates of volatile compounds, among other effects, which affect the performance of the biological treatment. In this work we performed a stability analysis at different equilibrium points in different scenarios of temperatures, recycle flowrate and wastage flowrate. The eigenvalues of the state matrix of dynamical model of the petrochemical wastewater treatment were evaluated. The model describes the effect of the temperature on the soluble and volatile pollutants removal rate, heterotrophic organism growth rate, oxygen uptake rate and biomass concentration in the settler. The model was obtained from tests that were done in laboratory reactors with 14 L of capacity, which were operated at different temperatures (from 30oC to 45oC), with the same conditions as the actual system. Simulations were done based on the model obtained manipulating the operation variables Qw and Qr. The stability analysis was also done to the reactor without recycle (Qr = 0 m3/d). In all the cases the stability analyses was performed. The stripping of VOC´s is an important process of COD removal at high operation temperatures.

The efficient characterization of the nonlinear dynamical response of kinetic molecular simulations is discussed. Following ideas originally proposed by Kevrekidis et al. [1, 2], one can empower molecular simulations as model-free equations and use them as a reference to perform bifurcation detection. Such a procedure requires the use of trajectories from the molecular simulation to generate low-order models (e.g. polynomial) that allow one to infer the location of a bifurcation. If such identification step can be performed robustly, a feedback control policy that drives the molecular simulation to the bifurcation point can be constructed. In previous work, the identification of the low-order model has been singled out as the key element in handling noise trajectories, such as those generated by low-resolution molecular simulations. Here, a procedure motivated by the use of Kalman Filter observers is proposed as a means to give robustness to the identification procedure. The potential of the technique to characterize the dynamical response of kinetic molecular simulations is illustrated using examples related to the CO oxidation over a catalytic surface.

Driven by the global competition and market demand, the petroleum industry is under increasing pressure to produce clean, high-value products from low-cost feedstocks with minimum capital outlays. This has triggered an urgent need to develop robust process models capable of predicting the effects of feedstock composition, operating conditions, and catalyst properties on product quality. To develop such models, one generally splits the reaction mixture at the molecular level, examines microscopic interactions among individual reactions, and then works all the way up to the macroscopic level. Along the way one performs lumping to keep the size and complexity of the problem at bay. This is not only crucial for practical applications, but also important for fundamental understanding, since one does not want to lose sight of the forest for too many trees. This paper gives an overview of new and emerging theoretical tools for building kinetic models for hydrocarbon processing. While the emphasis is on reaction kinetics, the question of reducing hydrodynamic complexities is addressed as well.

Typically a Ziegler-Natta catalyst formulation includes the use of alkyl aluminum compounds and this investigation involves the substitution of triethyl aluminum (TEAL) with n-BuLi. It was found that this lithium compound, used as first co-activator, produced polyethylene at rates high enough for solution polymerization technology. Catalyst activity and molecular weight data were collected from a lab-scale continuous polymerization reactor at 150, 170 and 190oC. A balance between n-BuLi and a second alkyl aluminum compound (AA-2), used as second co-activator in the catalyst formulation, enabled the catalyst to achieve a performance (catalyst activity and polymer molecular weight) similar to that obtained with TEAL. It is proposed that the optimum balance between these two compounds is related to the difference in the reducing power of n-BuLi and AA-2, and one of the chain transfer reactions is to AA-2.