Showing papers in "Chemical Engineering & Technology in 1992"

Mixing times in stirred suspensions

Abstract: In stirred systems, the presence of solid particles leads to a pronounced lengthening of mixing time, sometimes to over 10 times that for the single-phase state. The mixing behaviour is strongly heterogeneous, since the slurry is mixed comparatively rapidly, while markedly slower homogeneization occurs in the solid-free zone. This is the consequence of different fluid velocities in the two regions. For particle settling velocities in excess of 5 cm/s, mixing times assume maximum values on reaching the state of complete suspension. By contrast, at lower settling velocities, maximum mixing times occur before suspension is complete. Mixing times are influenced only by the state of suspension and not by the mode of its generation. Consequently, for the fulfilment of the 90% suspended slurry height criterion, mixing times are independent of stirrer speed, solid concentration, type of agitator or diameter ratio d/D. The effects of particle diameter, viscosity and equipment dimensions on mixing time, when the 90% slurry height criterion is fulfilled, are reflected by the ratio of the vessel diameter to a representative liquid velocity

Xanthan production in stirred tank fermenters: Oxygen transfer and scale‐up

Abstract: The microbial polysaccharide xanthan was produced by cultivation of Xanthomonas campestris in four stirred tanks with volumes ranging from 0.072 to 3 m3. Many of the previously suggested correlations for the volumetric mass transfer coefficient described the present data very unsatisfactorily. When Rushton turbines were employed as stirrers, the experimental results agreed well with a correlation suggested by Kawase and Moo-Young; for the more energy-efficient INTERMIG impellers, this correlation was modified by a factor of three. Constant specific oxygen transfer rate was found to be the appropriate scale-up criterion for constant space-time yield and constant product quality, specifically, the molar mass which determines the viscosity yield.

Mixing characteristics in the torus reactor

Abstract: The hydrodynamic characteristics of propeller-induced toroidal flow in a loop reactor were investigated by performing an RTD analysis. The experimental determination of circulation time allows the calculation of the mean axial velocity with respect to the rotational speed of the impeller. RTD measurements are interpreted with the aid of the dispersion plug flow model, and it is shown that axial dispersion is relatively weak in the torus reactor. The mixing time was also determined experimentally and related to the circulation time. A direct relationship between mixing time and axial dispersion coefficient has been established, leading to a correlation for the mixing time in a torus reactor.

Experimental and analytical study of countercurrent flow limitation in vertical Gas/liquid flows

Abstract: An experimental and analytical study of adiabatic countercurrent flow limitation (flooding) in single vertical ducts is reported. The experiments were carried out in a rectangular channel using saturated liquid and vapour of Refrigerant 12 (CCl2F2). The steady-state liquid delivery (flooding) curves as well as local pressure drop and void fraction distributions in the countercurrent flow were measured in a range of system pressures from p/pcrit = 0.16 to p/pcrit = 0.31, and for various total liquid injection rates and locations. The measured flooding curves j1 = f(jg) as well as pressure drop and void fraction during partial liquid delivery (j1 < j1in) were not affected either by the axial liquid feed location or by the excess liquid rate carried upwards by the vapour. Moreover, for given flow conditions during flooding pressure drop and void fraction were essentially the same at different axial positions. Radial void fraction distributions evaluated from optical fibre probe data indicate an annular-type flow pattern. Based on this experimental evidence, a mechanistic core/film flow model was developed for the calculation of flooding. The analytical results are compared with the present high pressure and with comparable atmospheric pressure experimental data, showing reasonable overall predictions not only of the flooding curves, but also of the pressure drop in countercurrent flow.

Chemical reaction fouling of organic fluids

Abstract: Chemical reaction fouling of heat exchangers involves deposition that is caused by chemical reactions which occur in the fluid, or on the surface itself. The factors which influence fouling of organic fluids via autoxidation and thermal decomposition processes are described and the dependence of the rate of thermal fouling on chemical effects and process variables demonstrated. Mitigation techniques are outlined in brief.

Multicomponent‐liquid‐membrane permeation of organic acids

Abstract: The parameters of lactic acid separation from fermentation broths by Emulsion-Liquid Membranes (ELMs) have been optimized. Using these parameters, lactic acid can be separated up to 90%, concentrated up to 3 times and cleared of most of its by-products. Only about 1% of glucose and amino acids is permeated together with the lactic acid. With these parameters, not only lactic acid but also other monocarboxylic acids can be separated, the better and faster the lower the pKa-value of the acid the fewer polar side groups it carries. When separating dicarboxylic acids, the mass transfer is hindered by the second carboxylic group. The smaller the distance between the two carboxylic groups, the slower is the separation of the acid. When several organic acids are separated from the same fermentation broth, their mutual influence is not very strong, at least when the concentrations of the acids are about 100 mmol/l. However, sulphuric acid which is added to the fermentation broth throughout permeation to maintain the broth at a constant pH, does exert a strong influence. In order to reduce the competition between the organic and sulphuric acids during permeation to a minimum, as little as possible sulphuric acid is used. However, a certain quantity of sulphuric acid is necessary to maintain the pH of the fermentation broth during permeation at 4.5, otherwise organic acid separation would become too slow.

Experimental studies on parametric sensitivity of a batch reactor

Abstract: Runaway behaviour of exothermic reactions in a batch reactor has been studied experimentally on the example of acetic anhydride hydrolysis catalyzed by sulphuric acid, in a 700 ml batch reactor. High parametric senstivities of peak reactor temperature with respect to initial reactor temperature and catalyst concentration were obtained.

Development and start-up of a fixed bed reaction column for manufacturing antiknock enhancer MTBE

Abstract: A laboratory fixed-bed reaction column is presented, in which a heterogeneouslyl catalyzed reversible fluid-phase reaction and the separation of the reaction product from unreacted or inert components by distillation are performed simultaneously. The rection studied, the formation of methyl tertiary butyl ether (MTBE) from methanol and isobutylene, is catalyzed by strongly acidic macroreticular ion exchange resins. Packing elements are Raschig rings, manufactured for the first time from the above-mentioned ion exchange resins. This in-house made catalyst shows practically the same ion exchange capacity as the commercial products. The control system installed in the equipment is capable of correcting strong disturbances os the steady state. As shown by expwrimental results, the superimposed distillation process occurs in such a way that the chemical equilibrium of the MTBE synthesis does not limit isobutylene conversion.

Heat transfer mechanisms in gas fluidized beds. Part 1: Maximum heat transfer coefficients

Abstract: This contribution presents the prediction of maximum heat transfer coefficients in bubbling fluidized beds, which takes into account thermal and fluid-dynamic properties of particulate material and fluidizing agent. The analysis suggests that heat transfer between heating or cooling surfaces and bubbling fluidized beds consists mainly in a particular manifestation of convective heat transfer. Another feature is an appropriate modelling of the particle convective component leading to a two-phase Prandtl number.

Heat transfer in gas fluidized beds. Part 2: Dependence of heat transfer on gas velocity

Abstract: Part 1 of this report investigated the effects of system properties on heat transfer between heating or cooling surfaces and bubbling fluidized beds. This investigation produced six correlations, which define the respective maximum heat transfer. The present contribution shows that the heat transfer depends on superficial gas velocity, with the relationship expressed in terms of a dimensionless excess gas velocity which defines the ratio of effective thermal conductivity by particle mixing to that of the fluidizing agent. Simple procedures for a reliable prediction of heat transfer in bubbling fluidized beds are presented.

Estimation of dispersion coefficients in packed beds

Abstract: The design and performance of fixed beds are greatly influenced by fluid dispersion. Unfortunately, the existing design data do not provide an accurate picture of this phenomenon. This paper presents an attempt to characterize the dispersive features of packed beds by obtaining reliable estimates of the associated coefficients in the axial (DL) and radial (DR) directions. Such an objective is achieved by developing a representative two-dimensional pseudo-continuous dispersed flow model which is subsequently employed to compute the desired coefficients using data obtained from a refined experimental approach. The established values have been correlated to allow such coefficients to be reliably predicted under a variety of physical and operating conditions.

Experimental studies on semibatch reactor parametric sensitivity

Abstract: Runaway behaviour of an exothermic reaction in a semibatch reactor was studied experimentally on the example of hydrolysis of acetic anhydride catalyzed by sulphuric acid, in a 700 cm3 thinwalled stainless steel stirred reactor, 102 cm in diameter and 130 cm high The reactor was partially immersed, to a depth of 21 cm, in a thermostatic bath, in order to obtain sufficiently rapid cooling of its contents A reaction run was performed by first introducing water, acetic acid and sulphuric acid into the reactor After the contents attained bath temperature, acetic anhydride at room temperature was fed to the reactor for a fixed time period, at a constant flow rate When water was present in excess of the stoichiometric requirement and addition of acetic anhydride completed before the reactor attained peak temperature, the system showed sensitivity to coolant temperature and sulphuric acid concentration When the reactants were present in stoichiometric quantities and addition of acetic anhydride was completed after the reactor attained peak temperature, the system did not exhibit any significant parametric sensitivity

Significance of axial heat dispersion for the description of heat transport in wall-cooled packed beds

Abstract: two-dimensional pseudo-homogeneous model with axial dispersion of heat has been solved numerically with different boundary conditions at the inlet and outlet of the packed bed. The model solutions are fitted to experimental temperature profiles, determined in a wall-cooled packed bed in which a hot gas is cooled down, and best fit values of the effective axial and radial thermal conductivities and the wall heat transfer coefficient are obtained. In the range of Reynolds numbers employed, Re > 50, the axial dispersion of heat is found to be of no significance for the description of heat transport in wall-cooled packed beds without reaction, provided that the inlet boundary conditions are chosen appropriately. If a radially flat inlet temperature profile is assumed, while the actual profile is curved, an apparent improvement in the description of heat transport is observed when axial dispersion is incorporated into the heat balance and high effective axial thermal conductivities are obtained. If a Danckwerts type inlet boundary condition is used, assuming a flat temperature profile immediately in front of the inlet, an apparent improvement is also found on incorporation of axial dispersion of heat. This is caused by the temperature jump at the inlet, compensating for the overestimation of inlet temperature, in the case of cooling. The latter also explains why the inclusion of axial dispersion may eliminate the so-called length effect, often related to the effective radial thermal conductivity and the wall heat transfer coefficient. It is shown for the outlet boundary condition that deletion of the axial dispersion term from the heat balance at the outlet is a convenient boundary condition for the model being solved numerically.

Heat, Mass and Momentum Transfer in Packed Bed Distillation Columns

Abstract: This contribution presents the basic equations for heat, mass and momentum transfer in multicomponent packed bed distillation processes. In some situations, the use of strongly simplified models is justified, but when approaching more difficult and, at the same time economically more interesting regions of operation where non-linear effects are significant, these models are likely to fail. Consequently, a more rigorous vapor-liquid equilibrium model should be employed since the pressure drop in the column will not be negligible in those regions. Furthermore, neither constant parameter hold-ups nor heat and mass transfer coefficients are assumed. Simulations demonstrate some interesting process properties. The impact of the surrounding on the process is discussed and a three-dimensional model extension is outlined. (orig.).

Gas‐phase properties in stirred tank bioreactors

Abstract: Bubble properties in stirred tank bioreactors equipped with standard radial flow Rushton turbines have been investigated. Bubble velocity patterns within a stirred tank reactor were derived from measured local bubble velocity distributions. Local information on specific interfacial areas, bubble number densities and bubble diameters together with local gas hold-ups, measured in a model medium at gas flow rates which are employed in practice, is presented. All measurements were performed with two new ultrasound pulse techniques which can be used in model as well as during real cultivation processes.

Trajectories and impact velocities of grinding bodies in planetary ball mills

Abstract: The motion of grinding bodies in conventional ball mills has been repeatedly investigated, both theoretically and experimentally. It is well-known that, depending on mill filling and speed of rotation, different motion patterns occur and some of these patterns, especially that of cataracting, can be described by simplified theories. This contribution presents such a theory of the cataracting motion of grinding bodies in a planetary ball mill. An analytical method for the evaluation of trajectories is given which permits an iterative calculation of the time and impact location of the grinding bodies on mill shell or mill filling. This leads to the determination of the impact velocity of grinding bodies and its component normal to the mill shell. On the assumption that this component is decisive for the grinding effect, conditions for an optimal design of a planetary ball mill are deduced.

Fatigue of thick‐walled pipes from soft martensitic and semi‐austenitic chrome‐nickel steels under pulsating internal pressure

Abstract: Thick-walled components subjected to pulsating internal pressure are widely applied in high-pressure technology and in manufacturing processes such as fluid-jet cutting and high-pressure cleaning, mainly in conjunction with reciprocating pumps. Corrosive fluids require high-strength and tough chrome-nickel steels with soft martensitic or semi-austenitic structure. This contribution reports on the fatigue of thick-walled plain and cross-bored pipes made from high alloy chrome-nickel steels such as X5 CrNiMoCu 21 8 and X5 CrNiMo 16 5. The speciments, uniaxial standard form and thick-walled pipes, were cut from forged blocks in the three axial directions. For loading with pulsating pressure, a suitable, high-frequency piston pulsation machine has been developed. The fatigue tests on pipe specimens show typical Woehler characteristics with only slight scatter and relatively good isotropy. The surprisingly large admissible pulsating pressure can be explained for the applied steels by dynamic generation of residual stresses as a result of shake-down effects. Presentation in a Smith digram explains the occuring dynamic shake-down and its favourable results in comparison to the more brittle highly tensile steels. It also reveals that heat treatment to higher tensile strength does not always yield an increase in the admissible pulsating pressure. It will be shown that static autofretting and shake-down affect the fatigue strength of thick-walled pipe specimens in the same way. Tests with internal liners in the tube specimens provide indications on the sensitivity of material failures towards fluids. The investigation aids the understanding of the fatigue behaviour and the design of components made of modern high-strength corrosion resistant steels.

Problems in use of supported liquid‐phase catalysts in fluidized bed reactors

Abstract: Since supported liquid-phase (SLP) catalysts can become irreversibly deactivated due to evaporation of the solvent used to prepare the contacts, a saturation technique is proposed to stabilize the catalysts activity. Application of this technique requires isothermal reactions which can be achieved directly in fluidized bed reactors. Although the fluidization properties of SLP catalysts are mainly determined by interparticle forces, it can be shown that this type of reactor is suitable for an SLP-catalyzed gas reaction provided that a number of prerequisites are fulfilled. Taking a hydroformylation reaction catalyzed by an Rh complex dissolved in dimethyl glycol phtalate as an example, it is demonstrated that the activity of the prepared SLP catalyst can, indeed, be stabilized by saturation of the reactant gas mixture with the employed solvent

A study of mass transfer from single large oscillating drops

Abstract: A study was made of mass trasfer rates from single large oscillating drops of pure liquid-liquid systems, in the size range of 5 to 10 mm. A thrermostatically-controlled, 50 mm in diameter, 1000 mm long, rising drop column was used, in which mirrors in the jacket enabled front and side views of drops to be photographed simultaneously. The systems studied were (1) toluene and acetone (dispersed)-water (continuous), and (2) n-heptane and acetone (dispersed)-water (continuous). High concentrations of acetone (up to 3.75 kmol/m3) were used to examine the effect of different parameters on the mass transfer rate, frequency and amplitude of oscillation in countercurrent operation. Previous theories and empirical correlations [2–6, 12, 13, 15] for the prediction of overall mass transfer coefficients showed large deviations from measured values. These may have aarisen because the models do not represent droplet oscillation accurately, and/or apply only to oscillations of small droplets. Fair agreement was obtained for small oscillating droplets as low solute concentrations. The oscillations of a travelling drop were asymmetrical; the period of oscillation was uniform for mutually-saturated systems but changed when mass transfer was taking place. The periods were longer than those predicted by the Lamb [7] and Shroeder and Kintner [37] correlations. Terminal velocities predicted from literature correlations [32, 34] did not give reasonable agreement with experimental data when there was mass transfer of solute. The drag coefficient increased with increasing mass transfer rate from the drop. Correlation of the results and the dispersed phase mass trasfer coefficients by dimensional analysis resulted in the correlation 11 with a mean deviation of ±23%, by insertion of experimental oscillation frequency data. This will facilitate more accurate prediction of the dispersed phase mass transfer coefficients relating to equipment containing droplets in the oscillating regime, e.g. pulsed columns or agitated tanks.

Heat transfer mechanisms in gas fluidized beds. Part 3: Heat transfer in circulating fluidized beds

Abstract: Part 1 of this contribution reported on the effects of system properties on heat transfer between heating or cooling surfaces and bubbling fluidized beds. This investigation produced four correlations which define the respective maximum heat transfer. Part 2 of this study suggests that the heat transfer between exchanger surfaces and bubbling fluidized beds depends on superficial gas velocity, expressed as dimensionless excess gas velocity. The present paper shows that heat transfer coefficients in circulating fluidized beds can be predicted by evaluation of a state diagram, which combines three dimensionless groups: Nusselt number, Archimedes number and a dimensionless pressure gradient. A comparison of coal combustion experiments with own cold model measurements indicates that the radiative component of heat transfer coefficients is only evident at very low dimensionless pressure gradients.

Mass transfer improvements in single phase in regular packings

Abstract: Two ways of enhancing liquid to solid mass transfer in regular packings are proposed and investigated: the introduction of a fluidized bed of inert particles and of turbulence promoters by constructing the packing from expanded metal. It is shown that both systems can be of considerable benefit in improving the mass transfer performance: up to 200% in the first case and up to 130% in the second case. However, the energy dissipated by fluidized particles is excessive in comparison to that consumed by turbulence promoters.

Atomization of liquids and suspensions with hollow cone nozzles

Abstract: Large centrifugal forces of rotational flow are used in hollow cone nozzles to form a thin liquid film in the oulet, which disintegrates into relatively small droplets. The flow in the nozzle can be calculated by means of simple physically meaningful balances, based on the cyclone theory. The influence of wall friction is taken into account via a wall friction coefficient which depends on the Reynolds number of the nozzle flow. The break-up mechanism of the liquid film was investigated under the consideration of nozzle oulet velocity and film thickness as well as gas and liquid properties. With increasing velocity and film thickness, a transition from aerodynamic wave break-up to turbulent atomization was observed to take place. Equations presented allow the calculation of mass flow rate, pressure drop and drop size distribution of hollow cone nozzles with any given geometry

Investigation on pure shearing of cohesive limestone with true biaxial shear tester

Abstract: Pure shear deformation of a bulk solid sample was performed with the True Biaxial Shear Tester. The influence of stress history on pure shearing was examined by three different consolidation procedures. Apart from the influence of the consolidation procedure, the effect of volume change, i.e. the change in the bulk solid density during consolidation, on pure shear deformation was investigated. Cohesive limestone with x50 = 5 μm was used for the experiments.

Local heat transfer and heat flux distributions in finned tube heat exchangers

Abstract: A detailed analysis of the effect of local heat transfer distributions, which were determined experimentally, on temperature distributions in the fins and in the fluid between two adjacent fins of finned tube heat exchangers is presented. The heat transfer distributions, obtained by a mass transfer technique, prove to be very inhomogeneous over the fin surface, leading to variable results for fin efficiency calculated by the conventional method. In addition, the effect of temperature variation in the fluid, which is usually neglected in fin analysis, is shown in many cases to exert a very strong influence on fin efficiency and total heat flux. For the optimization of fin designs, a combined analysis of local heat transfer, conduction resistance in the fin and temperature variation in the passage is indispensable

Modelling of mass transfer phenomena of associated systems in packed distillation columns

Abstract: Mass transfer between liquid and vapour phases in packed distillation columns has been investigated by the two-film theory for binary associated systems. The effects of association reaction in systems with the association of one component in both phases, taking into account a concentration-dependent liquid-phase association constant, were examined. The mass transfer coefficients are modelled, taking into account the association constant. The experiments were carried out at atmospheric pressure with acetic acid-benzene and acetic acid-toluene systems in a column packed with Rasching rings.

Mass transfer during the removal of dissolved heavy metals from wastewater flows in fluidized beds of ion exchange resins

Abstract: A fluidized bed of cation exchange resin was used for the removal of cupric ions from a simulated wastewater effluent consisting of copper sulphate solution. Variables investigated were: superficial liquid velocity, particle diameter, bed height and copper sulphate concentration. These variables were studied with respect to their effect on the solid-liquid mass transfer coefficient. The coefficient was found to increase with increasing superficial liquid velocity. Increasing particle diameter, bed height and copper sulphate concentration were found to reduce the mass transfer coefficient. The experimental data can be correlated by the equation valid for the following conditions: 0.23 < Re1 < 2.27; 0.52 < ϵ < 0.87; 0.0127 < dp/d < 0.0417 and 0.0095 < dp/L < 0.125.