An investigation has been carried out, on a pilot plant scale, into the factors affecting the hydrodynamic behaviour, mass transfer efficiency, liquid and gas residence times distribution functions and the extent of fluid mixing in a two sieve plate air-water distillation column simulator. The hydrodynamic behaviour was investigated as a function of the column operating conditions and wherever possible the results have been reported as correlation equation suitable for design purposes. The air-aqueous glycerol-oxygen system was chosen for the mass transfer study because the low solubility of oxygen in glycerol solutions means that the mass transfer may be regarded as being controlled by the liquid film. An accurate continuous sample and analysis technique was employed and the efficiency of both the tray alone and the tray plus downcomer was measured. The liquid phase residence time distribution functions were measured using an improved dye injection and photocell detection technique. The degree of mixing was determined using the variance of the residence time distribution function and the eddy diffusion model. The residence time and degree of liquid mixing were investigated as functions of the column operating conditions. The gas phase residence time distribution functions were measured using an ethane tracer with a double sample infra-red detection. The degree of mixing was interpreted using an eddy diffusion model and the Peclet number was successfully correlated with the liquid momentum. The mass transfer efficiency has been successfully related to the hydrodynamic behaviour, the liquid residence time and the extent of liquid mixing on the tray. The liquid momentum has been shown to be the major variable affecting the degree of mixing in both the gas and liquid phases.

Mutual solubilities of 3 methyl 2 butanone-water have been determined analytically for temperatures from - 8 C to 180 C at low pressure and from - 8 C to 90 C at pressures up to 350 MN/m[2]. It is found that this system shows a phase behaviour at elevated pressure similar to that of the homologous system 2 butanone-water. The phase behaviour of the ternary system 3 methyl 2 butanone-water-propanone was studied analytically and optically at ambient pressure at 0 C and 40 C. Liquid-liquid equilibria of the binary system 2 butanol-water were thoroughly studied as a function of temperature and pressure. An hitherto unknown immiscibility phenomenon shown by this system was observed at 1.3.8 MN/m[2] isobar in which a domed curve with an U.C.S.T. is followed by a closed solubility loop having a L.C.S.T. and an U.C.S.T. The phase behaviour of the binary system propanone-carbon disulfide was investigated at - 12 C and - 14 C up to a pressure of 362 MN/m2. The results support an earlier investigation carried out at - 2 C which indicates that increasing pressure, raises the critical solution temperature. For each system above the relations existing among the consolute solution properties are discussed in the light of the observed results. The mutual solubility of the binary system thiodipropionitrile-toluene was studied at atmospheric pressure by a light transmission method. It appears that the demixing curve of this system exhibits a L.C.S.T. inbetween two U.C.S.T.'s over a small concentration range in an open miscibility loop.

Foam fractionation is a, method of separating components of a solution that differ in surface activity; in such a solution, surface active material concentrates at the liquid surface. When the solution is foamed, a very large additional surface is created in the foam in which the surface active material is therefore concentrated. Proteins are well known as foam stabilisers and foam fractionation has been used to separate enzymes on a laboratory scale. While the process has the potential for protein separation on a large scale, no satisfactory treatment exists that will give a quantitative prediction of the behaviour of a foaming system. The aim of this work was to develop an universal design equation capable of predicting the Enrichment ratio (the chosen criterion of separation efficacy) under any given set of operating conditions. Foam fractionation of an idealised pure protein system of Bovine Serum Albumen in buffer at pH 4.6 was carried out in the batch mode; the range of liquid volume employed being 0.2-11.5 litres. Firstly the operating parameters affecting Enrichment were identified by carrying out experiments over a wide range of operating conditions. The parameters were found to be bulk liquid concentration, foam column height, superficial gas velocity, foaming cell diameter and also bulk liquid depth. Using the concept of Ideal Foam a simplistic Theoretical treatment was developed to give a design equation which has four empirical constants, [mathematical equation] This equation predicts all the experimental results in the region where the assumption of Ideal Foam is valid, with the highest confidence limits using the chi[2] statistical test. The equation does not take into account any change in bulk liquid depth,while the mean foam bubble diameter was found to be essentially constant over the wide range for which the design equation applies. The model has not been adapted for use with binary protein solutions.

Tests have been performed on the modified oscillating baffle contactor (O. B. C.) in which different types of perforated baffles have been introduced, and the results obtained analysed. Two liquid systems have been studied. The systems carbon tetrachloride-acetic acid-water and Methyl ISO-butyl ketone (MIBK)-acetic acid-water were chosen partly from expediency and partly from their suitability and for comparison purpose with published data. Flammable and highly toxic materials were not possible due to prevailing laboratory conditions. Runs made with MIBK were limited strictly for safety and health reasons. Equilibrium data obtained in our laboratory has been extended and equations describing the equilibrium distribution of the systems have been established with high degree of accuracy and compares favourably with published data. For the true performance of the extractor detailed knowledge of the drop size and distribution was necessary. Some high speed cine-film of the drops in the column were taken and analysed. Hold up values of the dispersed phase measured by both the drainage and mano-meteric methods were recorded. The longitudinal dispersion (axial mixing) which occurs in both phases has been examined with the aid of dye injection technique, and the coefficients calculated from Peclet numbers based on the length of the extractor. The relationship between the true, measured and piston flow values of N. T. U. are established. Equations have been developed based on a modification of Rod's method, to predict the true concentration profile of the dispersed phase corresponding to that of the continuous phase along the column. Relationship existing between H. T. U. values and ka values have been presented graphically to explain the extractor performance. Equations to determine sectional values of transfer units and axial mixing have been derived. Data for power dissipated in the contactor have been obtained: by using highly sensitive torque - transducers and recorders. These have been treated both by dimensional analysis and modified existing correlations. Interfacial tension measurements were made for both systems. The, accuracy of these measurements have been found reasonable. They give the magnitude of the change in interfacial tensions with concentration. The measured efficiency of an experimental R. D. C. has been investigated for the purpose of comparison with the O. B. C. for the system CCL 4 - water - acetic acid. The. drop size distributions resulting from the measurement in the 2'6" long O. B. C. is correlated by two equations, which correlate the data with a maximum deviation of +/- 5%. The development was used to predict the interfacial area in the 4' column and hence calculate k values.

Differing experimental results have been obtained for gas absorption into horizontal stagnant liquid pools, with and without chemical reaction, using a novel simultaneous optical birefringent interferometric and pressure transducer technique. The thesis is divided into two parts. Section one is mainly concerned with physical gas absorption. The systems investigated were carbon dioxide/water, acetylene/water, ammonia/water, sulphur dioxide/water and carbon dioxide/propylene carbonate. The take-up of gas (M[t]) by the absorbing solutions, concentrations and concentration profiles of gas in liquid have been obtained relative to exposure times from both experimental and,generally, theoretical considerations. The possibility of a surface resistance (1/k[s]) for the absorption of carbon dioxide by water and propylene carbonate, and acetylene by water, has been considered. The method of analysis indicates an appreciable resistance to exist in all cases. The effect of a soluble surfactant on the transfer process has also been studied. It has been demonstrated that for all the systems studied, convective disturbances (perhaps in the form of microflows or eddies) are produced almost instantaneously (0.04 sec was the shortest time of investigation possible in this study) upon exposure of absorbing gas to liquid. This convective disturbance could be responsible for a much larger transport of solute from the liquid surface than can occur by molecular diffusion alone. Further studies on convection involved examining the effect of pool depth on the absorption of carbon dioxide and acetylene by a 3% aqueous solution of wetting agent (pool depths from less than 0.4 mm to 25 mm), and sulphur dioxide by pure water (approximately 2 mm to 25 mm). In all cases, there was a peak depth at which absorption was greatest. This phenomenon is suggested to be due to the presence of convection in the liquid pool. An analysis, based on hydrodynamic stability theory, was carried out for buoyancy-driven convection in the sulphur dioxide/water system. Calculated values of critical Rayleigh numbers from experimental results were shown to be larger than theoretically predicted values. Section two of the thesis deals with interfacial turbulence and absorption of an acid-gas in several systems of absorption accompanied with chemical reaction. These were carbon dioxide absorption by aqueous solutions of mono-di- and tri-ethanolamine, sodium and potassium hydroxide, and hydrogen sulphide absorption by aqueous solutions of mono-di- and tri-ethanolamine. All the systems were examined optically and by the transducer technique. The carbon dioxide/monoethanolamine system for concentrations less than IN was studied using a laminar jet absorption apparatus. Photographs of turbulence are shown, and a time-scale for the onset of visible convection in the form of roll cells is presented for all the systems. A minimum time of several seconds for this onset time, which varied with concentration, was found for all the systems. However, analysis of interferograms for absorption of carbon dioxide by IN monoethanolamine, suggests that some form of convection is almost instantaneously generated upon exposure of gas to liquid, as in the case of physical absorption. Transducer absorption studies, when compared to theoretical predictions, suggest there is an accumulation of reaction product at the gas/liquid interface until viscous forces in the liquid bulk are overcome. At this point, roll cell convection commences and the amount of gas absorbed begins to exceed that predicted by theory.

Studies of the mass transfer mechanisms which occur when carbon dioxide is absorbed into flowing water films in an inclined cell are described. The amount of gas absorbed is small and therefore requires a highly sensitive detection apparatus. For this purpose a new modified Michelson interferometer was constructed which enabled direct concentration readings to be obtained. Gas concentrations in the liquid were also obtained by titration. The experimental work was mainly confined to angles of inclination less than 5°. Hydrodynamic results showing increased flow rates at the sides of the inclined cell, due to meniscus effects, are presented. Hydrodynamic "end effects" at the liquid exit from the cell, which cause increases in the absorption rate, have been successfully minimised by the design of the apparatus and the experimental technique employed. The take up of gas, concentrations and concentration profiles have been obtained relative to the gas-liquid contact time, from both experimental and theoretical considerations. It has been shown that there is an appreciable surface resistance to the absorption process. The applicability of certain theoretical equations to the results has been studied. It has been demonstrated, for angles of inclination less than 3°, that convective disturbances (in the form of microflows, eddies or perturbations) are present. This causes an increased transport of solute from the liquid surface than can occur by molecular diffusion alone. Beyond 3° eddies or disturbances are present due to hydrodynamic instabilities eventually leading to "observable" wave formation. The effect of waves on the concentration profile, and the amount of gas absorbed, has been considered. Experimental results have also been obtained for desorption of gases from horizontal stagnant water pools, using a simultaneous birefringent interferometric and pressure transducer technique. The systems investigated were carbon dioxide-water, acetylene-water, ammonia-water, sulphur dioxide-water and acetone-water. The amount of gas desorbed, concentrations, concentration profiles and surface resistance were obtained relative to the desorption times from experiment; A theoretical analysis has been made. The effect of pool depth and pool area on the desorption of acetone from water were investigated (for pool depths between 0.3 mm and 25 mm; pool areas between 2.2 cm[2] and 17 cm[2]). The desorption study was performed in order to assist the interpretation of results obtained from the flowing liquid cell. This was mainly in connection with the nature of the surface resistance and convective disturbances encountered during absorption into thin liquid films.

Ammonia in the form of dissolved salts can he removed from industrial waste water by successive biological nitrification (oxidation to nitrate) and denitrification (reduction of nitrate to nitrogen gas). Nitrification of ammonium sulphate solutions was studied in a hatch activated sludge system, at low substrate concentrations (less than 60 g/m3 ammonia-nitrogen) to investigate substrate limitation and at high concentrations (up to 5000 g/m3 ) to investigate substrate inhibition. The kinetics of the system were considered in terms of a "specific substrate removal rate", analogous to microbial specific growth rate, to allow for the effects of non-viable but biochemically active bacterial cells in the sludge and the probable variation of bacterial yield coefficient with specific growth rate. The effects on this parameter of substrate concentration were summarised as mathematical models by applying a pattern-search curve-fitting procedure to the experimental data relating substrate concentration, product concentration and time obtained from batch runs at 25°C covering a range of initial substrate concentrations, mostly at a controlled pH of 7. At low concentrations, the effects of substrate concentration on the specific substrate removal rate were described better by a model based on the discontinuous function first proposed by Blackman than by one based on the Monod relation. At high concentrations, stagewise analysis of the hatch concentration-time data suggested that nitrification was subject to hysteresis, in that the specific substrate removal rate depended on the initial batch substrate concentration as well as the transient substrate concentration. A model based on the Haldane equation for enzyme inhibition gave a reasonable description of the effects of initial substrate concentration on the specific substrate removal rate. A correlation including three empirical parameters was found to describe the variation of specific substrate removal rate with transient substrate concentration for individual batch runs. When inocula from dilute stock sludge culture were introduced into concentrated substrate solutions, an acclimatisation phase was observed in which the specific substrate removal rate fell rapidly to the inhibited level. Inocula which were subjected to preliminary acclimatisation before inoculation did not show this effect. In additional work at low substrate concentrations, the optimum pH for nitrification at 25°C was found to be 8. 5, and empirical expressions were found to describe the relation between pH and specific substrate removal rate. The contribution to ammonia removal by chemical decomposition of ammonium nitrite (the van Slyke reaction) was found not to be significant in the conditions prevailing in microbiological nitrification.

This thesis is concerned with the principal features of solvent resistant nanofiltration (SRNF), and in particular its potential in fuel processing. In the recent and more established literature, dense polydimethylsiloxane (PDMS) has been identified as a potential SRNF membrane with separation properties at a molecular level. This thesis investigates the transport and separation mechanisms of dense PDMS membranes in SRNF applications relating to the filtration of organic solvents (mixtures are mixed to simulate fuel). Simulated fuels were created which comprised representative organic solvents with organometallic (OM) and poly-nuclear aromatic (PNA) solutes. The solvent flux and solute rejection behaviour of the solvent-solute systems were studied using a cross-flow air-driven filtration apparatus and a range of operating regimes.

No description available.

No description available.