Robert W. Field

Critical, sustainable and threshold fluxes for membrane filtration with water industry applications.

Critical flux theory evolved as a description of the upper bound in the operating envelope for controlled steady state environments such as cross-flow systems. However, in the application of UF membranes in the water industry, dead-end (direct-flow) designs are used. Direct-flow is a pseudo steady state operation with different fouling characteristics to cross-flow, and thus the critical flux concept has limited applicability. After a review of recent usage of the critical flux theory, an alternative concept for providing design guidelines for direct-flow systems namely that of the threshold flux is introduced. The concept of threshold flux can also be applicable to cross-flow systems. In more general terms the threshold flux can be taken to be the flux that divides a low fouling region from a high fouling region. This may be linked both to the critical flux concept and to the concept of a sustainable flux. The sustainable flux is the one at which a modest degree of fouling occurs, providing a compromise between capital expenditure (which is reduced by using high flux) and operating costs (which are reduced by restricting the fouling rate). Whilst the threshold flux can potentially be linked to physical phenomena alone, the sustainable flux also depends upon economic factors and is thus of a different nature to the critical and threshold fluxes. This distinction will be illustrated using some MBR data. Additionally the utility of the concept of a threshold flux will be illustrated using pilot plant data obtained for UF treatment of four sources of water.

Enhancement of liquid phase adsorption column performance by means of oscillatory flow: an experimental study

Abstract Oscillations have been applied to the bulk fluid flowing upwards through a column packed with spherical beads of 3A zeolite that is used to dry an ethanol solution containing about 3.2 wt.% water. For a given bulk flow rate, concentration and temperature of the feed, the application of oscillations leads to a delay in the breakthrough time and a sharper breakthrough curve than for the case when there is no oscillation. Use of oscillatory flow, therefore, leads to an increased bed capacity up to the point of breakthrough, a lower length of unused bed and a reduced mass transfer zone length (MTZL). Improvements in all these parameters of up to 20% have been found for oscillations with frequencies up to 2.4 Hz and amplitudes up to 0.001 m. The best improvements were found for the highest frequencies and amplitudes that could be achieved with the apparatus. In the development of the process it is believed that a further 20–30% improvement could realistically be achieved. However, practical limitations to the upper bounds of these two oscillatory parameters arise from the strength of the adsorbent particles and the risk of movement of particles within the bed. The main reason for the improvement in the process, when oscillations are applied, is the increase in interparticle transport. The increase in interparticle heat transfer is potentially as relevant as that in interparticle mass transfer . The results obtained for this model chemical system show that process improvements through the use of oscillatory flow can be obtained with well-rounded particles and it is speculated that the method could be applicable more generally to all liquid phase adsorption processes.

Integration of vacuum and sweep gas pervaporation to recover organic compounds from wastewater

Pervaporation with hydrophobic membranes has been widely recognised as a possible process to recover organic compounds from wastewater. Compared to vacuum pervaporation, on which many researchers have focused, sweep gas pervaporation has received little attention. The aim of this study was to analyse opportunities for integrating and optimising both process layouts for the treatment of wastewater. The focus was on hollow fibre modules. Two module configurations of hollow fibre modules were considered: (1) shell-side and (2) tube-side feed flow. An advanced simulation program based on a phenomenological/semi-empirical model was used. The influence of (1) process parameters such as permeate pressure and the size of the sweep stream per module, and of (2) module design parameters such as void fraction or module configuration was determined for two model substances pyridine and phenol. Based on the simulations, guidelines for the optimisation of pervaporation are presented. These include the observation that for vacuum pervaporation shell-side feed flow is superior, whilst for sweep gas pervaporation tube-side feed flow should be selected. In the former case and for a given feed rate per module, the void fraction within the module should be selected as low as possible to reduce the effect of concentration polarisation. This approach is, however, limited by the pressure resistance of packed fibres causing an increasing pressure gradient on the feed side. For hydrophobic pervaporation of wastewater, sweep gas pervaporation should be combined with a moderate vacuum (of around 0.1 bar) to improve the pervaporation performance; the performance at atmospheric pressure for the conditions selected leads to excessive membrane areas. Similar to vacuum pervaporation the void fraction should be selected as high as possible for tube-side feed flow, and as low as possible for shell-side feed flow.

Fundamentals of Pressure-Driven Membrane Separation Processes

Publisher Summary This chapter presents some basic concepts related to membranes and membrane processes. Porous membranes are divided into two types according to their structures: microporous membranes and asymmetric membranes. Microporous membranes are characterized by the membrane pores throughout the membrane bodies. The pores are of uniform size (isotropic) or nonuniform size (anisotropic). Microporous membranes are designed to reject all the species above their ratings. In terms of materials, membranes can be classified into polymeric or organic membranes and ceramic or inorganic membranes. Organic membranes are usually made up of various polymers, among which the typical ones are cellulose acetate (CA), polyamide (PA), polysulfone (PS), polyethersulfone (PES), polyvinylidene fluoride (PVDF), and polypropylene (PP). Polymeric membranes are relatively cheap, easy to manufacture, available in a wide range of pore sizes, and they have been widely used in various industries. Inorganic membranes with the high mechanical strength and chemical and thermal stability over the conventional polymeric membranes have extended the application of membrane technology into many new areas. Membrane module is the way the membrane is arranged into devices and hardware to separate the feed stream into permeate and retentate streams. There are four kinds of membrane modules widely used in industry: tubular modules, hollow fiber modules, flat sheet modules, and spiral-wound modules.

Comparative performance of various plasma polysiloxane films for the pervaporative recovery of organics from aqueous streams

Abstract Thin SiO x C y H z film membranes were deposited from two different organosilicon precursors, one cyclic and one linear, in a low-frequency plasma polymerisation process. The ability of the films for effective recovery of various organic contaminants from aqueous waste was tested using the process of pervaporation. Four separate organic components, phenol, chloroform, pyridine and methylisobutylketone (MIBK), each representative of an industrially significant family of chemicals, were chosen for evaluation. Trends between the pervaporation performances of plasma polysiloxane membranes and the composite plasma parameter V / FM ( V : input voltage; F : monomer flow rate; M : monomer molecular weight) were found. Pervaporation measurements reveal that polysiloxane membranes synthesised by the plasma deposition process are sufficiently permeable for the mass transfer boundary layer above the membrane to be dominant. The permeability of plasma membranes formed from cyclic octamethylcyclotetrasiloxane were up to an order of magnitude more permeable than those formed from linear hexamethyldisiloxane. Future work should involve vapour permeation.

Preparation of modified difunctional PDMS membranes and a comparative evaluation of their performance for the pervaporative recovery of p-cresol from aqueous solution

Abstract Cross-linked polydimethylsiloxane (PDMS) membranes modified by the sequential introduction of two different side-arm functional groups, (CH2)3OC2H5 and (CH2)3NMe2, have been prepared and evaluated for the pervaporative recovery of p-cresol from aqueous streams. The loading of the ether group was varied from 5 to 20% and that of the amino group from 5 to 15%. These loadings are expressed in terms of the mole percentages of Si atoms within the membrane containing the identified functional group. The performances of these difunctional membranes are compared with that of pure PDMS and with membranes modified with a single functional group. It was found that various combinations of ether and amino groups improved p-cresol flux while maintaining the high separation factor exhibited by PDMS modified by a high loading of the amino group. These new materials also facilitated the preparation of defect-free films. Membranes containing AEE (10%) +AMI (5%) gave a very similar performance to monofunctional membranes containing 20% amino, but unlike those membranes the new difunctional membranes possess good film forming properties. Another combination of note is the AEE (20%) + AMI (10%) membrane that produced both an excellent cresol flux (that was 30% higher than that achievable using either AEE or AMI modified membranes) and a high separation factor ca. 50.

Tailoring wall permeabilities for enhanced filtration

The build-up of contaminants at the wall of cross-flow membrane filtration systems can be detrimental to the operation of such systems because of, amongst other things, the osmotic backflow it may induce. In this paper, we propose a strategy to avoid the negative effects of backflow due to osmosis by using 2D channels bounded by walls with a combination of permeable and impermeable segments. We show that preventing flow through the final portion of the channel can increase the efficiency of filtration and we determine the optimal fraction occupied by the permeable wall that maximizes efficiency. Our analysis uses a combination of numerical techniques and asymptotic analysis in the limit of low wall permeabilities. Finally, we consider how the energy cost of filtration depends on the Peclet number and show that the energy cost per unit of filtered water may be minimized by appropriately choosing both the Peclet number and the permeable-region fraction.