No abstract

No abstract

ab s t r ac tThe specific energy consumption (SEC) of pressure-driven liquid-phase membrane processes, in particular the reverse osmosis (RO) process, has usually been estimated using a phenomenological approach, which does not explicitly consider the membrane properties and operating parameters. This paper presents a new analytical approach that has been derived, from a well-established theory, to estimate the SEC and to quantify the effect of membrane properties; namely, membrane permeability and surface area as well as the effect of process parameters such as feed pressure, recovery rate, membrane element permeate rate, and feed osmotic pressure. The SEC is also presented in terms of a dimensionless parameter, namely, the specific energy indicator (SEI), which can be used as a membrane property to indicate the SEC of the membrane element for a given process recovery rate and feed osmotic pressure. The SEC calculations are presented for desalting a NaCl solution with a salinity of 35,000 mg/L over a wide range of recovery rates and membrane element permeate flow rates. The calculations showed that for a membrane element with a permeate flow rate of 2 m 3 /h operating at 50% system recovery rate, the SEC of the RO process can be reduced by more than 35% if the membrane element flow rate factor is doubled, for example, from a value of 20 to 40 L/h.bar.

This study is a combination of experimental and theoretical works in an attempt to produce a new useful empirical model for the mass transfer in pressure-driven membrane separation processes. Following on from our previous work in Part I, this part II paper introduces three new permeability models when using aqueous solutions as feed. The Solution-Diffusion Pore-Flow Concentration-Polarization (SDPFCP) model, which is a combination between the Solution-Diffusion Pore-Flow (SDPF) model [1] and the Concentration Polarization (CP) model, is presented. The SDPFCP model examines the CP model to represent the transfer phenomena outside the membrane by merging its effect within the water permeability coefficient. A further development for this model, the SDPFCP+, is obtained by adding an additional resistance to the system in series with the membrane resistance and the CP. The second model shows fair representation of the experimental results. The Solution-Diffusion Pore-Flow Fluid-Resistance (SDPFFR) model is then proposed to provide better representation for the system. The feed solution resistance to water flux, the Fluid Resistance (FR), is suggested to replace the CP and the additional resistance. The latter model shows excellent fitting to the experimental results; it may be useful for development and design applications, when based on experimental data. Crown Copyright (C) 2010 Published by Elsevier By. All rights reserved

No abstract