Development of membrane testing protocols for characterisation of RO and NF membranes

The CSIRO Flagship Cluster project Advanced Membrane Technologies for Water Treatment is a collaboration of nine Australian Universities and the CSIRO looking at improving the energy efficiency of desalination of seawater using membrane processes. One of the initial aims of the cluster is to develop a standardised suite of tools and techniques for the characterisation of membranes for seawater desalination. Measures used to assess membrane separations are speed, selectivity and stability. These three performance measures depend upon the physical and chemical properties of the membrane, the feedwater and the type of filtration process being used (i.e. Dead end batch filtration versus continuous crossflow filtration). Our proposed testing protocol will cover three main categories; transport properties, pore characteristics and surface properties. The transport properties of membranes are strongly related to the pore and surface characteristics, however no general predictive capability exists yet that is widely accepted in the literature.

The development of an appropriate membrane testing protocol is an essential component of the CSIRO Membrane Desalination project. The testing protocol will provide a framework for direct comparison of the physical and chemical characteristic of newly developed membrane materials. As a result of this literature review, we have developed a preliminary testing protocol as follows: • Primary characterisation of speed and selectivity will use dead end filtration for rapid analysis, supported by crossflow filtration work for data validation. This primary characterization includes the evaluation of flux and rejection data for model seawater (32,000 mg/L as NaCl) and brackish water (2000 mg/L NaCl) solutions. Both sodium chloride and mixed sodium/ calcium ion solutions will be tested across a range of pH values and operating pressures. • Further tests of membrane stability will be conducted using crossflow systems to estimate changes in flux and rejection when a foulant such as alginic acid or humic acid is present. We still have considerable work to do to determine the most realistic foulant. • Membrane pore characteristics will be evaluated using PALS analysis. • Surface roughness will be evaluated using AFM measurements. • A combination of captive bubble contact angle and streaming potential measurements will allow the determination of membrane surface chemistry and membrane–solute affinity.