Other Water Related Research

Since reverse osmosis membranes are very sensitive to foulants such as colloids, inorganic scale and biofouling, proper pre-treatment process therefore becomes a critical factor for a successful long-run seawater reverse osmosis (SWRO) plants. Recently, low pressure membrane has been successfully used in the pre-treatment for wastewater reclamation by reverse osmosis (RO). This is because membrane pre-treatment offers several advantages such as smaller plant footprint, better quality of feed water for RO unit and less chemical consumption. As a result, the use of low pressure membrane is now being considered as a viable solution for pre-treatment to SWRO plants but further improvements in membrane configurations and operations need to be investigated to reduce fouling and energy consumption.

The first aim of this study is to investigate the efficiency of pre-treatment using MF and UF submerged hollow fibre system by varying the operation parameters. All experiments were operated on a bench scale rig which can automatically control the filtration, backwash cycle. The influence of different operation parameters, such as filtration time, backwash duration, backwash strength, air scouring during backwash is investigated. The following information is used to evaluate the efficiency of the different pre-treatment: the cycle time, the net flux, TMP profile, the permeate composition and the newly developed membrane filtration index (MFI) value for the permeate. The cake deposition and removal will also be investigated. A new MFI was developed to predict better predict fouling propensity of feed waters, particularly where small particles are present. A cross-flow sampler was used to simulate the hydrodynamic conditions close to RO membrane in terms of particle capture. The fouling resistance of the captured was then assessed by dead-end ultrafiltration. Comparison between fractionated particles and bulk solutions were compared with model solutions of mixed silica particles. Further validation of this new fouling index (CFS-MFI) will be carried out during reverse osmosis filtration.

Other aspects of the project include optimal design of spacers to reduce pumping costs in SWRO membrane modules.

First, the relative effect of individual processes featuring in the hybrid process was assessed and the effectiveness of UV photolysis was found to be strongly dependent on lamp configuration with the current setup (8x8W UVA external illumination) achieving a modest 5%. The high Rf observed for TiO₂ adsorption (5.5x1011m^-1) was anticipated since TiO₂-and-HA-laden particles were reported to form a dense fouling layer. The highest Eff (69%) was obtained when all three processes were combined together, highlighting the synergetic effect of the hybrid system. Also, the hybrid process resulted in lower Rf (4.6x1011m^-1) than TiO₂ adsorption, indicating its selectivity in destroying HA constituents with higher membrane fouling propensity. The effect of using membranes with different average pore sizes (from 0.45 down to 0.04µm) on the performance of the hybrid process was investigated. As expected, tighter membrane not only resulted in higher Eff (up to 82% with the 0.04 µm membrane), but also lead to higher Rf (5.5x1011m^-1). When the influence of operational parameters on the process was examined, the optimum TiO₂ concentration was found to be 0.6g/L (Eff = 83%). Although increasing TiO₂ concentration leads to higher surface area for organic adsorption, it also increases turbidity and weakened UV penetration. Increasing initial organic concentration was found to decrease Eff and was expected to increase Rf as more abundant organic molecules resulted in the denser cake deposit. . The importance of the influent nature and pollutant characteristics in this type of treatment was also assessed as various water sources were tested. Membrane degradation was not observed in any of these experiments, indicating the relative stability of polymeric membranes in this photocatalytic environment.

Removal efficiency of humic acid with combinations of filtration, UV, and UV+photocatalysis

Trace organic chemicals, like endocrine disrupting compounds (EDCs), pharmaceutically active compounds (PhACs) and personal care products (PCPs), present in wastewater effluents are known to potentially cause detrimental effects to human health and to the biotic environment if not removed during the treatment process. High-pressure membrane processes such as nanofiltration (NF) can be used efficiently in applications where a high water quality is required. Previous research indicated that the fouling layer formed on the membrane surface during filtration could significantly affect the rejection of trace chemicals and could either improve or jeopardize the quality of the treated water. Conflicting results on the effect of fouling on solute rejection have been reported and studies revealing the mechanisms leading to either improved or diminishing rejection of trace organic chemicals by fouled NF membrane are limited so far.

Using a commercially available NF membrane (NF-270, Dow/Filmtec), accelerated organic fouling was achieved by using a variety of natural organic matter (NOM) fractions ranging from humic acids, extracted from river water and from soil, surface water, protein (bovine serum albumin) solution, and wastewater effluent from a tertiary treatment process (membrane bioreactor). Different concentrations of NOM and operating modes (such as constant flux and constant pressure operation) were considered. A mixture of 18 trace chemicals representing a wide range of different physicochemical properties was added at the nanogram-per-liter range to the different feed water qualities and their level of rejection was assessed by a gas chromatography-mass spectrometry (GC-MS). According to their physicochemical characteristics, the trace chemicals were grouped into three categories: (1) hydrophilic non-ionic, (2) hydrophilic ionic, and (3) hydrophobic non-ionic. Variations in hydraulic resistance, membrane surface charge, roughness and relative hydrophobicity were measured for each experiment. Preliminary results indicated that feed water matrices and operational modes were the major factors governing the trace chemicals rejection. Under constant flux operation, rejection of contaminants increased after fouling, as compared to those obtained under constant transmembrane pressure. Changes of the membrane surface characteristics due to the formation of an organic fouling layer were confirmed by the observed increased hydrophobicity and decreased surface charge, which could explain the rejection mechanisms of compounds targeted in this study: (1) the main rejection mechanism for the hydrophilic non-ionic compounds was size exclusion, as their rejection remains relatively constant throughout the experiments; (2) In the case of the hydrophilic ionic chemicals, the initial rejection mechanism was electrostatic exclusion, which became offset by size exclusion as fouling occurred. Rejection of small compounds declined by 10%, while rejection of larger chemicals decreased only by 2%; (3) Hydrophobic non-ionic compounds presented high initial rejection due to their adsorption on the membrane surface. During long-term operation, their rejection decreased, as they were able to diffuse through the hydrophobic fouling layer.

The pulp mill industry consumes 10-40 liters of water per kilograms of produced paper and ranks as third (99,238 ML) in overall water consumption for industry in Australia (Australian Bureau of Statistics 2007). There are 22 mills across Australia. Disposal of wastewater from pulp and paper mills provide a significant challenge, particularly for inland plants. Recycling however requires improved treatment processes to restore the quality of the water to a standard comparable with the fresh water. In order to produce high quality water from the pulp and paper mill effluent, reverse osmosis (RO) needs to be used due to the high level of soluble contaminants. However the presence of large amount of organics in paper mill effluent will decrease the performance of RO. By introducing an appropriate pre treatment method prior to RO can minimize the pollutant level and increase the efficiency of RO. The aim of this study is to focus on nanofiltration as a designated pre-treatment option for RO and predict performance of RO with nanofiltration permeate. Also challenges on using RO such as silica fouling will be compared with alternative discharge and treatment options using cost modeling.