Performance evaluation of different ultrafiltration membranes for the reclamation and reuse of secondary effluent

Abstract

This study investigates and compares the performance of two different types of ultrafiltration (UF) membranes in the recovery of water from secondary treated wastewater. Filtration experiments were carried out on a pilot scale cross-flow unit using synthetic wastewater similar to the quality of secondary treated wastewater by varying the operating parameters such as transmembrane pressure (TMP), feed composition and membrane configuration. The filtration experiments demonstrated that the flux recovery through spiral polymeric UF membrane was more sensitive to the variation in TMP compared to the tubular ceramic UF membrane over the range of TMP studied. The resistance in series model was used for the evaluation of the resistance to the permeate flux. The fouling resistance, particularly irreversible resistance compared to reversible resistance plays a major role in the total resistance for the tubular ceramic membrane. In contrast clean membrane resistance is the major contributor for the total resistance of the spiral polymeric membrane. Finally, the effectiveness of the filtration treatment was determined by evaluating the rejection coefficients for various pollution indices of the wastewater. Significant differences in the performance of the membrane types were observed which are likely to impact on the selection, operation and maintenance of the membrane system.

Introduction

Water recycling is a way of moving towards sustainable management of our resources and environment by adopting the concept of using water that is ‘fit for purpose’. The reuse of treated municipal effluent for non potable purposes is therefore considered vital in alleviating the demand on existing and limited water supplies while helping to protect remaining water sources from being polluted. In recent decades, advanced membrane technology has become increasingly attractive for reclamation of secondary wastewater as it is highly efficient, easy to operate, economical and satisfies more stringent discharge regulations and supply limitations [1], [2]. Porous membranes like ultrafiltration (UF) and microfiltration (MF) membranes present significant operational advantages such as reduced sludge, a high quality of permeate and a small space requirement due to high packing density. Moderate capital costs and ease of operation make it a very competitive alternative to conventional technologies [3], [4]. In many cases, UF alone or in combination with physico-chemical or biological treatment, have provided reclaimed wastewaters that were able to meet the established wastewater reclamation criteria [5].

One important issue that presents a major impediment to the progress of the membrane technology is membrane fouling by organic, inorganic and biological fouling materials. Membrane fouling causes a decrease in membrane performance and in turn limits the application of membrane filtration of wastewater effluent during wastewater reclamation/reuse [6]. Membrane fouling is characterised as a reduction of permeate flux through the membrane or higher TMP across the membrane, as result of increased flow resistance due to pore blocking, concentration polarisation, and cake formation [7]. This results in the reduction of the productivity of the membrane, a significant requirement for increase in transmembrane pressure (driving force), biodegradation of membrane materials and system failure, and ultimately increases the cost of operation [8]. While several studies have been performed to enable a better understanding of the fouling phenomenon, it is generally accepted that the performance of UF is influenced by the feed water characteristics, membrane type and operational conditions [9]. Currently membrane fouling encountered in reclamation of municipal wastewater is a major operational concern as a result of the large variability of wastewater quality and the complex and unstable nature of the organic materials present in the wastewater. Several investigations have been conducted considering size distribution, characterisation and quantification of organic materials that are assumed as major foulants in low pressure membrane processes [10], [11].

Currently, hollow-fiber membranes made up of polyvinylidene fluoride (PVDF) and polyether sulfone (PES) are predominantly used in large scale water treatment facilities because of their high chlorine and acid resistance. Structural modification of these membranes has also been a common practice used to reduce the membrane fouling. However little information is available in the literature on the specific chemical and physical factors that govern the fouling of low pressure tubular or spiral PVDF or PES treating secondary wastewater. Further the use of UF to treat municipal secondary treated wastewater has almost exclusively focused on polymeric membranes, while little research has been conducted with ceramic membranes. It is well known that ceramic membranes have inherently superior physical integrity, chemical resistance and thermal stability compared to polymeric membranes. Recent research using ceramic membranes treating secondary wastewater demonstrated a high permeate flux, high water recoveries and with less frequent chemical cleaning. [9], [12].

It has also not been specifically established previously whether ceramic and polymeric membranes tested under similar conditions on pilot scale will yield similar results with respect to fouling and removal efficiency. The objective of this manuscript is to describe pilot scale experiments where tubular ceramic and spiral polymeric UF membranes were tested with the same water sources to determine their filtration performance under similar operating conditions. Therefore this study has focused on (1) selecting commercially available polymeric and ceramic UF membranes as representative samples (2) carrying out comparison of their operation and performance (3) interpreting the results. The effectiveness of filtration treatments under different operating conditions were assessed on the basis of a decrease in various pollution indices of the wastewater such as chemical oxygen demand (COD), colour, turbidity and absorbance at 254 nm. Finally, the fouling mechanism in the present case was assessed by fitting the experimental data to a resistance in series model [12], [13].