Batch Recycle Membrane Test as Alternative to the Rapid, Bench-scale Membrane Test for Evaluation of Nanofiltration of Surface Drinking Waters

Nanofiltration is being widely investigated to remove natural organic matter (NOM), the precursor to disinfection by-products in drinking water. Bench-scale tests are common before more expensive pilot-scale tests of membranes. However, even at bench-scale, a large volume of water will be needed for a complete testing program if the rapid, bench-scale membrane test (RBSMT) is used. A direct comparison of the RBSMT and an alternative, batch recycle membrane test (BaReMT) showed that similar results could be obtained regarding flux decline and NOM rejection. The BaReMT is a more desirable test method because of its ease of operation and smaller sample volume requirement. While a batch volume as small as 4L is acceptable at a recovery of 90%, a larger volume of 15 L is required for recoveries of 80% and less to avoid a foulant limitation. The relationship of recovery and rejection showed that the mass transfer of NOM across the membrane surface was controlled by either diffusion or convection. Several equations were derived to predict the ability of the RBSMT to simulate the performance of a full-scale array configuration. It was shown that at high recoveries (>75%)the RBSMT underestimates the solute concentrations of a full-scale array configuration. Atomic force microscopy and scanning electron microscopy were used to examine the surfaces of several membranes. Microbiological and NOM foulants were found on the surface of the membrane. Finally, it was determined that under conventional pretreatment conditions, the use of nanofiltration (NF) for NOM removal is not feasible for Durham and Fayetteville, NC due to the frequency of cleaning required to maintain an acceptable flux.Master of Science in Environmental Engineerin