Pilot evaluation of the potential log removal credit using chemical markers for reverse osmosis and nanofiltration

Abstract

Reverse osmosis (RO) and nanofiltration (NF) are widely applied membrane treatment technologies due to their ability to effectively remove particulates, dissolved compounds, bacteria, and viruses. However, in potable reuse applications, RO is given minimal log removal value (LRV) credit of 1–2 for pathogens due to the inability to demonstrate higher LRV performance daily. Therefore, systematic single-element membrane pilot tests (3 different RO and 1 NF membranes) were performed to assess the potential LRVs (> 4) for three chemical markers (i.e., sulfate, sucralose, and uranine) that could potentially enable membrane-based potable reuse systems to achieve higher credit. Lead- and tail-element tests, as well as new and chlorine-oxidized membranes, were tested to assess the effect of membrane location and damage on the marker LRVs. Undamaged and damaged membranes demonstrated LRVs of >5 for MS2 bacteriophage with no statistically significant differences observed for any of the 4 membranes tested regardless of membrane placement (i.e., lead vs. tail) or condition (i.e., new vs. oxidized). Statistical analysis by optimal cluster assignments for each membrane indicated there were little differences in effective LRVs between the three chemical markers. Though the LRVs could not achieve >4, likely due to diffusion limitations, they were > 2 for the three chemical markers that ranged from 2.7 to 3.6 across non-damaged lead- and tail-element tests for all three RO membranes. Uranine, while achieving LRVs of 2.9–3.6 for RO and 2.2–2.3 for NF for non-damaged lead- and tail-element tests, was deemed less preferable due to its high dose requirement and difficult chemical handling. Sulfate and sucralose proved to be effective as well but preferable with non-damaged, lead- and tail-element test LRVs of 2.7–3.5 for RO and 1.4–2.3 for NF. Damaged-membrane testing demonstrated the three conservative chemical markers were sensitive to the chlorine-oxidized membrane compromise, more so for RO than NF, before any virus breakthrough occurred.