Robustness of a demonstration potable reuse facility was evaluated through a series of system-wide chemical challenge tests spiking neutral low-molecular weight compounds (acetone, formaldehyde, N-nitrosodimethylamine (NDMA), and 1,4-dioxane) which are known to be challenging for removal through advanced treatment. Two advanced treatment train arrangements were compared: O3/BAC-MF/UF-RO-AOP and RO-AOP. Ozone and biological activated carbon (O3/BAC) offered significant attenuation of the smallest and most biologically degradable compounds tested: acetone and formaldehyde. These two compounds had limited reduction across the reverse osmosis (RO) membrane barrier and the different advanced oxidation process (AOP) setups used. 1,4-Dioxane was partially reduced across the oxidation barriers: 62% across ozonation and up to 95% across AOP depending on oxidant used and oxidant dose. Both a hydrogen peroxide (H2O2) based AOP (UV/H2O2) and a sodium hypochlorite (as HOCl) based AOP (UV/HOCl) demonstrated sufficient oxidation, providing no less than 0.5-log (68%) 1,4-dioxane attenuation required by regulators in the USA. NDMA was reduced across both UV/H2O2 and UV/HOCl from 157 to 267 ng/L to below the 10 ng/L established notification level for drinking water in California. Overall, addition of O3/BAC enhanced cumulative removal of all the spiked trace organic chemicals, providing greater protection against the spiked contaminants than RO-AOP alone. In addition, online total organic carbon (TOC) monitoring successfully captured the presence of the spiked chemicals.

Constituents in the feed of reverse osmosis (RO) processes pass into the product by two mechanisms: diffusion through the membrane and advection through defects in the system. While minerals and tracers pass via both mechanisms, pathogens pass only via advection. Using mass and flow balances, a rigorous model is developed to characterize the response of both viruses and tracers in RO systems. The model is used to assess several tracers being considered today. Early virus testing and the commercial requirements of desalination are reviewed, leading to the conclusion that some degree of advection can be expected to be present in RO systems but that commercial incentives can be expected to maintain pathogen removal due to defects somewhat greater than 4.0 log. Model results suggest that the best tracers currently in use can detect defects near this log reduction value but that there is still room for improvement.

The number of potable reuse projects in the United States has increased the need for operators who are knowledgeable about advanced water treatment processes, including membrane treatment technologies.