Pilot UV-AOP Comparison of UV/Hydrogen Peroxide, UV/Free Chlorine, and UV/Monochloramine for the Removal of N-Nitrosodimethylamine (NDMA) and NDMA Precursors

Three different ultraviolet-advanced oxidation processes (UV-AOPs) (UV/hydrogen peroxide, UV/free chlorine, and UV/supplemental monochloramine) used in full advanced treatment (FAT) for potable reuse were evaluated at pilot scale using reverse osmosis (RO) permeate from a FAT facility. Oxidant concentrations were varied, and each AOP was assessed for its ability to remove N-nitrosodimethylamine (NDMA) and NDMA precursors. For all AOPs, NDMA was removed well (to near 1.2 ng/L). However, UV/free chlorine exhibited less apparent NDMA destruction, and this level of destruction decreased as the free chlorine concentration increased. Removal of NDMA precursors was variable with no oxidant showing superior performance, ranging widely from 1% to 84% removal. The level of NDMA precursors increased in only two of 38 UV-AOP pilot tests, suggesting that formation of NDMA precursors by UV-AOP is limited. Adjustment of the UV-AOP pH was evaluated as a strategy for controlling NDMA precursors. Increasing the RO permeate pH prior to UV-AOP from 5.5 (ambient pH) to ∼7.0 using NaOH, to decrease the dichloramine concentration, decreased the level of destruction of NDMA precursors (i.e., increased the level of NDMA formation) in the UV/HOCl AOP (likely due to an increased level of radical scavenging by OCl⁻). Increasing the pH to ∼8.5 via the addition of lime decrease the level of NDMA destruction in the UV/H₂O₂ AOP due to the lower NDMA quantum yield at higher pH.