The Effect of Natural Organic Matter on UV/Hydrogen Peroxide Treatment and the Effect of UV/Hydrogen Peroxide Treatment on Natural Organic Matter

Ultraviolet light with hydrogen peroxide (UV/H2O2) produces hydroxyl radicals that degrade organic micro-pollutants. However, radicals react non-selectively with natural organic matter (NOM). This research effort quantified the effect of NOM variation on the efficiency of UV/H 2O2 contaminant destruction, explored the kinetics of hydroxyl radical/NOM reactions, determined the effect of UV/H2O2 on biofilm formation potential, measured UV/H2O2 impact on trihalomethane (TTHM) and haloacetic acid (HAA5) formation potential, and evaluated UV/H2O2 effects on TTHM speciation after chlorination. Granular activated carbon (GAC) adsorption was investigated to improve process efficiency and reduce by-product formation potential without creating brominated THM problems.;A year-long UV/H2O2 pilot study was conducted to study seasonal variations in NOM and multiple GAC breakthrough conditions. Pilot-scale reactors consistently achieved 80% atrazine degradation, allowing comparison of low pressure (LP) and medium pressure (MP) lamps for contaminant destruction efficiency and unintended by-product formation.;The effect of NOM on UV/H2O2 destruction of atrazine, metolachlor, methyl tert-butyl ether (MTBE), methylisoborneol, ibuprofen, gemfibrozil, and 17a-ethynylestradiol was evaluated. UV absorbance scans demonstrated changes in NOM from UV/H2O2 that increased under certain NOM conditions. As NOM increased, electrical energy per order (EEO) requirements for contaminant destruction increased; requirements increased similarly for all contaminants. UV/H2O2 followed by GAC eliminated the contaminants, throughout the year. LP lamps had lower E EO requirements than MP lamps. UV/H2O2 destruction of MTBE was evaluated with bench-scale experiments using waters with varying NOM. Destruction and EEO values correlated well with specific-ultraviolet absorption for pilot-scale and bench-scale experiments. Changes in the kinetics of NOM/hydroxyl radical reactions were observed with different types of NOM. Total assimilable organic carbon (AOC) concentration increased through UV/H 2O2 by 14 to 33%, more with conventionally treated (CONV) reactor influent than with Post-GAC influent. The AOC concentration increases generated by MP and LP processes were similar. The Spirillum strain AOC increased through UV/H2O2 50 to 65% due to formation of smaller more soluble compounds, e.g., organic acids. Pseudomonas fluorescens strain AOC concentration increased when CONV water served as pilot influent, but not when Post-GAC water was used. GAC effluent streams receiving UV/H 2O2 pretreatment produced biofilms with greater heterotrophic plate counts than controls. The GAC effluent stream following the MP reactor produced the most viable biofilm.;Three-day simulated distribution system (SDS) TTHM concentration increased through the UV/H2O2 reactors (20 to 118%). Post-GAC reactor influent produced lower 3-day SDS TTHM concentration than CONV influent after UV/H2O2. Three-day SDS HAA5 concentration increased for CONV UV/H2O2 pilot influent, but not for Post-GAC influent. No difference in 3-day SDS DBP concentrations was observed between LP and MP processes. Brominated THMs are more toxic than chloroform, thus minimizing them is desirable. UV/H2O2 did not shift 3-day SDS THMs towards the brominated species. UV/H2O2 increased the TTHM contribution of 3-day SDS chloroform by 7 to 13%, while 3-day SDS bromoform TTHM contribution decreased by 0.5 to 7%. GAC adsorption after UV/H2O2 insignificantly increased 3-day SDS bromoform concentration from 0.01 to 0.02 mumole/L.;UV/H2O2 can be used with GAC for excellent contaminant removal and minimal adverse effects.