Assessing Adsorbable Organic Halogen Formation and Precursor Removal during Drinking Water Production

Adsorbable organic halogens (AOX) were measured at the outlet (${\mathrm{AOX}}_{\mathrm{o}}$) and after 24 h contact time with residual disinfectant (${\mathrm{AOX}}_{24\mathrm{h}}$) for 10 samples obtained from 7 drinking water treatment plants (WTPs) that employ chlorine-based disinfectants. ${\mathrm{AOX}}_{\mathrm{o}}$ ranged from 0.75 to 7.60 μM ($26.2–266.0\mathrm{\mu g}\mathrm{Cl}/\mathrm{L}$) and increased by up to 170% after 24 h contact time. The results were also compared with AOX formation potential (${\mathrm{AOX}}_{\mathrm{FP}}$) measured in the laboratory for the samples collected prior to disinfection at the same WTPs. In general, ${\mathrm{AOX}}_{24\mathrm{h}}$ and ${\mathrm{AOX}}_{\mathrm{FP}}$ values varied by less than $\pm 20\%$. A suite of disinfection byproducts (DBPs) were individually measured to elucidate the AOX composition. Trihalomethanes and haloacetic acids explained the largest fraction of AOX; however, the percentage of AOX not accounted for by measured DBPs varied greatly between samples (14.5–62.0%). Additionally, the fate of ${\mathrm{AOX}}_{\mathrm{FP}}$ was assessed across the seven WTPs. On average, coagulation removed $67\pm 7\%$ of ${\mathrm{AOX}}_{\mathrm{FP}}$ across the WTPs, while dual media filtration did not substantially remove ${\mathrm{AOX}}_{\mathrm{FP}}$. The effect of ozone on ${\mathrm{AOX}}_{\mathrm{FP}}$ varied between WTPs, while biological activated carbon filtration removed $33\pm 13\%$ ${\mathrm{AOX}}_{\mathrm{FP}}$. The ratio of known versus unknown ${\mathrm{AOX}}_{\mathrm{FP}}$ did not change across the WTPs.