Comparing a silver-impregnated activated carbon with an unmodified activated carbon for disinfection by-product minimisation and precursor removal

Abstract

During disinfection, bromide, iodide and natural organic matter (NOM) in source waters can lead to the formation of brominated and/or iodinated disinfection by-products (DBPs), which are often more toxic than their chlorinated analogues. The objective of this study was to compare the efficiency of a silver-impregnated activated carbon (SIAC) with the equivalent unimpregnated granular activated carbon (GAC) for the removal of bromide, iodide and NOM from a matrix of synthetic waters with variable NOM, halide, and alkalinity concentrations, and to investigate the impact on DBP formation. An enhanced coagulation (EC) pre-treatment was employed prior to sample exposure to either carbon adsorbent. Excellent halide removals were observed by the SIAC treatment across the sample matrix, with iodide concentrations consistently reduced to below the method reporting limit (< 2 μg/L) from as high as 25 μg/L, and 95 ± 4% removal of bromide achieved. Bromide removal by unimpregnated GAC was poor, however iodide removal was comparable to that achieved by SIAC. The combination of EC with SIAC treatment removed 77 ± 8% of the dissolved organic carbon (DOC) present, across the sample matrix, which was similar to removals by EC/GAC (67 ± 14%). Combined EC/SIAC treatment reduced both total trihalomethanes (tTHMs) and total dihaloacetonitriles (tDHANs) formation by 97 ± 3%, while also achieving a greater than 74% removal of two chloropropanones and a 92 ± 8% decrease in chloral hydrate (CH), compared to untreated samples, regardless of the sample's starting water quality (bromide, alkalinity and NOM concentration). Combined EC/GAC treatment led to similar DBP removals to EC/SIAC for the fully chlorinated DBPs, however, brominated DBPs were less efficiently removed, or experienced concentration increases.