Removal of disinfection byproduct (DBP) precursors in water by two-stage biofiltration treatment

Abstract

The removal of precursors of 36 disinfection byproducts (DBPs) in effluents from flocculation/sedimentation process was evaluated across a pilot-scale two-stage biofiltration process, i.e., a sand/anthracite (SA) biofilter (empty bed contact time (EBCT) of 7.5 min) coupled with a biologically-active granular activated carbon (GAC) contactor (EBCT of 15 min). The biofiltration process exhibited a good capacity for removal of the total DBP formation potential (DBPFP) (by 25.90 ± 2.63%), and GAC contactors contributed most to the DBPFP removal (accounting for 60.63 ± 16.64% of the total removal). The removal percentage of DBPFPs of different structure types was in the following order: halonitroalkanes (58.50%) > haloaldehydes (33.62%) > haloacetic acids (HAAs, 28.13%) > haloalkanes (20.46%) > haloketones (13.46%) > nitrosamines (10.23%) > halonitriles (−8.82%) > haloalkenes (−9.84%). The precursors of bromo-DBPs (containing only bromine atoms) and maximal halogenated DBPs (containing 3 & 4 halo atoms) were removed largely compared to other DBPs. Among the total DBPFP, trihalomethanes (THMs), HAAs, and chloral hydrate were the dominant DBPs, and they accounted for >92% of the total targeted DBPs by weight. Pearson correlation analysis (CA) and principal components analysis (PCA) indicated a significant association among these dominant DBPs. Canonical correspondence analysis (CCA) revealed specific ultraviolet absorbance (SUVA254) could serve as a good surrogate parameter for DBPFP. Pre-chlorination upstream of the biofilters may not greatly impact the overall removal of DBPFP by SA/GAC biofiltration. In addition, results showed that SA/GAC biofiltration was a useful procedure to remove the inorganic DBP chlorite.