Influence of factors affecting the removal of chlorine and chlorination byproduct precursors in organic filters

The addition of an ozonation step followed by a second stage biological act ivated carbon (BAC) filtration to a conventional treatment (coagulation-flo cculation, settling and sand-anthracite filtration) has been identified as a good mean for controlling chlorination by-product precursors. Lately, the addition of an ozonation step before filtration and the replacement of san d-anthracite filtration by sand-BAG filtration has been proposed as an econ omical solution when retrofitting treatment plants. The accumulation of flo cs and particulate matter in first stage sand-BAG filters may cause neverth eless some problems of substrate or oxygen diffusion to the biomass fixed o n the BAG. The main objectives of this study were to verify the impact of flocs/partic le accumulation in first stage sand-BAC filters by: (1) studying the effect s of filter backwash on the removal of chlorination byproduct precursors; ( 2) comparing the efficiency of sand-BAG filtration with that of first stage sand-anthracite filtration and second stage BAC filtration. Figure 1 shows the flow; chart of the pilot facilities used for this study. The raw water, from the Mille-Iles River, was settled in the St. Rose filt ration plant (nominal capacity of 110,000 m(3)/d). This settled water was t hen pumped to the pilot plant (nominal capacity of 48 m(3)/d) where it was ozonated in two sequential ozone reactors. The ozonated water stream was th en split by open weirs to three filtration trains. Four pilot filters were used: two replicate first stages and-BAC filters (filtres sable-CAB #1 and sable-CAB #2), one first stage sand-anthracite filter (filtre SA) followed by one second stage BAC filter (filtre CAB). Table 1 shows the design and t he operational parameters of the filters. Properties of filter media are de scribed in Table 2. Table 3 shows additional sampling information. Analytical monitoring included chlorine demand, dissolved organic carbon (D OC), ammonia, hexane extractable disinfection by-product precursors (HEDBPp ) and haloacetic acid precursors (HAAp). Table 5 indicates the HEDBPp and t he HAAp measured. All these parameters were measured on water samples taken at different filter depths according to empty bed contact time (EBCT). Results showed that floc and particle accumulation in first stage sand-BAG filters impaired the removal of some chlorination by-product precursors, su ch as HAAp which are considered to be easily biodegradable. Fortunately, bi ter backwash restored these removals. Chloropicrin precursors were also ide ntified as being easily removed in biological filters when compared to the biodegradation kinetics of dissolved organic carbon. Chloroform and trichlo ropropanone precursors were found to be removed as the global pool of disso lved organic carbon. The replacement of first stage sand-anthracite filter by sand-BAG filter an d the addition of an interozonation step could improve the quality of the w ater produced by the treatment plant. In fact, this study showed that sand- anthracite filtration was not suitable for the removal of chlorine demand, chloroform precursors, chloropicrin precursors and trichloropropanone precu rsors. Furthermore, the water quality of the effluent of first stage sand-B AG filtration and of BAG filtration was shown to be very similar. However, the coagulation-floculation step must be enhanced in a manner to prevent th e accumulation of Aocs and particulate matter in first stage sand-BAG filte rs. Results also suggested that the potential of removal of biodegradable p recursors of chlorination by-products is limited and should be part of a gl obal strategy including the optimisation of natural organic matter removal by other treatment processes (coagulation, settling, membranes). (C) 1999 E lsevier Science Ltd. All rights reserved.