Influence of ozonation and adsorption processes onto the chlorine demand of surface natural water

The aim of this work was to quantify the influence of the adsorption process, alone or coupled with ozone, in reducing the chlorine dose required in the disinfection stage of a drinking water treatment train. The raw water of the Blue reservoir (Paimpont, France) was fed to a pilot system consisting of a constant pre-treatment (coagulation-flocculation and sand filtration) but with different polishing processes (adsorption alone or combined with a pre-ozonation stage). Chlorine demand, at 2, 6 and 24 hours of reaction time, was determined for the effluent collected from the sand filter, according to the break-point method. The clarified and sand-filtered effluent was subsequently treated either by ozonation (15 min, 3 mg O-3/L) coupled with adsorption (EBCT=8 min) or by adsorption alone. Two types of activated carbon were evaluated (NC35, coconut shell and PICABIOL, wood). The final effluents of these two treatment trains were collected to determine their chlorine demand. The adsorption process with PICABIOL halved the chlorine dose (2 mg Cl/L), compared with 4 mg Cl/L for the sand filtered effluent, required to obtain a residual concentration of free chlorine of 0.5 mg/L. Pre-ozonation coupled to adsorption also reduced significantly (by a factor of two and four depending on the type of GAC used) this chlorine dose, after 2, 6 and 24 hours of reaction time. Moreover, the residual chlorine content was the same for these three reaction times because ozone and activated carbon efficiently removed most of the reducing inorganic nitrogenous pollutants. The removal of aromatic organic compounds, quantified as Abs254 nm, by these processes decreased also significantly the calculated THM FP (trihalomethane formation potential) to the extent that this value for the ozonated and or adsorbed effluents has always been lower than the guideline proposed by the EPA (100 mu g/L) for total THMs (yearly average).