Inactivation des virus dans l'eau sur une filiere de production a ozonation etagee

Abstract

The removal and inactivation of indigenous viruses for individual water treatment processes were investigated both in a pilot (3 m 3 h −1) and a full-scale (4000 m 3 h −1) plant, treating water from the Seine and Oise rivers respectively. Viruses were recovered by the previously described methods of adsorption-elution using electronegative Balston filters (Joret and Block, 1981) for the pilot plant study or alternatively using either electronegative or electropositive 1 MDS Cuno filters (Sobsey and Glass, 1981) for the full-scale water supply. Depending on the efficiency of filtration, the volumes of water samples concentrated directly in the field varied from 30 to 100 litres for raw, preozonated, stored river water and decanted water, and from 550 to 1000 litres for treated water downstream from the clarification step. BGM cells were used for virus detection by the plaque assay method. During this study, the initial virus concentration in the Seine river water entering the pilot plant, varied from 190 to 1420 PFU/1000 l. The virus removal observed after coagulation-flocculation with 35–50 g m −3 of water aluminum chloride (WAC) (a basic aluminum chloride) and settling, was 31–90% effective (130–390 PFU/1000 l. of decanted water), and did not exceed 1–2 logarithms after sand filtration at a rate of 5 m h −1 (83% of 600–800 samples were positive for viruses, maximum concentration = 80 PFU/1000 l.). The preozonation of river water (0.8 ppm during 2–3 min without residual oxidant) inactivated 64–97% of viruses (maximum viral concentration in preozonated water being 300 PFU/1000 l.). Under these conditions, no viruses could be isolated from samples of 550–1000 litres of sand filtered water. The treatment process for the industrial water supply located at Mery sur Oise was preozonation (0.5–1 ppm for 1 min), storage, coagulation-flocculation, settling, sand filtration, second stage ozonation (1.4–1.6 ppm for 10 min, residual oxidant: traces to 0.3 ppm), activated carbon filtration, postozonation (0.4 ppm free residual for 4 min) and chlorine dioxide (0.1 ppm). Nine series of samples were taken after various stages of treatment and tested for viruses. All samples of raw river water yielded positive results with relatively low numbers (10–146 PFU/1000 l.) compared with the number of enterovirus in the Seine river. The efficiency of preozonation before storage could not be evaluated in this study due to its intermittent operation. Enteroviruses were found in 83% of samples after storage (7–100 PFU/1000 l.), 77% after settling and 55% after sand filtration (3–10 PFU/1000 l.). Enteroviruses were not detected after the second stage ozonation in any of the nine samples. The subsequent postozonation and final treatment with chlorine dioxide serves as an additional barrier to ensure virologically safe drinking water. During studies on pilot and full scale water treatment plants, viral isolates are as follows: ECHO virus (49 and 37% respectively), coxsackie B virus (42 and 49% respectively) and poliomyelitis virus (9 and 14% respectively).