Faecal bacteria and bacteriophage inactivation in a full-scale UV disinfection system used for wastewater reclamation

A study was carried out to compare the inactivation of faecal bacteria and one type of bacteriophage in a full-scale UV disinfection system. The syste m is part of a water reclamation facility for effluent reuse in golf course and agricultural irrigation. Influent and effluent samples were taken over two sampling periods (three consecutive days in July and one day in August ), with three different UV doses applied each day (ranging from 10 to 40 mW .s/cm(2) and 20 to 80 mW.s/cm(2) in July and August, respectively). Effluen t samples were also taken from a chlorine disinfection channel (5 mg Cl-2/L dose) operating in parallel to the UV system. Total coliforms (TC), faecal coliforms (FC), faecal streptoccoci (FS) and somatic coliphages (SC) were measured in each sample. F-specific RNA bacteriophages and bacteriophages o f Bacteroides fragilis were also measured one day in July. The decay ratio observed for all the microorganisms was similar when UV doses applied were low (July), ranging from 1.15 to 1.25 log-units. This suggests that bacteri al indicators may be suitable for virus inactivation control when low UV do ses are applied; however, such low doses are inadequate to achieve effluent quality requirements for unrestricted irrigation. At higher UV doses (Augu st), decay ratios for TC and FC were 3.1 and 2.8 log-units respectively, in dicating that they were more susceptible to UV exposure than SC and FS, wit h decay ratios of 2.6 and 1.0 log-units, respectively. Nevertheless, these higher doses were also inadequate to achieve water quality requirements for unrestricted irrigation. The decay ratio of SC during chlorine disinfectio n was clearly lower than that of the other microorganisms. Bacteriophages o f Bacteroides fragilis were more resistant to UV disinfection than SC and F -specific RNA. In fact, bacteriophages of Bacteroides fragilis were not aff ected during UV exposure. A UV dose ranging from 40 to 80 mW.s/cm2 marks th e borderline beyond which inactivation rates of SC are clearly lower than t hose of bacterial indicators.