AWWA WQTC62538

The goal of this study was to provide pilot-scale information about the ability of ultraviolet (UV) to disinfect unfiltered surface water. A pilot-scale low pressure UV system was operated with raw water from the Pardee Reservoir at the East Bay Municipal District in California. Monitoring of particle counts, turbidity, and UV transmittance at 254 nm was performed for 12 months with data recorded hourly. The online particle count data were collected in 8 different size ranges between 2 m and 50 m, but for reporting purposes the data was divided into two: greater than and smaller than 10 m in diameter. In addition to monitoring the water quality properties for evidence of particulate matter that could conceivably inhibit UV disinfection (i.e. particle counts and turbidity), the inactivation of indigenous total coliform bacteria, total aerobic spore-formers (TAS), and somatic coliphages across the pilot UV reactor was periodically monitored in an effort to determine if inactivation was achieved at a level consistent with expectations based on UV treatment in particle-free water, and if not, to try to correlate impaired inactivation with turbidity or particle counts. These three groups of organisms were selected to include those that are of a size consistent with Giardia and Cryptosporidium and might therefore experience a similar degree of particle shielding (i.e. coliforms and TAS), and virus-sized organisms (i.e. somatic coliphage). Previous work had also indicated that these organisms might be present in surface waters at high enough ambient concentrations to allow several log inactivation to be measured across the UV system. Total coliform bacteria were enumerated using the membrane filtration technique based on Standard Method 9222 (APHA et al. 1998). TAS were enumerated using a membrane filtration technique (Rice et al. 1996, Verhille et al. 2003). Somatic coliphage were enumerated using the spread plate technique, Standard Method 9224 (APHA et al. 1998). The pilot UV system was a 5 to 15 gpm low pressure high output (LPHO) Trojan UVMax®. The reactor had been biovalidated by the manufacturer with fluences available as a function of flow rate, UVT, and lamp age. The UV fluence was confirmed onsite by biovalidation using MS-2 phage on three occasions over the study period, and ranged from 90 to 100 mJ/cm². Methods for the biovalidation procedure followed Bolton and Linden (2003) for the collimated beam work. Includes 8 references, figure.