RICHMOND, VIRGINIA INVESTIGATES UV DISINFECTION OF LARGE CSO OUTFALL

After the completion of the Phase II CSO control improvements, the City of Richmond, Virginia conducted a re-evaluation study on its original 1988 CSO Long Term Control Plan (LTCP). The purpose of the study was to reassess the last phase of the original LTCP in light of EPA's 1994 National CSO Control Policy and state-of-the-art technologies. The study identified that after the completion of the Phase II CSO controls approximately 79% of the entire CSO volume is discharged through the City's largest CSO outfall, Shockoe Creek, at a peak flow rate of 5,000 MGD. The study acknowledged that reliable disinfection of such a large flow rate would be challenging and that inactivation efficiencies greater than 80% may not be feasible or practical. The receiving water quality model showed that an 80% reduction of bacteriological loading would result in a significant improvement in water quality of the James River. The study recommended that a comprehensive disinfection pilot study be conducted to evaluate the feasibility of existing technologies in achieving 80% or possibly higher disinfection efficiency at the Shockoe outfall.Two disinfection technologies, ultraviolet (UV) irradiation and chlorination/dechlorination, were evaluated during this disinfection study. This paper presents the findings from the UV disinfection pilot study conducted between April 2003 and April 2004. The UV pilot study was conducted on raw CSO with a wide range of water quality characteristics. UVT varied from 20 to 61%, and TSS ranged from 10 to123 mg/L. Influent fecal coliform concentration was between 230,000 and 6,130,000 cfu/100 mL, and influent E. coli concentration varied from 86,700 to 2,100,000 cfu/100mL. The 12-month UV disinfection pilot study demonstrated that UV appears to be a feasible technology for disinfection of Shockoe CSOs with 80% efficiency. Multivariable regression models are being finalized to predict the delivered UV dose and/or log reduction in the reactor. The models will determine the approximate number of UV lamps required given target disinfection efficiency (or delivered dose), flow rate, power setting, and water quality (UVT).