Variability in UV Disinfection System Performance in Municipal Wastewater Treatment
Abstract
Several factors are known to affect the performance of UV disinfection systems, including: flow rate, UV transmittance (UVT), lamp power, the concentration of viable target microbes in the undisinfected water (N0), and the UV dose-response behavior of the microbial target. Numerical models based on integration of computational fluid dynamics (CFD) and fluence rate field (I) simulations have been demonstrated to be able to account for the effects of variability of flow rate, UVT, and lamp power. As such, the focus of this research was on variations of N0 and UV dose-response behavior. UV254 dose-response behavior and initial concentration of viable bacterial indicators (E. coli, fecal coliforms, and Enterococcus ) and a viral indicator group (coliphage) were measured for undisinfected secondary effluent samples collected from operating wastewater treatment facilities. Corresponding water quality and operating characteristics (TSS, turbidity, BOD, flow rate, precipitation) were also quantified. Bacterial UV dose-response experiments for all three types of bacteria indicated that UV254 dose required to comply with effluent discharge regulations is considerably lower than is called for in conventional design criteria, which suggests that existing UV disinfection systems that have been implemented using these conventional design criteria are likely to be overdesigned. Phage dose-response data provided valuable information on UV254 sensitivity and variation of UV254 dose response behavior of coliphage in wastewater; this information is expected to be of immediate relevance, given that the EPA is planning to regulate infective virus concentration in wastewater in the near future. UV254 dose-response data were fit to the Phenotypic Persistence and External Shielding (PPES) model of Pennell et al. (2008) using two different regression methods. The results of these regression analyses allowed for quantification of variability in UV dose-response behavior, which in turn can be incorporated into CFD-I models via a stochastic method to estimate variability in overall disinfection process performance. This analysis was also used to develop linear correlations between PPES model parameters and water quality data.
Degree
M.S.
Advisors
Blatchley, Purdue University.
Subject Area
Microbiology|Environmental engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.