AN ASSESSMENT OF THE BIOASSAY CONCEPT FOR UV REACTOR VALIDATION

While UV dose delivered by UV reactors is often expressed as the product of the average UV intensity within the reactor and the theoretical residence time, several authors have pointed out that hydraulic profiles and intensity gradients within UV reactors give rise to a distribution of delivered doses as opposed to a fixed value (Qualls et al, 1989; Scheible, 1985; Chiu et al, 1997). Recently, computational fluid dynamics (CFD) has been advocated as the best and possibly only way to fully characterize dose distributions delivered by UV reactors. In this paper, CFD predictions of dose distributions are made for several UV reactor configurations. Predictions of the inactivation of Cryptosporidium, E. coli, rotavirus, and MS2 bacteriophage by those reactors are made assuming first order kinetics and published inactivation constants. Predictions of fecal coliform inactivation by those reactors are also made using biphasic UV dose-response inactivation kinetics obtained with various wastewaters and measured using a collimated beam apparatus. The inactivation achieved is compared to collimated beam dose-response data to estimate the delivered UV dose that would have been predicted if the reactors were assessed using bioassay methods. It is shown that the bioassay dose varies depending on the inactivation constant of the challenge microbe and that this variance increases as the reactor deviates more from the ideal hydraulic condition of plug flow with radial mixing. The results show that UV doses predicted using bioassay methods may be easily misinterpreted if due consideration is not paid to ensure that the challenge microbe has inactivation kinetics similar to the indicator microbe or pathogen being targeted by regulations. An interpretation for bioassay results is proposed and its implication for compliance with drinking and wastewater regulations discussed. A modified bioassay procedure is recommended.