BIOASSAY VALIDATION OF COMPUTATIONAL DISINFECTION MODELS USED FOR UV REACTOR DESIGN AND SCALE-UP

UV reactors for disinfecting drinking water are designed to deliver a specified dose of UV radiation. Dose in an ideal reactor is defined as the product of average UV intensity and residence time. In a real reactor, however, there is a distribution of residence times with a spatial dependence, interacting with a spatial intensity profile. This makes calculation of delivered dose very complex. As a consequence, computational disinfection models (which integrate fluid dynamic simulations, irradiation distributions, and microbial kinetics) have been employed to predict reactor performance and to optimize reactor design.We used computational disinfection models (CDM) in the design of a new drinking water reactor with nearly double the efficiency. The models were used to design a specific reactor with a predicted increase in disinfection dose of 2.2 times over that of an existing reactor with the same power input. Bioassay results confirmed a 1.9 fold dose increase at a particular operating condition. Scaled-up versions of the new reactor were designed using the above CDM and then bioassayed. The computational disinfection modeling and bioassays were done for two sizes of the new reactor design: one treating up to 3 mgd and one up to 19 mgd at the same dose. The CDM accurately predicted bioassay performance in both size reactors. The paper will further address the importance and advantages of using CDMs to accurately predict performance of larger reactors as compared to using conventional average dose calculations to predict scale-up.