Assessment of Reduction Equivalent Fluence Bias Using Computational Fluid Dynamics

Abstract

A study was performed to investigate the use of computational fluid dynamics (CFD) coupled with a fluence rate model to determine the bias in UV biodosimetry measurements using inactivation of a challenge microbe instead of the target pathogen—termed the reduction equivalent fluence (REF) bias. Simulations were performed using two fluence rate models: RADial Line Source Integration (RADLSI), and UVCalc3D, a multiple segment source summation (MSSS). CFD UV models were carried out in several closed-conduit UV reactors using either mono- or polychromatic lamps. Simulations were also performed with different UV transmittances, flow rates, lamp power settings, and with lamps out of service. In addition, the Cryptosporidium fluence response kinetics was approximated using two nonlinear models: a P-factor model, and a Rational model. The REF bias was computed using the predicted fluence distribution from a Lagrangian particle tracking routine and compared with the REF bias computed using the EPA UV disinfection guidance manual (UVDGM) approach. Overall, the results show that the REF bias using the predicted fluence distribution for the specific UV reactor was generally 45 to 66% of the REF bias predicted using EPA UVDGM. The REF bias was slightly sensitive to the selected fluence rate model. In addition, the REF bias was sensitive to the Cryptosporidium fluence response model when a large fraction of the fluence distribution was higher than 50 mJ/cm².