Standard Methodology for Transient Simulations of UV Disinfection Reactors

Using ultraviolet (UV) light to disinfect water has recently become more commonplace, and using computational fluid dynamics (CFD) is increasingly being used to model, optimize, and test UV disinfection reactors. The vast majority of CFD-based UV reactor models assume the steady-state because it was thought that transient phenomena will not lead to significant differences in performance estimates. Recent studies have shown that this assumption may not be valid under certain circumstances, especially when large eddy simulation (LES) is utilized. However, there is no standard methodology to conduct a transient simulation of a UV reactor. The objective of this paper is to propose a standard methodology to model the performance of UV reactors using transient CFD. A simple single-lamp cross-flow UV reactor was used to perform a series of computational experiments to develop a proposed standard methodology for transient simulations, and this method was then verified using two more-complex reactors. The paper examined mesh, time step, initial conditions, and particle tracking to determine model convergence using transient turbulent k - epsilon models. In addition, transient reduction equivalent dose (tRED) was introduced as a means to determine the variability in reduction equivalent dose (RED) over time in a UV reactor. The results showed that although the tRED does not vary significantly for the single-lamp reactor, there is upwards of 10% of variability in the RED for more-complex reactors. The proposed methodology developed in this paper leads to a structured and efficient approach to model UV reactor performance based on transient CFD. (C) 2016 American Society of Civil Engineers.