Application of a quasi-steady flow CFD model to simulate BMP response

In the last several years, computational fluid dynamics (CFD) has emerged as a tool to accurately predict the response of best management practices (BMPs) to highly unsteady rainfall runoff (stormwater), particulate matter (PM), and metal and nutrient loadings. Commensurately, such predictive capability for highly unsteady loadings can require much higher computational time than steady loadings. Therefore to reduce computational time for a given level of accuracy, this study tests a quasi-steady modeling approach to reproduce the unsteady load response of three stormwater BMPs. These units are a hydrodynamic separator (HS), a granular media-based volumetric clarifying filtration system (VCF), and a primary clarifier (PC). The quasi-steady steps were determined based on creating a frequency distribution of flow rates loading each BMP. The quasi-steady steps required to accurately represent the unsteady PM clarification were a function of influent unsteadiness and the mechanisms of each unit. Modeled PM clarification reproduced measured PM data from each physical model tested in situ with a relative percent difference (RPD) of less than 10%. Results indicate that with up to 15 quasi-steady steps the measured load-response data generated from unsteady rainfall-runoff events was reproduced for the VCF but generated deviation for the HS and PC. The quasi-steady method reduced computational time by an order of magnitude, several hours, compared with a fully unsteady method. Results suggest that additional quasi-steady steps are required to reduce deviation for the PC and HS.