CFD Modeling of a Storm-Water Hydrodynamic Separator

Proliferation of manufactured unit operations to separate particulate matter (PM) transported in storm water has generated interest in modeling of such units beyond conventional “% removal” and “black-box” type evaluations. One such unit operation is a screened hydrodynamic separator (HS) combining settling and to a lesser extent inertial separation by means of an annular static screen. However, performance and mechanistic evaluations of HS units are challenged by loadings with wide PM particle size distributions (PSDs) at a range of flow rates. Wastewater-based methods such as total suspended solids (TSS) and auto samplers further challenge such evaluations. This study utilized computational fluid dynamics (CFD) to model PM separation by a HS coupled with measured PSDs, flow rates [1–125% of hydraulic capacity (Qd)=15.9L∕s], and head loss through pilot-scale testing with representative sampling and material balances. The static screen was modeled as porous cylinder, a k-ε model accounted for flow turbulence, a Lagrangian discrete phase model examined PM fate, and the resulting CFD model was stable and grid independent. CFD results were validated with pilot-scale data across the flow range with a relative percent difference between model and measured data of less than 10%. For analysis or design, CFD facilitates a spatially distributed examination of PM fate as a function of PSD and flow rate. CFD results are contrasted to conventional methods with PM as a lumped gravimetric index (TSS) and HS as a lumped overflow rate type of best management practice.