Development and Application of Particle Separation Efficiency Model for a Hydrodynamic Separator with Cumulative Probabilistic Distribution Functions

The presented study demonstrated that the particle separation efficiency (PSE) for a hydrodynamic separator (HS) as a function of particle size and flow rate could be best modeled using two-parameter cumulative gamma distribution (CGD). The developed PSE model provides continuous 3-D surface response with particle size and flow rate as two input variables. Goodness of fit test and model adequacy check by statistical analysis for model residuals are used to select the appropriate hypothesized distribution functions. It was found that modeled continuous performance curve did closely follow the experimentally measured PSE points across the range of tested flow rates for the tested HS loaded. With the developed PSE model for the HS and frequency distribution of influent hydrologic and granulometric loads, the statistical assessment of the tested HS in terms of particle removal capability, resulting alteration of PSDs across the BMP before discharge to downstream water bodies is provided. In addition, the uncertainty of performance of a particle separation BMP (i.e. statistical distribution of BMP effectiveness from a number of unsteady hydrologic events) is identified.