Performance modeling of stormwater best management practices with uncertainty analysis
Date
2009
Journal Title
Journal ISSN
Volume Title
Abstract
Best management practices (BMPs) contain many uncertainties that make it difficult to determine their performance with a model. Moreover, predicting BMP performance with existing methods is not easy. The major research objective of this dissertation is to incorporate uncertainty analysis in a BMP performance model to better represent its treatment performance. The k-C* model is used in this study to simulate BMP performance, and the study assumes that the influent event mean concentration (Cin) and aerial removal constant (k) include uncertainty. Both Cin and k represent data and model uncertainty. To evaluate the model, three different uncertainty cases, uncertainty in Cin, k, and both Cin and k, are applied to the total suspended solid (TSS) data of detention basins and retention ponds. To evaluate uncertainty values, three different uncertainty analysis methods, the derived distribution method (DDM), the first-order second-moment method (FOSM), and the latin hypercube sampling (LHS), are applied to each case. TSS, as a representative pollutant, and detention basins and retention ponds, as representative BMPs, are utilized in this study. The observed datasets are selected from the International Stormwater BMP database. By incorporating uncertainty analysis into the k-C* model, the effect of BMP surface area and inflow on the effluent event mean concentration (Cout) of TSS can be quantified for detention basins and retention ponds. These effects are not large in detention basins but are noticeable in retention ponds. In addition, the k-C* model with uncertainty analysis is applied to a hypothetical watershed to show how uncertainty might be used improve the probability of compliance with TMDLs.
Description
Rights Access
Subject
best management practices
stormwater management
total suspended solids
civil engineering
environmental engineering