Statistical characterization of stormwater runoff and environmental implications: Oil and grease, heavy metals

This dissertation examines stormwater quality from highways, and is a part of a larger study by the California Department of Transportation. As a first goal, 22 storm events were examined to determine when a single oil and grease sample (O&G) grab sample most closely approximates a flow-weighted composite sample. The analysis revealed that samples collected within the first hour of the storm overestimate the O&G event mean concentration (EMC) by 20 mg/L or more. The best time to collect a single grab sample ranged from 1 to 6 hours after the beginning of runoff. Correlations also exist between O&G EMC, antecedent dry days, and total rainfall. Correlating with chemical oxygen demand or dissolved organic carbon (DOC) EMCs collected by automated samplers may be the best method for estimating O&G EMC.; The second and third goals of the dissertation examines metals concentrations in highway runoff. Initially 59 storm events from the three UCLA sites were used. The analysis was then extended to 83 additional Caltrans sites. The first set of metal regressions analyzed the relationships between metal event mean concentrations and storm, site, and traffic characteristics, such as antecedent dry days, rainfall, average daily traffic, imperviousness and catchment area. Regression models developed for the individual and combined UCLA sites were generally good, (R2 ∼ 0.4 to 0.6) with total rainfall and antecedent dry days being the most significant parameters. The regression models were then applied to the Caltrans sites, grouped together by geographic proximity. The correlation coefficients were not as high but still significant, with total rainfall and average daily traffic being most significant.; The second set of metal regressions examined the relationship between metals and conventional constituents, such as total suspended solids (TSS) specific conductivity, and DOC. Regressions were generally good (R 2 ∼> 0.5 to 0.65), with TSS and DOC being the most significant.; The regressions provide a generalized way of estimating metals concentrations in highway runoff water quality from site and event specific parameters, or from conventional constituents. The regressions should be useful in assessing the potential impacts of highway runoff and for developing management plans or total daily mass loads.