Modeling of highway stormwater runoff

Abstract

Highways are stormwater intensive landuses since they are impervious and have high pollutant mass emissions from vehicular activity. Vehicle emissions include different pollutants such as heavy metals, oil and grease, particulates from sources such as fuels, brake pad wear and tire wear, and litter. To understand the magnitude and nature of the stormwater emissions, a 3-year study was conducted to quantify stormwater pollutant concentrations, mass emission rates, and the first flush of pollutants. Eight highway sites were monitored over 3 years for a large suite of pollutants. The monitoring protocol emphasized detecting the first flush and quantifying the event mean concentration. Grab and flow-weighted composite samples, rainfall, and runoff data were collected. A new runoff model with four parameters was developed that to describe the first flush of pollutants for a variety of rainfall and runoff conditions. The model was applied to more than 40 events for 8 pollutants, and the parameters were correlated to storm and site conditions, such as total runoff, antecedent dry days, and runoff coefficient. Improved definitions of first flush criteria are also presented.