A Granulometry-Based Selection Methodology for Separation of Traffic-Generated Particles in Urban Highway Snowmelt Runoff

Abstract

Traffic activities generate a wide gradation of anthropogenic particulate matter that is entrained in highway snow and snowmelt. Granulometric-based analyses of this particulate mattercan provide guidance for selection and design for snowmelt solid/liquid separation processes. This study presents a granulometry-based analysis of snow melt particles from 10 highway shoulder sites in urban Cincinnati (State of Ohio, U.S.A.) generated from a 46 cm, 48 hr snowfall with a highwayshoulder residence time of 102 hr. Each site was exposed totraffic and maintenance activities (plowing and de-icing saltsonly). Despite variations in gradations and particle indices, measured lnv-Nt (number-volume mean size-particle counts) values for all sites were closely modeled using a two-parameter power law model. The model exponent, β, ranged from 2.22 to 2.64 indicating that a significant fractionof surface area is associated with the coarser gradation of particles. Modeled results support previous experimental data indicating the predominance of surface area and heavy metal mass is associatedwith the coarser fraction of particles. Results indicated a statistically significant difference in densities for coarse (2.86 gm cm-3) and fine particles (2.75 gm cm-3) (P < 10-11).A methodology is presented in the form of a process selection diagram to evaluate mechanisms of particle separation. It was determined that sedimentation could remove 90% of particles, bymass, within 2 hr for a typical roadway drainage design. The influence of particle specific gravity on sedimentation efficiency is illustrated for a range of specific gravities from 1.1 to 4.0. Although the very high concentrations of particulate matter in highway snow melt do not make it amenableto straining or direct filtration, such processes may be feasible as a secondary treatment for snowmelt or storm water effluent discharges to receiving waters.