Research Areas > Stormwater > Dynamics and Partitioning of Stormwater Particles

Project: Dynamics and Partitioning of Stormwater Particles

Background and Objectives

Because the majority of metal and organic contaminants in stormwater are associated with particles, assessment and management of pollutant loading requires an understanding of the behavior of suspended sediments. Previous research by SCCWRP and others has indicated that different pollutants are preferentially associated with specific particle size fractions. Moreover, particle density and size distribution can change dramatically over the course of a storm. The dynamics of contaminants associated with various particle sizes over the course of a storm have not been well described in southern California or elsewhere. Such information is lacking in four main areas:

The objective of this research was to characterize the particle size distribution of stormwater discharge from both characteristic land uses and mass emission sites, and then quantify the differential partitioning of pollutants of concern into various particle size fractions.

Status

This project was initiated in 2007 and completed in 2010.

Methods

An early objective of this research was to adapt and validate near-real time particle density analysis methods using an in-situ laser scattering particle analyzer (LISST). The LISST has been used in marine environments and for low volume analysis. However, application to urban stormwater has been limited due to the complexity of the stormwater matrix and difficulty of deploying it in a fast moving river. Thus, the first two years of this project focused on completing a series of laboratory and field experiments to validate LISST analysis under highly turbid, high volume pumping conditions. Once the sampling approach was perfected in the lab, it was tested and evaluated during three storms in Ballona Creek. The LISST’s measurements were then validated by comparing the instrument measurements of pumped stormwater particle sizes against filter masses of samples taken directly from the stream and pumped through the LISST.

Following successful method development, four storms were sampled at Ballona Creek over a two-year period, in which stormwater particulates and their associated pollutants were characterized. Samples were continuously analyzed throughout the storm by a LISST instrument which measures particle concentrations in 32 size classes. Collected samples were filtered to five size classes, and the particle concentration within each size class was calculated. Bacteria (Total Coliform, Fecal Coliform and Enterococcus) concentrations were measured for three size classes using IDEXX Enteroalert and Colilert trays. Additional storms were sampled in Ballona Creek to investigate metal partitioning into the predominant particle size fractions, and how those patterns vary within and between storms. Particle analyses from characteristic land use sites and BMPs were also conducted to aid in model refinement.

Findings

Researchers refined and tested the methodology for sampling highly turbid stormwater with the LISST. Method refinements included testing various approaches to sample collection (i.e pumped vs. grab samples), different dilution techniques, and evaluation of LISST performance versus traditional TSS measurements from filtered samples. The majority of particles found in stormwater runoff were sands but smaller particles represented a larger fraction during the early part of the storm. The LISST measurements correlated to the concentrations seen on the laboratory filters.

During small storm events (<0.3” rain), the highest concentration of metals and bacteria were associated with a <6 µm filter fraction, which accounted for 70% of the contaminant mass, but made up <20% of the total particle mass. The pollutant-particle association changed with storm size, as during a 2” rain event most metals were associated with >35 µm size particles. However, E. coli remain associated with small particles even during larger storm events. These results suggest that much of the contaminant load in storm water runoff will not be captured by the most commonly used best management practices, since the majority of these devices (e.g., hydrodynamic separators) are unable to capture particles <75 µm.

Partners

Loyola Marymount University

Presentation

• Stormwater pollutant-particle dynamics in Ballona Creek (Video)- January 2011 presentation to SCCWRP member agencies describing early results of particle size analyses.