FRACTIONATION AND REMOVAL OF PARTICLES AND ASSOCIATED METALS IN A HIGHWAY STORMWATER DETENTION BASIN
Particles on the roadway surface play a major role in delivering and transporting chemical constituents in stormwater runoff. Particles entrained in highway stormwater runoff are typically comprised of small particles that exhibit a first flush. Additionally, organic and inorganic pollutants
tend to adsorb onto mineral particle surfaces and/or absorb into organic matter in surface coatings, often leading to an increase in pollutant concentrations with decreasing particle size. To improve water quality of highway stormwater runoff through effective application of best management
practices (BMPs), it is necessary to understand the distribution of pollutant concentration and mass across the particle size continuum and the extent to which particle removal through treatment processes (e.g., settling and filtration) relates to pollutant removal. Currently, little information
exists on the size-distributed mass of pollutants passing through highway stormwater BMPs, particularly in the smaller size fractions (e.g., <30 micron), and the appropriate monitoring protocols for resolving the size distribution of suspended particles and associated pollutants. Accordingly,
this study was initiated to determine the size distribution of stormwater runoff particles and the associated trace metals at the inlet and outlet of a highway detention basin using different methods of sample collection and size fractionation for subsequent elemental analysis.
stormwater runoff samples were collected at the inlet and outlet of a dry extended detention basin during four storm events from February 27, 2006 to April 17, 2006. The monitored detention basin is located in the San Gabriel River watershed along the intersection of interstate 605 and state
route 91 in the greater Los Angeles area. The 0.4-hectare watershed area is approximately 95% impervious surface and supports an annual average daily traffic volume of 187,000 vehicles. The detention basin was designed for treatment of a 25-mm storm within a 70-cubic meter water quality
volume and a 72-hour drawdown period. Hourly wholewater grab samples were collected at the inlet and outlet of the detention basin for analysis of particle size distribution (by number concentration) and elemental composition. These samples were further fractionated through sequential filtration
to determine the size-resolved elemental composition on particulate material captured by membrane filters of successively smaller pore sizes (100, 20, 8, and 0.45 microns), as well as in the filtrate (<0.45 micron) fraction of the sample. Composite stormwater runoff samples were also collected
for the purpose of immediately separating particles from stormwater runoff through a continuous-flow centrifuge. Centrifuged particles were further fractionated through gravitational settling. In settling experiments, a pipette method was used to withdraw subsamples for near-simultaneous measurement
of particle size distribution (by number concentration) and elemental composition. In addition, the size distribution of elemental concentrations was estimated based on theoretical settling velocities determined through Stoke's law and associated assumptions. The different methods of
sample collection and related fractionation were used to compare available techniques in measuring the finer particle size distribution and the associated metal pollutant mass in highway stormwater runoff.
Study findings indicate that a large proportion of elemental mass resides in the
fine particulate fractions of highway stormwater runoff at the inlet and outlet of the detention basin. Particle-based concentrations of suspended material retained by sequential filtration were typically highest in the 8–20 micron fraction. For example, this size fraction comprised
approximately 35 to 70% of total concentrations of copper and lead in the four monitored storm events. When converted to volumetric concentrations (in microgram/liter), the distributional patterns of high concentrations in the 8–20 micron fraction were preserved. In addition,
the filtrate fraction (<0.45 micron) comprised the largest proportion of concentration of the more soluble metals, such as cadmium, nickel, zinc, and copper, in most samples, and a relatively small proportion of the more particlebound metals, aluminum, iron, and lead. Comparing inlet and
outlet samples, the proportion of elemental concentrations associated with the two finest particulate fractions (0.45–8 and 8–20 microns) and filtrate fraction were generally higher in the outlet samples. Elemental fractionation based on settling experiments supported the general
patterns found by sequential filtration marked by a predominant distribution of concentrations of all metals (≥ 75%) in the fraction of particles less than 20 micron in diameter, which is even more pronounced in the outlet samples compared to inlet samples. Although most of the particulate
mass entering the detention basin is composed of particles larger than 20 microns, the size-resolved elemental concentrations measured in this study elucidate the preferential association of trace metals with fine particle fractions, particularly in the downstream region of the detention basin.
the limitations of a standard fractionation methods noted in this study, the general agreement between the two methods presented in this study provided a basis for elemental size distribution and partitioning of small particle size ranges in highway stormwater runoff transported through a
detention basin system. These findings highlight the importance of removing fine particles from stormwater runoff when improvement in downstream water quality is the target of stormwater treatment. Moreover, study findings were used to further evaluate the implications of BMP selection, placement,
operation, and design on effective treatment of stormwater runoff in the highway environment.
More about this publication?