Fate and Transport of Metals and Particulates within the Roadside Environment – A Review - page 3



Meso-scale mechanisms

Localized traffic conditions

Traffic generated winds can provide sufficient kinetic energy to keep small particulates in suspension in the turbulent region within traffic paths until they are blown far enough away to settle onto vegetation and soils as dry deposition. Re-suspension of particulates by road traffic during dry conditions is influenced by particle size, vehicle speed, and the number of vehicle passes (Nicholson and Branson, 1990). Deposited metals concentrations on vegetation and soils (Ter Haar, 1970; Dedolph, 1970; Daines and Motto, 1970; Ward et al., 1975; Ward et al., 1977; Wheeler and Rolfe, 1979; Ylaranta, 1995) exhibit characteristic decreases in concentration with distance from the road, as have directly measured atmospheric deposition (Backstrom et al., 2003). During wet weather, traffic generated spray scours the undercarriage of vehicles increasing metals and particulate release, with the intensity of the spray being a function of the traffic speed. In addition to releasing contaminants, spray influenced by traffic volume and speed can be a significant mechanism for delivering contaminant laden stormwater to roadside environments (Cristina and Sansalone, 2003).

Localized road corridor geometry

Metals and particulates may also settle and aggregate on the road surface (Harrison et al., 1981). Location of dominant particulate deposition and particle size distribution is a function of roadway geometry and transverse and longitudinal slope (Sansalone and Tribouillard, 1999). Transportation corridor geometries influenced by stands of trees have also been demonstrated to reduce the atmospheric transport of contaminant laden particulates by reducing wind velocities (Heath et al., 1999).


Micro-scale mechanisms

Factors influencing metal speciation and behavior

While a detailed description of the complex settling and scouring processes of dry and wet atmospheric deposition of metals and particulates to the roadside environment could be provided, it is beyond the scope of this paper. Delivery of these contaminants via stormwater runoff is more relevant and warrants discussion. Metals speciation in stormwater is a function of residence time, pH, particulate size and organic content of suspended particulates, alkalinity, salinity and temperature (Warren and Zimmerman, 1994; Sansalone and Buchberger, 1997a; Sansalone and Glenn III, 2000; Glenn III et al., 2001). Pavement type, either asphalt or Portland cement, may also influence metals speciation as a result of the influence of Portland cement on carbonate chemistry (Dean et al., 2005). These physical, chemical and temporal factors will cause the metal species to dynamically adsorb onto and desorb from particulates according to conditionally specific reaction kinetics and partitioning coefficients (Sansalone and Glenn III, 2000). If under a steady state, low alkalinity and slightly acidic conditions, partitioning equilibrium can be reached in as little as six hours (Sansalone and Buchberger, 1997a), with a small fraction existing in a dissolved phase and the major fraction being bound to particulates. The order of partitioning affinity for each metal species varies according to system specific conditions (pH, soil type, etc.) mentioned above.

First flush of contaminants

Metal species and particulates are lifted and scoured from the road surface, predominantly during the initial interval of a precipitation event, exhibiting what is characteristically known as a 'first flush' phenomenon (Sansalone and Buchberger, 1997a). Metals species may be primarily in the dissolved form in runoff at the pavement edge, and are more apt to exhibit a first flush phenomenon than particulates, as particulate transport is a function of runoff velocity and hence precipitation intensity (Sansalone et al., 1998). Initial transport of dissolved metals may therefore be considered flow driven while particulate transport is primarily intensity driven (Cristina and Sansalone, 2003). As smaller particles are more readily transported by lower rainfall intensities and smaller particles have greater surface area to volume ratios and thus a higher metal adsorption capacity, a dominant portion of the particulate bound metals fraction is also more strongly correlated with first flush phenomena as a result of being bound to more readily transported particulates (Sansalone and Buchberger, 1997b). However, mid range sized particles may actually have a greater specific surface area due to the rough physical characteristics of that particulate fraction and thus more adsorption sites for metal species (Sansalone and Tribouillard, 1999).

With short duration events, contaminant mass may not be fully washed from the roadway surface, resulting in a flow limiting event with regard to the complete and characteristic development of what is known as a 'pollutograph' – a graphical representation of contaminant concentration versus time. Conversely, long duration events may succeed in removing the mass of contaminants from the roadway surface deposited during the antecedent period, and a runoff event may be considered contaminant limited. With contaminant limited events, once antecedently deposited contaminants are removed from road surfaces, and as the contaminants generated during an event are concurrently flushed from roadways with runoff, the pollutograph asymptotically approaches equilibrium (Cristina and Sansalone, 2003).

Overland flow of metals to roadside environments

As stormwater leaves the pavement edge as overland flow, aqueous phase and particulate bound metals contaminants introduced to the roadside environment are subject to complex and dynamic functions of sediment transport, filtration, and adsorption onto soil matrices and existing vegetation. Hydraulically, sediment transport and erosive potential is a function of runoff velocity which is a function of precipitation intensity, roadside slope, and surface roughness. Surface roughness is dominated by the fraction of bare ground and litter cover (Hart and Frasier, 2003).

Much research has examined sediment transport through vegetation in both simulated laboratory experiments (Pearce et al., 1997; Munoz-Carpena et al., 1999; Deletic, 1999) and simulated field experiments (Pearce et al., 1998; Backstrom, 2002). Generally, greater slopes increase flow velocities, and higher velocities can transport larger particles. With increasing vegetated thickness (increased Manning number) sediment transport capacity is reduced (Gross et al., 1991). Velocity slows with greater roughness, and particles are subject to sedimentation processes, with the heavier (typically larger) particles settling out first, followed by smaller particles which settle out further down-slope as velocities are reduced.