Editor’s Note: This article first appeared in the September 2010 issue of Stormwater Magazine.
Polycyclic aromatic hydrocarbons (PAHs) commonly occur in the environment (Boehm 2006, Stout et al. 2004, Yunker et al. 2002), and as a result they are common constituents of municipal stormwater (Hwang and Foster 2006). Because PAHs are typically associated with particles, stormwater runoff can be a source of these compounds to surface waters and to sediments within those water bodies. The presence of PAHs in stormwater and sediments raises concerns because of the potential risk and impact of these contaminants on aquatic organisms. As a result of these risks, the presence of these compounds can result in regulatory demands for sediment assessment and remediation, or requirements for stormwater management controls or treatment. In addition, the presence of PAHs in sediments can result in increased costs of sediment remediation and pond solids disposal. The purpose of this article is to describe the class of compounds known as PAHs, to describe the potential sources of PAHs to urban stormwater and sediments, and to describe methods that can be used to differentiate between sources of PAHs found in sediments. In addition, because of the recent interest in the issue of coal-tar-based pavement sealants as a source of PAHs (Mahler et al. 2005, DeMott and Gauthier 2006), research on this specific topic is reviewed. Methods for treating stormwater are discussed.
PAH Chemistry
PAHs are a subset of the class of organic compounds known as hydrocarbons, which contain only hydrogen and carbon. The term aromatic refers to the presence of a six-carbon ringed compound, benzene. Further, PAHs, as their name implies, are made up of two or more fused benzene rings. Regulated PAHs range from the two-ringed compound naphthalene to the six-ringed benzo(ghi)perylene. When the US Environmental Protection Agency (USEPA) began its regulatory mission, it designated 16 specific PAHs as “priority pollutants.” Of this list, the National Toxicology Program’s Report on Carcinogens lists six as reasonably anticipated to be carcinogens (NTP 2005). These six and an additional PAH are considered by the EPA as probable human carcinogens (USEPA 1993). These 16 compounds are unsubstituted or “parent PAHs,” which means that they contain no hydrocarbon side chains, also known as “alkyl” side chains. Though most regulatory focus has been on the 16 priority pollutants, alkyl-substituted PAHs are abundant in the environment.
The relative concentration of individual PAHs in environmental samples and the relative amount of alkylated versus parent PAHs found in environmental samples depends on the source(s) of the PAHs. When high temperatures (i.e., greater than several hundred degrees Celsius) are involved in a process that produces PAHs, these high temperatures tend to eliminate the alkyl side chain, so processes such as fossil fuel combustion (in gasoline or diesel powered vehicles, or in wood burning), the manufacture of steel, and coal gasification result in a predominance of unsubstituted or “parent” PAHs. Specific known sources of such combustion-related or “pyrogenic” PAHs include combustion-linked atmospheric deposition from diesel and automobile vehicle exhausts, smelters, power plants, legacy manufactured gas plants, creosote from wood treatment applications, forest fires, and residential wood burning. The effect of temperature on the formation of PAHs is elegantly explained in Lima et al. (2005). Crude petroleum and refined petroleum products and unprocessed coal are also major potential sources of PAHs. However, unlike pyrogenic PAHs, these fossil fuel or “petrogenic” PAHs are formed at lower temperatures over a longer period of time. As a result, petrogenic PAH sources are characterized by a higher relative abundance of alkyl-substituted PAHs (Figure 1). Petrogenic sources to stormwater commonly include motor fuels, lubricants, asphalt, and petroleum-based solvents.
Sources of PAHs in Stormwater and Urban Sediments
Sources of pollutants to urban waterways and sediments are numerous (Boehm et al. 2002) and include municipal and industrial effluents and discharges, stormwater runoff whether direct or as part of combined sewer overflows, direct deposition on water, and spills, among other pathways. With regard to stormwater pathways in particular, atmospheric deposition on roadways and other paved surfaces has been found to be a major source of PAHs in stormwater and sediments (VanMetre et al. 2000, Stein et al. 2006). The mass (Moon et al. 2006) and specific chemistry (Yunker et al. 2002, Boehm 2006) of the PAHs deposition depends on land use, urban density, and industrial processes, among other factors. Combustion products from transportation, power generation, and coking processes are significant manmade sources that result in atmospheric deposition. Other potential sources of pyrogenic PAHs to stormwater are the runoff from historical sites such as manufactured gas plants, wood treatment facilities, and smelters.
In addition to combustion, releases of petroleum can also be a source of PAHs in stormwater. Petrogenic PAHs can originate from point sources–such as oil spills or petroleum handling facilities–from asphalt pavements, and from more dispersed sources such as leaking oils from individual vehicles (Boehm 2006). Typically, the percent of PAHs in petroleum products ranges from zero to about 20% by weight. While uncombusted gasoline contains almost none, PAHs are present in diesel fuel, home heating oil, motor oil and other lubricants, and crude oils. In addition, as crankcase oil is used and heated, the concentrations of PAHs increase with time (Pruell and Quinn 1988). Because of the role of combustion products in atmospheric deposition, it has been observed that the average PAH flux (g/m2) from urban areas is up to 45 times as great as flux from undeveloped areas (Stein et al. 2006). In addition, mass per unit time of PAHs flowing into stormwater systems per storm event increases with the amount of time between events (Stein et al. 2006). In areas with wet and dry seasons, the early wet-season storms have the highest stormwater PAH concentrations. Concentrations in stormwater also change over the course of a storm. Because much of the PAHs in a watershed are associated with small particles, they are quickly mobilized in the first flush of a storm, resulting in the highest concentrations early in a rain event. In one study, up to 60% of the flux of PAHs resulted from the first 20% of the storm flow (Stein et al. 2006).
The presence of PAHs in stormwater is important because stormwater is a major delivery system of PAHs (and other contaminants) to the sediments of rivers, harbors, and bays. PAHs are ubiquitous in sediments and, as such, pose potential environmental risks. As a result, the chemistry, behavior, and sources of these compounds have been the subject of significant research. The general PAH compositional makeup or pattern in sediments is quite similar across urban systems and is commonly known as “urban background” (Figure 1). Urban background consists primarily of the larger pyrogenic compounds, with lesser concentrations of mid-sized pyrogenic and petrogenic contributions. Based on typical concentrations found in urban sediments, some have suggested that total background priority pollutant PAH concentrations (i.e., unrelated to specific sources) range up to about 20 mg/kg (parts per million) (Stout et al. 2004).
The primary sources of PAHs to this urban background relate to atmospheric deposition of combustion products within the watershed (VanMetre et al. 2000). As summarized by Stout et al. (2004), “Previous studies have long since and unanimously demonstrated that the PAHs in urban runoff are largely associated with combustion-derived particulate matter.” These sources can be ascribed largely to emissions from transportation, industry, and power generation. Although direct atmospheric deposition of particulate PAHs onto water bodies occurs to a significant extent, for the most part this atmospheric deposition finds its way into urban sediment via deposition on land and transport via stormwater systems. Regional differences in these processes result in some differences in the specific source patterns.
Point sources, such as manufactured gas plants, creosote-treated piers and docks, coking units and smelters, and coal-fired power plants, are also common on urban waterways and can result in localized PAH hot spots in sediments because of the discharge of wastes and wastewaters.
Manufactured gas plants were common facilities in towns and cities across America, with estimates of more than 50,000 individual sites (Hatheway 2009). The wastes from these plants are known to have impacted many sediment systems including the Hudson River (Khalil et al. 2006), Oregon’s Portland Harbor (Sower and Anderson 2008), Lake Champlain (Kreitinger et al. 2007) and urban creeks in Austin, TX (Carmody and Ward 2003). Activities related to smelters and aluminum production have also been shown to result in PAH-contaminated sediments (Booth and Gribben 2005). Although contaminated sediments associated with these facilities have been remediated at some sites, issues such as increasing environmental exposure as a result of suspending impacted materials and concerns about recontamination of treated areas by impacted background sediments need to be considered when making remedial
decisions (Sower and Anderson 2008).
Regulatory Issues
The presence of PAHs in both stormwater and sediments often raise regulatory concerns. More stringent application of two sections of the Clean Water Act (CWA) has implications for stormwater managers. First, under CWA §402, discharge permits can be required for stormwater systems. Depending on the specific permit condition, there can be concentration limits for specific PAHs. Permits can impose treatment requirements if the limits are exceeded. In addition, under CWA §302, states must designate a water body as “impaired” if it exceeds established water-quality standards. Such a designation can result in requirements to set a total maximum daily load (TMDL), which limits the amount of a pollutant that the water body can receive. TMDLs for PAHs have been proposed or approved for a number of surface water systems across the country (DC DOH 2004, USEPA 2005, VA DEQ 2007). Once TMDLs are established, management plans must be developed to achieve the identified requirements. Because the PAH load in surface water systems is influenced by historical, existing sources such as contaminated sediment hot spots (Sower and Anderson 2008), a TMDL approach may have limited effect on overall water quality as it is focused only on reducing future inputs.
Urban sediments have been a major focus of the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), commonly known as Superfund. For example, the Commencement Bay (WA) Superfund site was designated as such in 1983 and has been divided into many sections known as operating units, all of which have focused on sediments and sources of PAHs and other contaminants. Many other complex aquatic systems have been designated as Superfund sites: Lower Passaic River, Hudson River, Portland (OR) Harbor, and many others. Because stormwater systems are implicated as sources of contamination in all of these Superfund sites, municipalities may incur environmental liability (Luria 2002). A federal court recently found that a highway department could be liable under CERCLA for the discharge of stormwater that contained hazardous substances (United States v. WSDOT, No. C08-5722, slip op. [W. D. Wash. June 7, 2010]). Being named a Potentially Responsible Party can result in significant assessment and remediation costs. The need to minimize future liability and prevent recontamination of remediated sediments will increase pressures to reduce sources of PAHs and other contaminants in stormwater.
What, When, Where, Who? Environmental Forensics of PAHs
Because PAHs have many different sources, and because the composition of PAHs (numbers, types, concentrations, and ratios of individual PAH compounds) in these different source types and specific sources is varied yet established in the literature, these differences in the relative concentrations of the individual PAHs make it possible to identify and distinguish sources by using appropriate analytical and data evaluation methods (Boehm 2006). The application of such tools is a key component of the field of environmental forensics. Traditionally, only some or all of the 16 priority pollutant PAHs are measured for regulatory purposes, but these have limited value for source identification. By including other parent PAHs and by including an extended list of alkyl-substituted PAHs, it is possible not only to readily separate pyrogenic and petrogenic sources, but also to distinguish different petrogenic sources from each other and from pyrogenic sources as well. An environmental forensic approach to PAH source identification and apportionment is to analyze a suite of 30 to 50 compounds inclusive of one carbon (C1) through four carbon (C4) substituted isomers of the two- through four-ringed PAHs in addition to the priority pollutant PAHs.
Data analysis approaches for evaluating the results of forensic chemical analyses include concentration histograms, diagnostic ratios, double ratio plots, and multivariate statistical analysis (Boehm 2006). Concentration histograms, sometimes called “fingerprints,” are bar graphs showing the relative concentration of individual compounds–i.e., the overall pattern. As shown in Figure 1, the PAH histogram of a petrogenic crude oil is clearly distinct from that of pyrogenic tar from a manufactured gas plant or from sediments impacted by urban background. Diagnostic ratio analysis depends on knowledge of how the relative concentration of specific PAH pairs differ depending on their source. There is published information on a wide range of diagnostic ratios for numerous natural and manmade sources of PAHs (Yunker et al 2002, Boehm 2006). Double ratio plots take this concept further by allowing the comparison of results within a sample set to each other and to potential sources (Figure 2). By creating plots with the ratio of one pair of PAHs on one axis and another pair on the other axis, the results of the entire data set evaluated. Samples from similar sources group together and may be separated from samples with different source ratios. These approaches have been carried further into the differentiation among similar sources as well.
Multivariate statistical analysis methods serve to both explore data sets (i.e., examine similarity in groups of samples) (Johnson et al. 2007) and to quantitatively apportion the PAHs in an environmental sample to its component sources (e.g., Burns et al. 2006). The output of exploratory techniques such as principal component analysis (PCA) includes information on which samples are similar and which compounds are most useful to distinguish sources, and evidence concerning the relative contribution of these sources to each sample. When the output of such methods is combined with mixing models, it is possible quantify the impact of various sources. These approaches were used to derive the contribution of unburned petroleum, motor vehicle exhaust, and other combustion sources to PAHs in surface water sediments (Sofowote et al. 2008).
Pavement Sealant as a Source of PAHs
Pavement sealants have been discussed as a source of PAHs in stormwater and sediments (Mahler et al. 2005, Gauthier and DeMott 2008, Richardson 2006). These products are used to protect and extend the life of asphalt and are typically used in parking lots or on driveways. There are two basic types of sealants, those made using a petroleum base and those made using a distilled coal tar. The refined coal tar used in pavement sealers is a byproduct of steel manufacturing and can consist of greater than 10% PAHs by weight. To reduce the amount of PAHs entering surface waters and to reduce disposal issues for retention pond solids, some municipalities, such as Austin, TX, and Washington, DC, have recently banned the use of refined coal-tar-based sealants.
Some of the research on this issue has been done by a US Geological Survey (USGS) research team in Austin, TX (e.g., Mahler et al. 2005, VanMetre et al. 2009). The results of these studies indicate that the concentration of PAHs in runoff from parking lots with refined coal-tar sealants is greater than from unsealed lots and from those sealed with a petroleum-based product. As PAHs are associated with solids, physical and chemical wear of the seal coats result in PAH-enriched particles that become mobile during rainfall. While refined coal-tar sealants can lead to elevated concentrations in water collected from treated parking lots, their impact on the overall loading of PAHs in a watershed and their contribution to PAHs found in urban sediments is far less certain (DeMott and Gauthier 2006, Mahler et al. 2006). One study has suggested that sealants can be a significant source of PAHs in streams (Mahler et al. 2005). Paired sediment samples taken upstream and downstream of sealed parking lots immediately adjacent to creeks detected elevated downstream PAH concentrations in some, but not all, creeks (City of Austin 2005). The effect decreased quickly with distance from the lot and with the time since a lot was sealed. An investigation of one small (< 2.8 km2) Austin watershed found that the PAH and coal-tar-pitch concentrations of commercial site soils and local sediments were similar, indicating that sources other than sealants may be involved (Yang et al. 2010).
Mahler et al. (2005) used double ratio plots to attempt to link sealant and sediment chemistry. However, the study did not include other known sources of stormwater PAHs such as atmospheric deposition or combustion by-products, so other potential PAH sources could not be eliminated (DeMott and Gauthier 2006). Using data from a similar study conducted by the USGS in Madison, WI (Selbig 2009), a double ratio plot was generated to compare the runoff from parking lots, roads, and roofs to the PAH profile of refined coal-tar sealants (Figure 2). These results indicated that sources other than the sealants very likely contribute to the PAHs in stormwater, as the sealant could not account for the distribution of ratios found in surface runoff and urban pond sediments. Another application of a PAH environmental forensic approach is shown in Figure 3. In this figure, the PAH histogram of a representative Austin creek sediment sample (DeMott et al. submitted) was compared to a PAH histogram of a refined coal-tar-based sealant. It indicated that while the PAHs in the sediment were primarily the four- to six-ringed compounds, as is typical for atmospheric deposition influenced urban sediments, the sealant had a higher ratio of the small three-ringed compounds. Differential weathering (defined as changes in PAH composition resulting from physical, chemical, and biodegradative processes after environmental release, in this case application of the sealant) of the sealant PAHs (that is loss of the three-ringed PAHs relative to the higher molecular weight, and less readily degraded four-, five-, and six-ringed PAHs) cannot account for the differences observed in Figure 3. Thus, the sources are very likely different.
The results of another study conducted two years after Austin’s coal tar sealant ban indicated that the ban has not resulted in an expected decrease in the sediment PAHs (DeMott et al. submitted). Even though hydrological conditions result in a rapid replacement of creek sediments, there has not been a noticeable decrease in PAH concentrations, implying that the PAH sources to the sediment have been constant and unrelated to coal-tar-sealant usage. Forensic analysis using PCA indicated that the sediments collected before and after the sealant ban group together, but are distinctly different from the PAH composition in the runoff of coal-tar test plots. Another study examined the linkage of sealant usage to the PAH compositions in sediment across various regions of the US (VanMetre et al. 2009). The authors found that the pyrogenic PAH signal is greater in samples collected in Midwest and East than in the West. They attempted to related this difference to sealants based on the claim that “coal-tar-based sealants are reported to dominate use” in these areas, but no actual data on regional differences in sealant use were provided. The lack of consideration of other regional differences, such as the use of coal versus natural gas for power generation and the density of heavy industry, including legacy manufactured gas plants, lessens the relevance of the authors’ findings to the debate concerning sealants. Examples of differences in the source and chemistry of atmospheric depositional load is indicated by comparison of the results from regional studies. Combustion products from power generation and steel production are responsible for as much as 86% of PAHs in Chesapeake Bay sediments, while automobile exhaust makes up the remaining 14% (Dickhut et al. 2000). Similarly, in the Midwest, coke oven emissions have been shown to be the source of up to 90% of the PAHs in some freshwater systems (Su et al. 2000). But in the West, which has significantly fewer coal-fired facilities, atmospheric deposition from mobile sources such as cars and trucks are the primary source of PAHs in stormwater and sediments (Stein et al. 2006).
In summary, the studies conducted by VanMetre and Mahler have raised important questions about the role of pavement sealants as sources of PAHs in stormwater and sediments, but they are far from definitive or rigorous. Their work suggests that freshly sealed pavement can result in runoff with elevated PAH concentrations and may serve as a short-term point source of PAHs to adjacent streams. However, the use of more rigorous environmental forensic methods, including more extensive sampling, PAH forensic chemical and data analysis methods, and a consideration of other very similar pyrogenic sources, needs to be undertaken. As the relative contribution of these materials to the overall urban PAH loading is inconclusive, additional research is required to further assess the potential environmental impact of pavement sealants.
Treatment of PAHs in Stormwater
Best management practices (BMPs) for the treatment of stormwater are outlined in a USEPA guidance document and can roughly be grouped into two approaches: 1) source control, and 2) treatment control (Clar et al. 2004). A variety of effective and straightforward stormwater treatment control technologies exist including ponds, wetlands, infiltration, vegetative biofilters, and filters. The traditional focus of the different treatment technologies has been mostly the removal of suspended solids, nutrients, oil, and grease in addition to metals. The ability of some of these technologies to remove PAHs has not been investigated until quite recently.
Because PAHs are largely associated with particulates (DiBlasi et al. 2009, Hwang and Foster 2006), a treatment technology that reduces particulates would also be effective in reducing PAHs. The most often employed BMPs, ponds, mainly remove contaminants by settling out particulates, although other physical, biological, or chemical mechanisms also contribute (Clar et al. 2004). Filtration technologies rely on filtering out the particulates in addition to sorption and biological mechanisms. Examples of these include bioretention cells, multichambered treatment trains (MCTTs), and proprietary stormwater filtration devices.
Bioretention cells consist of layers of soil, sand, and organic matter. In a recent study, bioretention cells reduced PAHs concentrations by 31% to 99% and mass discharge by an average of 87% (DiBlasi et al. 2009). An underground MCTT has been developed to treat stormwater from relatively small-scale critical source areas, including vehicle service facilities, parking lots, paved storage areas, and fueling stations (Pitt et al. 1999). The MCTT consists of a catch basin for removal of large particles, a settling chamber for fine sediment removal, and finally a mixed-media filter of sand and peat (or other combination) designed to polish the final effluent by removing some of the soluble constituents.
Conclusions
PAHs are ubiquitous and have multiple sources in urban environments. As a result, PAHs are common constituents in stormwater and in urban sediments. While the atmospheric deposition of combustion products is typically the most significant pathway that introduces PAHs to urban sediments, a number of other point and nonpoint sources also contribute. Similar PAH compounds originate in multiple sources types, thus underscoring the need to carefully and rigorously differentiate sources in environmental studies. This is especially true for the coal-tar types of PAHs. Because PAHs are regulatory compounds of concern, their presence often raises issues and drives sediment remediation under both state and federal law, and as a result may result in the need for stormwater treatment and source management. Successful management requires an understanding of the critical sources within a watershed; a number of environmental investigation and forensics tools are now commonly used to identify and apportion PAHs in sediments to component sources. While pavement seal-coating can introduce PAHs into runoff and have measurable localized contributions, the available literature suggests that other sources, especially combustion sources, represent the vast majority of PAH inputs to urban sediments.
