• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

An Integrated Case Study for Evaluating the Impacts of an Oil Refinery Effluent on Aquatic Biota in the Delaware River: Advanced Chemical Fingerprinting of PAHs

Posted on: Thursday, 6 October 2005, 06:00 CDT

By Uhler, Allen D; Emsbo-Mattingly, Stephen; Liu, Bo; Hall, Lenwood W Jr; Burton, Dennis T

ABSTRACT

More than one thousand samples were collected and analyzed to evaluate the potential impact of Motiva's oil refinery effluent on the receiving water, sediment, and biota of the Delaware River. The data collected from these samples were used with advanced chemical fingerprinting of polycyclic aromatic hydrocarbons (PAHs) in Motiva's oil refinery effluent to differentiate Motiva-related PAHs in sediment and biota from other sources. The PAHs released from the refinery between 1999 and 2002 were dominated by petrogenic 4-ring PAHs. Specifically, the refinery signature exhibited relatively high levels of fluoranthenes/pyrenes with two (FP2) and three (FP3) alkyl groups and benz(a)anthracene/chrysenes with two (BC2), three (BC3), and four (BC4) alkyl groups. This PAH signature, attributed to accelerated degradation of low molecular weight PAHs in the Motiva wastewater treatment plant, exhibited little variability over time relative to the background patterns in the Delaware River. This distinctive feature of the Motiva effluent allowed the identification of this source in other samples. Water and sediment samples identified a range of PAH characteristics associated with the Delaware River urban background signature. These characteristics included varying levels of 2- to 3-ring PAHs (likely from weathered automotive fuel, marine fuel, or bilge tank discharges), pyrogenic 4- to 6-ring PAHs (from partially combusted organic material like soot), and perylene (diagenetic product of terrestrial plant decomposition). The Motiva hydrocarbon signature was only evident at moderate to low levels in selected near-field sampling stations for sediment, bivalves, and effluent/near field water. PAHs in the river sediments beyond the near-field area were consistently associated with samples containing the Delaware River urban background signature, and exhibited little to no effect from the Refinery.

Key Words: PAH fingerprinting, oil refinery effluent, Delaware River.

INTRODUCTION

Polycyclic aromatic hydrocarbons (PAHs) are ubiquitous contaminants in coastal and riverine sediments. Because PAHs are low- level constituents of crude and many refined petroleum products, concern has been raised about the potential for inputs to the Delaware River of petroleum-derived PAHs arising from wastewater discharges from the Motiva Refinery as described in Hall and Burton (2005). However, because PAHs found in river sediment can arise from numerous sources, detailed analyses of the PAH chemical signatures in the sediments must be carried out in order to determine the source signatures of these hydrocarbons. Such a forensic study is commonly referred to as a "chemical fingerprinting" investigation.

This article describes the results of detailed chemical analysis of PAHs found in environmental samples that were collected in support of a number of investigations conducted for the Delaware City Refinery case study as described in this series of articles (Table 1; see Hall et al. 2005; Salazar et al. 2005; Alexander et al. 2005; Alden et al. 2005 in this special issue). In particular, PAH data and forensic interpretation are described for the following investigations: monthly water monitoring; sediment survey study; sediment Triad study; sediment core study; and bivalve study.

The background and technical approach for each of these studies are described in separate articles in this special issue. In this article, we focus on interpretation of the environmental chemistry of PAH compounds-vis avis the nature and source (s) of those PAHs- as they pertain to each of the studies.

BACKGROUND

The sources of PAHs are difficult to determine in urban waterways due to their ubiquitous nature in the environment. However, it is possible to determine the sources through an understanding of the fundamental environmental chemistry of PAHs, the chemical patterns of PAHs in various source materials (e.g., materials containing PAHs that find their way into the sediments), and the changes these chemical patterns undergo as the PAHs alter, or weather, once in the environment. In addition, a firm understanding of the chemical features of ubiquitous, non-point source-derived PAHs found in sediments (referred to as Urban Background) is essential for understanding the results presented in this article. As such, we present a brief background on the environmental chemistry of PAHs prior to the discussion of the fingerprinting investigation in the Delaware River.

Table 1. Studies and numbers of samples analyzed for PAH compounds.

General Chemistry of PAHs

Recognizing the sources of PAHs can be difficult in sediments and it requires a basic understanding of their chemistry and nomenclature. As their name implies, /jolycyclic aromatic / iydrocarbons literally: (1) contain multiple ring structures, (2) which are aromatic in nature, and (3) comprised of hydrogen and carbon. Naphthalene, consisting of two fused benzene rings, is the simplest PAH. The arrangement and number of fused rings is used to distinguish different PAHs (Figure 1).

Some PAHs contain carbon side chains of varying numbers, lengths, and locations. These carbon chains are termed "alkyl groups." PAHs that do not contain alkyl groups are termed "nonalkylated" or "parent" PAHs, such as naphthalene. PAHs with one or more of these groups are said to be "alkylated." Other heteroaromadc compounds co- occur with PAHs. These compounds contain either a nitrogen, oxygen, or sulfur atom, such as dibenzothiophene.

PAHs are rarely found in the environment as single compounds; rather, petroleum- or combustion-derived PAHs occur as groups of 2- , 3-, 4-, 5-, and 6-ring PAH compounds and their alkylated derivatives (commonly referred to as "alkyl homologues"). The relative distribution of the 2- through 6-ring PAHs in an environmental sample is a function of the nature or type of material (s) that are the source of the PAHs; similarly, the relative distribution of the alkylated homologues among PAH compounds depends on the distinct chemical features of the source materials. Examples are given in this article that illustrate how these diagnostic chemical features can be used to catalog and discern among different PAH source materials in the environment.

Source Categories of PAHs in Urban Sediments

Sources of PAHs in urban sediments can be separated into several categories. They originate from a large number of sources, which can be broadly classified as either (1) diagenetic, (2) petrogenic, or (3) pyrogenic. Diagenetic PAHs arise from rapid (years) rearrangement of natural hydrocarbons that are not ordinarily recognized as significantly impacting sediment quality (Wakeham et al. 1980). Petrogenic sources are anthropogenic (def., derived from man's activities) sources of PAHs that are derived directly from crude oil or refined petroleum products. Pyrogenic sources are anthropogenic sources of PAHs that include those derived from fires, combustion of petroleum products, combustion and conversion of coal, roasting of organic matter, and metallurgical processing.

Figure 1. Structures of the parent (non-alkylated) 2- to 6-ring EPA Priority Pollutant PAH compounds. see Table 2 for abbreviations.

Sources of Background PAHs in Sediments

PAHs are introduced into sediment environments through a variety of anthropogenic activities from both point and nonpoint sources that may exist along urban waterways. Common point sources of PAHs in many urban sediments include: direct or indirect discharges from refineries, petroleum terminals, shipyards, aluminum smelting, manufactured gas production facilities, tar distillation plants, rail yards, loading/unloading facilities, treated wood pilings, marinas, discharge canals, and stormwater outfalls. These point sources can spill or seep petroleum, tar, and distillation products at various rates into the environment. Common nonpoint sources include atmospheric (soot) particulates and dripped/leaked petroleum washed from the surrounding urban roadways, parking lots, vegetation, and structures during rainfall events. Other nonpoint sources of PAHs in urban waterways include recreational boat traffic, ship traffic, general surface and stormwater runoff (i.e., not entering from a specific outfall location), direct atmospheric particulate deposition (soot from petroleum combustion, forest fires, wood stoves, coal-fired power plants smelters, etc.), and rainout of vapor phase PAHs to the waterway.

On a global basis and in areas remote from urban influence, background PAHs generally are limited to pyrogenic PAHs derived from atmospheric particles transported over large distances (Ohkouchi et al. 1999). In selected geologically active environments, oil seeps and erosion from petroleum source rocks and coal can result in elevated concentrations from natural sources of petrogenic PAHs (Boehm et al. 200Oa,b). The concentrations of background PAHs in these remote areas are generally much lower than background PAH concen\trations in urban waterways, where direct deposition of combustion-related PAHs from proximal sources and urban runoff have occurred for much of the last century.

In urban sediments, background PAHs associated with pyrogenic sources usually are more abundant than those associated with petrogenic sources, due to the high volume of fossil fuels combusted in urban areas (Gonzalez et al. 2000; Van Metre et al. 2000; Stout et al. 2001; Wu et al, 2001; Mai et al. 2001). Concentrations of pyrogenic PAHs normally are highest in upper sediment layers, and decrease to a relatively constant "natural background" concentration at depths corresponding to deposition prior to urbanization. In most settings, the most highly PAH-contaminated sediments were deposited between the start of heavy industrial use of fossil fuels and the present, indicating that the fraction of the total "urban background" PAHs above natural background levels is derived from the combustion of fossil fuels. Forest and brush fires probably are major sources of the "natural background" pyrogenic PAHs deposited before the industrial revolution.

Background petrogenic PAHs are also present in most urban sediments, although they usually are less abundant than background pyrogenic PAHs. Their occurrence is largely attributable to uncombusted petroleum spilled or dripped onto roadways and parking lots (e.g., crankcase oil) that enters a waterbody via runoff following storm events (Stout et al. 2001).

Overview of Advanced Chemical Fingerprinting

Advanced chemical fingerprinting (ACF) is an umbrella term for identifying and distinguishing hydrocarbons-and particularly PAH sources-in the environment. Numerous investigators have used detailed chemistry and chemical fingerprinting techniques to identify diagnostic features and differentiate among sources of hydrocarbons in sediments (Almini et al. 2003; Stout et al. 2001; Wang et al. 1999; Douglas et al. 1996; Sauer and Boehm 1995; Peters and Moldowan 1993; Volkman et al. 1992; Eganhouse et al. 1982; LaFlarnmc and Hiles 1978). Thse techniques were used in this investigation to understand the nature and source (s) of PAHs in sediments of the Delaware River near the Motiva Refinery.

The fundamental analytical tool in ACF of PAHs is detailed analysis of parent (non-alkylated) PAHs and their alkylated homologues using gas chromatography/mass spectrometry (GC/MS). It is important to distinguish between the utility of PAH measurements made following standard EPA compliance methods of chemical analysis versus those accomplished using ACF. The USEPA lists 16 nonalkylated PAHs of principal environmental concern on its Priority Pollutant List, and has developed a standard method (SW-846 Method 8270, Semi- Volatile Organic Compounds by Gas Chromatography/Mass Spectrometry) to analyze for these PAHs in various media. However, hundreds of PAHs exist, and many of these are essential to PAH source investigations as diagnostic tools to differentiate target PAH analytes from complex mixtures. Table 2 lists common PAH analytes targeted during PAH source investigations (and utilized in the investigation described in this article); the 16 USEPA Priority Pollutants are indicated by asterisks.

One common ACF technique (employed in this investigation) is to modify USEPA Method 8270, which targets only the 16 Priority Pollutant PAHs in its standard form, and ignores all alkylated PAHs and related PAHs. In the modified method, the GC is operated with a very slow oven temperature program to optimize separation of target PAH compounds, and the mass spectrometer is operated in the selected ion monitoring (SIM) mode to minimize interferences from non-target compounds and lower the detection limits. These simple modifications reduce PAH detection limits from 660 jug/kg (wet) for the standard method to approximately 1 /Ltg/kg (dry) for the modified method. These lower detection limits for PAHs are particularly important when attempting to define background concentration levels. In addition, modified Method 8270 measures the ions characteristic of the PAHs and PAHs useful in PAH source investigations. Internal surrogate and recovery standards are used to measure performance and concentrations, relative to an external calibration solution containing the parent PAHs on the target analyte list. The response factors for the parent PAHs are applied to the appropriate alkylated PAHs. This method is fully described in the NOAA Status and Trends methodology and numerous peer-review publications (e.g., Sauer and Uhler 1994; Page et al. 1995; Boehm et al. 1997; Stout et al. 2002).

Another ancillary ACF technique sometimes used to compliment direct measurement of PAHs is a modified USEPA Method 8015B, which employs gas chromatography in combination with flame ionization detection (GC/FID). The primary modification to Method 8015B requires use of a slow GC heating rate, which provides a more detailed chromatographic "fingerprint" of the concentration and character of the total extractable hydrocarbons in sediments. These data can be useful to describe the general hydrocarbon characteristics of the often complex mixture of hydrocarbons present in sediment samples.

Figure 2 is an example of the results of using USEPA Method 8270, both modified and unmodified, to characterize PAHs in the same petroleum source (fuel oil #6). The bottom histogram is an analysis for only the 16 Priority Pollutant PAHs, whereas the top histogram illustrates an analysis for a more inclusive suite of PAHs listed on Table 2 using modified Method 8270. The additional "fingerprinting" information obtained in the full suite of PAH analytes is of significant benefit in PAH source studies.

Using ACF to Distinguish PAH Sources

Because of their nature of formation and similar physical/ chemical properties, groups of petrogenic or pyrogenic PAHs tend to co-occur in sediments. This knowledge allows the investigator to recognize specific PAH assemblages, or "fingerprints," as being derived from a certain source. In order to recognize the distinguishing features of PAH distributions in environmental samples and potential source materials, it is necessary to measure a range of PAH compounds that include both unsubstituted (parent) PAH compounds, as well as their alkylated (substituted) homologues. Generally, a target compound list of approximately 45 PAH and related hetrocyclic compounds are necessary for chemical fingerprinting of PAHs (Stout et al. 2002 and references therein; Boehm and Loreti 2001; Saner and Boehm 1995; Wang et al. 1999; Bence et al. 1996).

Table 2. PAHs used in advanced chemical fingerprinting to identify and distinguish among PAH sources.

Figure 2. Comparison of PAH histograms for the same Fuel Oil #6 using two different PAH analyte lists. See Table 2 for PAH analyte identification.

Pyrogenic and petrogenic PAHs can be readily distinguished on the basis of their alkyl group distributions (LaFlamme and Hites, 1978; Wakeham et al. 1980). Figure 3 shows the basic relative distributions of variously alkylated PAHs, depending on the temperature (and rate) of formation. Petrogenic PAH profiles form a "bell-shaped curve" due to the relative abundance of alkylated PAHs (Cl, C2, C3, C4) (Blumer 1976). Pyrogenic PAH profiles form a decreasing, or "sloped" curve, due to the domination of the nonalkylated, parent PAHs (CO), over the alkylated PAHs; this dominance increases with increasing temperature of formation. These characteristic curves are not exhibited in PAH histograms if only the 16 Priority Pollutant PAHs are measured. More detailed descriptions of PAH sources "fingerprinting" features are available in Stout et al. (2002).

Figure 3. Representative distribution of alkylated PAHs formed at different temperatures (after Blumer 1976). Sum the multiple CO isomers when drawing the curves for 3-ring and 4-ring PAHs.

Typical Petrogenic PAH Source Signatures

Petrogenic PAH source signatures are comprised of primarily lower molecular weight (2- and 3-ring) PAHs; higher molecular weight PAHs usually are present only at low concentrations (<100 /xg/kg; (Kerr et al. 1999). Petrogenic PAHs also are characterized as having higher concentrations of alkyl groups than pyrogenic PAHs, and exhibit "bell curve" profiles of PAH histograms (Figure 3). These PAHs commonly enter urban and coastal waterways from anthropogenic sources such as spills or leaks of crude oil and other fuels, but also are comprised of effluents from oil terminals and refineries, discharges of ballast and bilge water from ships, coalfired power plants, as a component within urban runoff and from municipal sewage treatment plants.

As crude oil is refined, it is subjected to heating under mild conditions (<550C) and separates into various distillates, including light distillates (gasolines, kerosene, and jet fuel), middle distillates (diesel fuel #2, fuel oil #2 and #4), and residuals (lube oils and heavy fuel oil). Because distillation heating is relatively mild, there is no significant formation of new (more highly condensed) PAHs. Thus, the resulting refined products only contain PAHs that were present in the parent crude oil.

Different crude oils, and the petroleum products refined from them, exhibit different PAH distributions, which can be useful in differentiating among individual petrogenic sources. Figure 4 shows the PAH profiles determined for two different crude oils (one fresh and one weathered) and for diesel fuel #2 (weathered). Some obvious differences in the distributions of PAHs are evident. For example, the unweathered crude oil is enriched in lower molecular weight, 2- ring PAHs (CO-C4 naphthalenes), which are depleted in the weathered crude oil. However, overall, the CO-C4 alkyl series in each oil exhibits a bell-shaped profile that is characteristic of petrogenic (i.e., petroleum-derived) PAHs. This profile is predictably altered in the weathered crude oil due to theeffects of evaporation, biodgradation, and solubilization following a release into the environment.

Figure 4. Histograms of selected petrogenic PAH sources. see Table 2 for PAH analyte identification.

It is noteworthy that crude oil and distillate petroleum contain only very low levels of high molecular weight 5- and 6-ring PAHs relative to the lower molecular PAH compounds (Stout et al. 2002; Kaplan et al. 1996). PAHs found in pyrogenic hydrocarbon products (e.g., coal and tar derivatives) and combustion residues found in urban runoff dominate the PAH assemblages in these materials (Emsbo- Mattingly et al. 2002). Thus, the occurrence and distribution of higher molecular weight PAHs are an important means to differentiate petroleum from combustion-derived PAHs.

It is important to note that PAHs comprise only a small fraction of most crude oils and petroleum products (a typical crude oil may contain from 0.2% to approximately 5% total PAHs). For example, total PAH concentrations (i.e., the sum of the PAH analytes listed in Table 2) in the petroleum products shown in Figure 4 range from only 1.3 to 2.4 weight percent of total petroleum.

Typical Pyrogenic PAH Source Signatures

Pyrogenic PAH source signatures are complex, and, unlike the signatures in petroleum, are dominated by higher molecular weight (4- , 5-, and 6-ring) PAHs. Pyrogenic PAH assemblages are characterized by a dominance of the unalkylated (parent) PAHs, and a decreasing abundance of PAHs with increasing degree of alkylation, thereby exhibiting a sloped profile on PAH histograms (Figure 3). These PAHs may be released to the environment in vapor phases, as airborne particles or in the solid byproducts of the heating process. Fossil fuel combustion, particularly the combustion of petroleum in gasoline and diesel engines, is an important and prevalent source of vapor and paniculate pyrogenic PAHs.

Pyrogenic PAHs are produced by the incomplete combustion (O2 is present) or pyrolysis (O2 is absent) of organic matter. A commonly encountered anthropogenic source of pyrogenic PAHs in sediments are the byproducts of the carbonization (coking) processes associated with historic manufactured gas production (MGP). These processes yielded coal- and petroleum-derived liquid tar residues (coal tar and petroleum tar) that were produced in the course of heating coal or crude oil for gas production (Gas Research Institute 1987).

Figure 5 shows typical PAH profiles for three pyrogenic materials: a typical unweathered coal tar, creosote, and coal tar pitch. As expected, each of these pyrogenic PAH source materials are enriched in higher molecular weight PAHs and include several 5- and 6-ring PAHs. Pyrogenic materials contain higher concentrations of PAHs than do petrogenic materials (i.e., petroleum products). The coal tar, creosote, and coal tar pitch shown in Figure 5 contain 103,000,142,000, and 141,000 g/kg of total PAHs (i.e., 10.3-14.2 weight percent), all of which are much higher than in most petrogenic source materials, as shown in Figure 4. This indicates that even small quantities of pyrogenic materials entering an urban waterway contribute significant quantities of PAHs to sediments.

Figure 5. Histograms of selected pyrogenic sources. See Table 2 for PAH analyte identification.

Typical Urban Background PAH Signatures

General Characteristics

Stormwater runoff is probably the largest chronic contributor of background PAHs to urban sediments. During storm events, rainfall washes countless small nonpoint PAH (and other pollutant) sources from the entire catchment basin and discharges this pollution at point sources (i.e., Stormwater outfalls) along urban waterways. These sources have impacted waterways for decades, and are often significant sources of the background PAHs detected in urban sediments.

Discharged Stormwater often contains the following types of PAHs: (1) urban dust/soot particles containing combustion-related (i.e., pyrogenic) PAHs, principally derived from incomplete combustion within automobile and truck engines, especially diesel-based engines; (2) used lubricating oils (i.e., petrogenic PAHs), principally from oil drippings from automobiles and trucks onto roadways and parking lots; and (3) waste oil and petroleum products (i.e., petrogenic PAHs) that are illegally or unintentionally discharged into a city's storm drain systems. Many of these particles settle out of the water column and enter the sediment column. Because of the variable effects of dilution and transport processes, total PAH concentrations in urban sediments near Stormwater outfalls cover very broad ranges and are highly site- dependent; however, they are typically in the 1-50 Mg/g (dry) range (Stout et al. 2003).

Although the character of PAHs within urban runoff varies between different catchment areas and at different times, the overall PAH signatures of urban runoff and the receiving urban sediments are typically dominated by pyrogenic PAH assemblages (see Figure 6). Two- and 3-ring PAHs (i.e., those PAHs more likely associated with petrogenic sources) are more water soluble and degradable than higher ring PAHs. Thus, although there is some degree of mixing of petrogenic and pyrogenic source materials in urban runoff (e.g., oils and combustion particles, respectively), urban runoff is predominantly a source of pyrogenic PAHs.

Characteristics of Urban Background PAHs

Recent work by Stout et al. (2003) reports PAHs in urban background sediments from 11 waterways around the United States. Representative profiles for PAHs in these urban sediments from that study are shown in Figure 6. The lower graph in this figure shows the median concentrations (and 25th and 75th percentiles) for all 334 sediments included in the Stout et al. (2003) study. The similarity in the PAH distribution between the median values and the selected sediments shown (Figure 6, A-F) further demonstrates the overall similarity in the distribution of PAHs in sediments impacted by urban background.

Figure 6. PAH histograms for sediments impacted with urban background from (A) Portland Harbor, (B) Elizabeth River, (C) Alameda Point, (D) Eagle Harbor, (E) Thea Foss, (F) Boston Harbor, and (G) median of all sediments studied. Error bars demonstrate 25th and 75th percentiles. see Table 2 for PAH analyte identification.

As illustrated in Figure 6, the most abundant PAHs in urban background sediment are high molecular weight (4- to 6-ring) compounds, particularly the fluoranthene and pyrene isomers. The fluoranthene and chrysene homologue series each exhibit the sloped pattern characteristic of pyrogenic sources (see Figure 3). Very few lower molecular weight (2- and 3-ring) PAHs are present in the PAH fingerprints; the most abundant of these are the anthracene and phenanthrene isomers. These homologues series also exhibit the characteristic sloped pattern. These features generally are distinct from those of other pyrogenic PAH sources (e.g., Figure 5), thereby allowing for the recognition of urban background.

The work compiled by Stout et al. (2003) documents the expected range of total PAH concentrations in urban background sediment. They report that the concentrations of total Priority Pollutant PAHs and total PAHs43 for the 334 sediments evaluated in that study. The vast majority of the urban background-impacted sediments studied contained less than 20 g/kg of Priority Pollutant PAHs and less than 30 g/kg of total PAHs43. Finally, when Stout et al. (2003) evaluated the chemical nature of the distribution of the PAHs43 analytes measured in the 334 regional urban background sediments in their study, they found a consistent predominance of pyrogenic PAHs (72.4% 6.8% of the samples were pyrogenic in character). This indicates that although the PAHs present in urban background may include a petroleum-derived fraction, it is considerably smaller than the PAHs attributable to pyrogenic background sources.

Chromatographic Characteristics of Urban Background

The chromatographic character of total extractable hydrocarbons (THC) from most urban sediments exhibit similar features, which indicates the overall similarity in the nature of urban background in different urban settings.

Stout et al. (2003) presented GC/FID traces of hydrocarbon extracts of sediment from four different urban areas that illustrate the nature of urban hydrocarbon "fingerprints" (Figure 7). Each of these sediments exhibit two characteristic chromatographic features: (1) numerous later-eluting resolved peaks and (2) an unresolved complex mixture (UCM) "hump" mostly within the higher boiling hydrocarbon range. These features have been previously observed in sediments impacted with urban background from many areas (Wade and Quinn 1979; Barrick et al. 1980; Eganhouse et al. 1982; Hostettler et al. 1999). The resolved peaks in these chromatograms represent various non-alkylated 3- to 6-ring PAHs, which are indicative of the combustion-derived particles in engine exhaust (LaFlamme and Hites 1978; Westerholm et al. 1988; Oahn et al. 1999). The UCM "hump" is characteristic of a (mostly) residual range petroleum, such as lubricating, hydraulic, and waste oil(s), which also expected to occur in urban runoff (Gogou et al. 2000). Also present in some urban sediments (Figure 7) are numerous odd-dominated normal hydrocarbons (n-C^sub 27^, n-C^sub 29^, U-C^sub 31^) associated with plant waxes derived from modern leaf debris in the sediments (Prahl and Carpenter 1984).

Together, these chromatographic features exhibit the ubiquitous presence of both naturally occurring and chronic anthropogenic hydrocarbons in modern urban sediments. Among the PAHs, urban sediments are dominantly pyrogenic in nature, that is, they arise overwhelmingly from combustion sources.

Figure 7. GC/FID chromatograms for extractable hydrocarbons for sediments impacted by urban background in (A) Thea Foss Waterway, WA (B) Eagle Harbor, WA (C) Portla\nd Harbor, OR, and (D) Elizabeth River, VA sites. Laboratory internal standards (IS) are not a component of the native material.

Data Exploration and Classification Techniques

The principle goal of an advanced chemical fingerprinting investigation of PAH contaminated sediments is the determination of the nature (i.e., type) and sources of the PAHs. This analysis is generally accomplished using a number of data exploration techniques- from simple graphic methods to complex multivariate data analysis. Collectively, these numerical analysis techniques are referred to as chemometrics.

Chemometric analyses have proven to be an especially effective means of comparing the chemical data from a large number of samples. In this investigation, we utilize Principal Components Analysis (PCA) as an important data exploration and classification tool. PCA is a powerful multivariate technique for visualizing inter-sample and inter-variable relationships. It achieves this by reducing the "n" dimensionality of the data (where n = number of variables or samples, whichever is smaller) by finding linear combinations of the variables in the data set that account for the maximum amounts of variance. These linear combinations are called the principal components. The 1st principal component (PC) accounts for the maximum amount of variance and each successive PC accounts for less of the remaining variance.

An Example-A PCA of Source Products Relevant to the Delaware River Sediment Investigation

In this example, we utilize PCA to evaluate and classify the distributions of PAH compounds found in 11 hydrocarbon product samples that include petroleumderived products (e.g., crude oil, diesel fuels, lube oil, bunker fuel, and oil-derived asphalt) and pyrogenic combustion wastes (e.g., coal tars, coal tar pitch, and soot). Figure 8 shows the factor score plot of the 1st principal component (Factor 1) versus the 2nd PC (Factor 2) for the PCA analysis of the PAH compounds measured in these samples. These first two principal components accounted for 52% and 22% of the variance in the data set. The Euclidean distances between the hydrocarbon product sample points in Figure 8 are a reflection of the chemical similarity between the samples. (Samples that are plotted close are chemical similar and vice versa.) For example, the "weathered diesel" and "bunker C" samples plotted close to one another and therefore, their PAH distributions are similar relative to compositional diversity expressed by the other samples. The fuel oil #2 appears most like petroleum products diesel fuel #2, lube oil, crude oil (fresh and weathered), and bunker C. In this group of samples asphalt possesses the unique petrogenic composition. Nevertheless, the petroleum products differ from the pyrogenic products. The wider spacing of pyrogenic materials suggests greater compositional complexity among the pyrogenic samples. The soot is somewhat intermediate between the coal-derived pyrogenic materials and the petroleum products. It is clear that the PAH distributions for the petroleum products (petrogenic) are distinct from the PAH distributions for the coal-derived products (pyrogenic).

Figure 8. Factor score plot for the 1st and 2nd principal components in the PCA example involving 36 PAHs in 11 product samples.

Figure 9. Factor loading plot of the 1st and 2nd principal components shown in Figure 8. Abbreviations refer to individual or groups of PAHs.

The chemical differences responsible for the separations shown in Figure 8 can be determined by an investigation of the corresponding factor loading plot from this example. Figure 9 shows the factor loadings plot for the 1st and 2nd PC. This plot reveals the individual PAH chemicals that cause the separation or similarity grouping among the samples. This plot shows that the variables causing separation in PC1 in the positive direction are parent (unsubstitued) 4- to 6-ring PAH such as benzo(e)pyrene (BAP), benzo(g,h,i)perylene (GHI), indeno(1,2,3,c,d) pyrene (IND), and chrysene (CO). Variables causing separation in PCl in the negative direction are alkylated PAHs such as C2-fluorenes (F2), C4- naphthalenes (N4), and Cl-phenanthrenes/anthracenes (PA1). Thus, creosote's position in Figure 8 is due to higher levels of naphthalene (N0), fluorene (F0), and phenanthrene (P0) and lower levels of C4- and C3- beno (a) chrysenes (BC4, BC3), C3- fluoranthenes/pyrenes (PF3), and C4-phenanthrenes/anthracenes (PA4) compared to other samples. By the same method, it can be determined that the coal tar pitch is relatively enriched in parent 4- to 6- ring PAH (i.e., pyrene (P0), chrysene (C0), benzo(a)pyrene (BAP), benzofluoranthenes (BAO, BFO), and depleted in various 2- and 3- ring PAH (i.e., naphthalenes-N0, N1, N2, N3, and N4) and fluorenes (F0, F1, F2, F3, and F4). Overall, the samples separate graphically based on their chemical compositions and/or their genesis (i.e., combustion or refining).

Given the fact that PAH distributions in environmental samples and in suspected sources of contamination are intrinsically complex mixtures, chemometric techniques such as principal component analysis are a logical-and perhaps most clear way-of understanding the nature and sources of PAHs in the environment.

METHODS USED FOR PAH FINGERPRINTING

This section summarizes the analytical methods used to measure diagnostic PAH residues (and in select cases, GC/FID fingerprints) in environmental samples collected in support of the Motiva Refinery investigation, notably sediments, biological tissues, water, and petroleum products. The results of the analyses of these various samples are discussed in this and other articles presented in this special issue (Hall et al. 2005; Salazar et al. 2005; Alexander et al. 2005; Alden et al. 2005).

The analytical methods described in this article are well- documented performance-based methods, optimized for measurement of an extended list of PAHs, or for high resolution GC/FID chromatographic fingerprints of environmental samples. These methods have been previously described in the literature (Sauer and Uhler 1994), and have been used for more than a decade by researchers and investigators focused on investigations of hydrocarbon contamination and the occurrence and fate of PAHs in the environment. These methods are fully described in Hall et al. (2004). All raw data pertaining to sample receipt, processing, and analysis are maintained in controlled laboratory record books. All laboratory work conducted at Battelle is monitored by an independent Quality Assurance Unit, who performs routine systems audits and data audits to ensure the tractability and accuracy of the data produced by the laboratory.

Sample Receipt and Storage

Field samples were collected by University of Maryland scientists or Refinery environmental contractors, and shipped under appropriate storage and chain-of-custody procedures to Battelle's Duxbury, Massachusetts environmental chemistry laboratory. Upon arrival at the laboratory, the samples were received and taken under custody of a trained Sample Custodian. The integrity of the samples was checked and documented, and the samples were transferred to the appropriate access-controlled storage refrigerator or freezer until removed for analysis. Procedures followed for sample receipt and log-in, storage, refrigerator/freezer temperature monitoring, and internal chain-of-custody procedures are fully described in Battelle Standard Operating Procedures.

Sample Extraction and Extract Clean-Up

Sediments

Sediment samples were prepared for analysis using solvent extraction and cleanup techniques followed by the NOAA National Status and Trends Program (NOAA 1998). Approximately 30 g (wet weight) of homogenized sediment was used for solvent extraction and an additional 5-10 g (wet weight) were collected for dry-weight determination. The surrogate internal standards (SIS) o-terphenyl, naphthalene-d8, phenanthrene-d^sub 10^, and chrysene-d^sub 12 ^ were added to each sample to document extraction efficiency. Sodium sulfate was added to absorb water from the sample and facilitate extraction with organic solvent. Additionally, activated copper was added to sediment samples to remove any sulfur that might have been present in the sample. The sediment homogenate was shaken/tumbled once for a minimum of 12 h, and then twice for at least 1 h, using dichloromethane (DCM) as the extraction solvent. The sample was centrifuged between the extractions, and the solvent decanted into a pre-cleaned, labeled Erlenmeyer flask. The combined extract was filtered and dried through a glass fiber filter containing sodium sulfate. The extract was then concentrated to 1 mL using Kuderna- Danish and nitrogen evaporation (N-Evap) techniques.

Extracts were weighed and processed through a 20 g alumina column in order to obtain a combined aliphatic and aromatic/unsaturated fraction (F1+F2). The combined F1+F2 were eluted with DCM. The combined F1+F2 fractions were then concentrated to 1 mL using the Kuderna-Danish and N-Evap techniques described earlier. The concentrated F1 + F2 fraction was further purified using gel permeation, high performance liquid chromatography (GP/HPLC). The purified F1 + F2 fraction was spiked with appropriate concentrations of recovery internal standard (RIS, containing 5a-androstane and fluorene-d^sub 10^) in preparation for PAH and optional total petroleum hydrocarbon (TPH) analysis.

Water from monthly samples

Samples of water were extracted for low-level PAH analysis using solvent extraction techniques consistent with EPA Method 3510, Separatory Funnel LiquidLiquid Extraction. Typically, 2 L of water were added to a 3-L separatory funnel and fortified with the surrogate internal standards were o-terphenyl, naphthalene-d^sub 8^, phenanthrene-d^sub 10^, and chrysene-d^sub 12^. The water was then serially extracted 3 times with 120-mL aliquots of dichloromethane. The DCM extracts were \combined and dried over sodium sulfate, concentrated to ~1.0 mL using Kuderna-Danish and nitrogen evaporation (N-Evap) techniques, and then spiked with appropriate recovery internal standard (RJS, containing 5a-androstane and fluorene-dH)) in preparation for PAH and optional total petroleum hydrocarbon analysis.

Petroleum products

Subsamples of petroleum products obtained from the refinery were quantitatively transferred into a 10 mL volumetric flask, spiked with the surrogate internal standards o-terphenyl, naphthalene-d8, phenanthrene-dH), and chrysene-di^ and the recovery internal standards 5a-androstane and fluorene-d]0 and diluted to a concentration of approximately 5 g/mL in preparation for PAHs and optional total petroleum hydrocarbon analysis.

Tissue (bivalves)

Tissue samples were prepared for analysis using solvent extraction and clean-up techniques followed by the NOAA National Status and Trends Program (NOAA 1998). Bivalve samples were shucked into a pre-cleaned container and thoroughly homogenized using a Tekmar Tissuemizer. Approximately 30 g (wet weight) of homogenized tissue was used for solvent extraction and an additional 5-10 g (wet weight) were collected for dry weight determination. The surrogate internal standards (SIS) o-terphenyl, naphthalenc-d^sub 8^, phenanthrene-d^sub 10^, and chryscne-d^sub 12^ were added to each sample to document extraction efficiency.

Sodium sulfate was added to absorb water from the sample and facilitate extraction with organic solvent. The tissue homogenate was macerate/extracted 3 times, each for 2 min using a tissuemizer, using dichloromethane (DCM) as the extraction solvent. The sample was centrifuged between the extractions, and the solvent decanted into a pre-cleaned, labeled Erlenmeyer flask. The combined extract was filtered and dried through a glass fiber filter containing sodium sulfate. The extract was then concentrated to 1 mL using Kuderna-Danish and nitrogen evaporation (N-Evap) techniques.

Extracts were processed through a 40 g alumina column in order to obtain a combined aliphatic and aromatic/unsaturated fraction (F1+F2). The combined F1+ F2 were eluted with DCM. The combined F1 + F2 fractions were then concentrated to 1 mL using the Kuderna- Danish and N-Evap techniques described earlier. The concentrated F1 + F2 fraction was further purified using gel permeation, highperformance liquid chromatography (GP/HPLC). The purified F1 + F2 fraction was spiked with appropriate concentrations of recovery internal standard (RIS, containing Sa-androstane and fluorene-d10) in preparation for PAH analysis.

Quality Control

A consistent quality control (QC) program was followed throughout the course of the investigation. Each batch of no more than 20 samples was accompanied by a suite of QC samples, analyzed in order to document the accuracy and precision of the measurement program. The following quality control samples were processed along with each batch of samples: 1 procedural blank (PB); 1 laboratory control sample (LCS); 1 matrix spike/spike duplicate (MS/MSD); and 1 field sample duplicate (DUP)

Data quality objectives (DQOs)-based on historical laboratory performance records-were followed (Table 3). If necessary, appropriate data qualifiers were applied to data points requiring data quality excursion explanations (Table 4).

Instrumental Analysis of Total Extractable Hydrocarbon (THE)-GC/ FID "Fingerprints"

Analysis of the F1+ F2 fractions of sediment extracts by gas chromatography with flame ionization detection (GC/FID) for TPH was performed following a modification of USEPA Method 8015, Nonhalogenated Organics Using GC/FID. Splitless analyses were carried out using a 30-m 0.25 mm i.d. fused silica capillary column coated with a 0.25-um 100% methylsilicone crosslinked stationary phase. The gas chromatograph was operated with an initial temperature of 35C, and an oven ramp rates of 6C/min to a final temperature of 320C.

Prior to sample analysis, a five-point response factor calibration was performed to demonstrate the linear range of the analysis and to determine the individual response factors at each calibration solution concentration. The calibration solution was composed of selected C^sub 8^ to C^sub 40^ n-alkanes, pristane and phytane. Analyte concentrations in the standard solutions may range from 1.0 g/mL to 100 g/mL. A mid-level calibration check and procedural blank was performed for every 10 samples. The individual response factors at each calibration concentration were determined and the TPH response factor was based on the average response factors of all the target analytes in the calibration solution.

Instrumental Analysis of PAHs

The analysis of the target parent and alkyl homologue PAH compounds was performed using gas chromatography/mass spectrometry (GC/MS) techniques that are modifications of USEPA Method 8270, Semivolatile Organic Compounds by Gas Chromatography/Mass Spectrometry (GC/MS). Splitless analyses were carried out using a 60 m, 0.25 mm Ld., 0.25 um film thickness 5% phenyl-95% methyl- silicone capillary column. The gas chromatograph was operated from an initial temperature of 40C, following an oven ramp rates of 6C/ min to a final temperature of 300C. The mass spectrometer was operated in the selected ion monitoring (SIM) mode. Diagnostic and confirmatory ions for each target compound were monitored for each parent and alkyl homologue. Alkyl homologues of PAHs were quantified using the straight baseline integration method, versus response factors assigned from the parent PAH compound.

Table 3. Data quality objectives for sediment, tissue and water analysis.

Table 4. Data qualifiers for sediment, tissue, and water analysis.

A minimum of a 5-point response factor calibration was run with analyte concentrations in the standard solutions ranging from approximately 0.02 ng/L to approximately 10 ng/L. The samples were bracketed by passing standard checks analyzed no less frequently than every 10 samples and at the completion of the sequence.

Quantification of individual compounds was obtained by the method of internal standards using the RIS compounds as quantification internal standards. Total PAH was determined as the sum of the individual PAH analytes. The homologous series of alkylated PAHs were quantified using the response factor of the parent PAHs. Target analytes are listed in Table 2. Method detection limits applicable for this investigation for sediment, tissue, and water (determined following EPA guidelines for determining the MDL (USEPA 1996)) are compiled in Table 5.

RESULTS AND DISCUSSION

The results of the advanced chemical fingerprinting investigation provide a basis to examine the nature and source signatures of PAHs in sediments proximal to, and distant from, the Motiva Refinery. As part of the analysis, the following were examined: (1) concentrations of diagnostic PAHs in the Refinery wastewater and sediments from the canal leading to the Delaware River (e.g., the potential Refinery PAH "source" signatures); (2) concentrations of diagnostic PAHs in river sdiments proximal to and distant from the Refinery (e.g., from near-field and far-field sampling stations); (3) compositional relationship of these diagnostic PAH compounds among candidate "sources" and those found in the Delaware River sediments, and (4) approximate spatial extent of Refinery-derived PAHs in the river sediments.

Table 5. Method detection limits (MDL) and reporting limits for target parent PAH and alkyl homologues.

Table 5. Method detection limits (MDL) and reporting limits for target parent PAH and alkyl homologues.

The results in this article follow the primary analytical tasks that governed the overall project. These tasks included the (1) monthly water monitoring; (2) sediment survey study; (3) sediment Triad study; (4) sediment core study; and (5) bivalve study.

Monthly Water Monitoring

The water monitoring program established a four-year record of hydrocarbon profiles along the potential migration pathway of fugitive petroleum residues from the Refinery (Hall et al. 2004). The historical extent of petroleum releases from the Refinery was determined based on the PAH composition of these samples. The field team collected samples from approximately eight stations on a monthly basis from 1999 to 2002 (Figure 10). Samples from two Refinery outfalls (601 and 001) and two remote Delaware River stations (06 down river and 10 up river) provided a comparative framework for tracking the extent of PAHs from the Motiva facility effluent canal (station 01) and into the near-field (stations 02 and 07). Approximately 45 samples were collected at the Cooling Water Intake Canal (CWIN) (duplicate), 601 (duplicate), 001 (duplicate), 02, 07, and 10. Fewer samples were collected at 01 (n ~35) and 06 (n ~15). The detailed PAH concentrations are presented in Hall et al. (2004). The station-specific average and maximum PAH composition was calculated for the following discussion of general PAH assemblages and source attribution.

Figure 10. Locations of monthly water monitoring samples collected from 1999 to 2002.

Outfalls 601 and 001 were sampled approximately 30 times each between 1999 and 2002. The repetitive and systematic sampling of outfalls 601 and 001 provided a basis to determine the average and maximum PAH composition in these effluent streams over time. The average sample of effluent from the wastewater treatment plant (WWTP) at the Refinery (outfall 601) was dominated by petrogenic 4- ring PAHs (Figure Ha). Specifically, the Refinery signature exhibited relatively high levels of fluoranthenes/pyrenes with two (FP2) and three (FP3) alkyl groups and benz(a)anthracene/chrysenes with two (BC2), three (BC3), and four (BC4) alkyl groups. On average, the WWTP process reduced the 2- and 3-ring PAHs to highly alkylated residuals. Several compounds used to differentiate Delaware River background (fluoranthene, benz(\a) anthracene and perylene) were not present or present at low levels in the WWTP effluent. The sample containing the highest level of EPAPAH among the outfall 601 monthly monitoring samples matched the average sample closely and indicated that the average signature was representative of the Refinery releases of greatest environmental concern (Figure 11b). In summary, the average (13.6 ug/L) and maximum (320 ug/L) samples from outfall 601 indicated that the majority of petroleum mass released from the Refinery consisted of petrogenic 4-ring PAHs.

Figure 11. PAH histograms of selected monthly water monitoring samples (GC/MS/SIM). See Table 2 for PAH analyte identification, (a) WWTP Effluent (Average): Outfall 601 contains the effluent from the Refinery wastewater treatment plant (WWTP). Its average composition was enriched in 4-ring petrogenic PAHs with little to no FLO, BAO, and PER. The 2- to 3-ring PAHs were severely weathered by the WWTP processing, (b) WWTP Effluent (Maximum): The sample collected in March 1999 contained the highest level of EPAPAH. Compositionally, it resembled the average sample closely with less 2-ring PAHs. (c) Diluted Effluent (Average): Outfall 001 contains less than 5% by volume 601 effluent plus more than 95% non-contact cooling water from the Delaware River. In addition to the WWTP signature above, the average 001 sample contained petrogenic 2- and 3-ring PAHs of unknown origin; pyrogenic 5and 6-ring PAHs; and diagenetic perylene.

Figure 11. (d) Effluent Channel (Average): Station 01 is in the tidal effluent channel. Delaware River background dominated the average profile in the form of 2- and 3-ring PAHs, pyrogenic 4- to 6- ring PAHs plus perylene. (e) Nearfield (Average): Station 02 is immediately outside the mouth of the effluent channel. Naphthalenes dominated the PAH profile. These compounds may constitute a weathered automotive or marine fuel. Magnified insert illustrates patterns dwarfed by enriched 2-ring PAHs. (f) Nearfield (Average): Station 07 is up river of 02. Naphthalenes dominated the PAH profile. Petrogenic 3-ring PAHs, pyrogenic 4- to 6-ring PAHs and diagenetic perylene were consistent with Delaware River background. Magnified insert illustrates patterns dwarfed by enriched 2-ring PAHs.

Outfall sample 001 partially represented the Refinery signature (Figure lie). However, it contained less than 5 percent of the 601 effluent commingled with more than 95 percent non-contact cooling water from the Delaware River. This compositional estimate reflected the approximate mixing volumes of water from the wastewater treatment system (e.g., 13 mgd at outfall 601) with non-contact cooling water from the Refinery (e.g., 490 mgd at outfall 001) (DNREC 1997). Consequently, sample 001 only possessed a small amount of the hydrocarbon signature from the Refinery. The average signature of the diluted effluent clearly retained the fingerprint of outfall 601 as evidenced in the high levels of petrogenic 4-ring PAHs. However, the diluted effluent also contained 2- to 3-ring PAHs plus higher levels of perylene and 5- to 6-ring PAHs. As discussed below, these compounds were consistent with the urban background PAHs collected from the up and down river locations.

Figure 11. (g) Clean Water Intake (Average): Delaware River water is drawn into the Refinery plant at the CWIN station. The various petrogenic . pyrogenic, and diagenetic PAH were consistent with Delaware River background. The levels of FLO, BAO, and PER are relatively high in the background samples and help identify the influence of Delaware River water, (h) Down River (Average): Station 06 is a Delaware River background reference point located approximately 10 miles south of the Refinery. Naphthalenes (2-ring PAHs) dominated this pattern with lesser amounts of petrogenic 3- ring PAHs, pyrogenic 4- to 6-ring PAHs, and diagenetic perylene. Magnified insert illustrates patterns dwarfed by enriched 2-ring PAHs. (i) Up River (Average): Station 10 is a Delaware River background reference point located approximately 5 miles up river from the Refinery. Naphthalenes dominated the mixed (petrogenic and pyrogenic) 3-ring PAHs, pyrogenic 4- to 6 ring PAHs, and diagenetic perylene.

Figure 11. (j) Up River (Maximum): The sample collected in September 1999 contained the highest level ofE PAPAH at station 10. Compositionally, it resembled the average sample closely with more 2- ring PAHs, less petrogenic 3-ring PAHs, and less perylene. Magnified insert illustrates patterns dwarfed by enriched 2-ring PAHs.

Once released to the effluent channel, the Delaware River diluted the WWTP effluent more intensively. The average sample from station Ol demonstrated the dominant influence of background PAH in the effluent channel (Figure lid). The 4-ring PAHs from the Motiva WWTP lost most of its petrogenic character in favor of the 2- to 6-ring PAHs from the Delaware River; that is, the relative abundance of refinery PAH declined rapidly below the background levels in the river sediment. Once in the nearfield (stations 02 and 07), the petrogenic 4-ring PAH signature of the Refinery largely disappeared (Figures lie and Uf, respectively). At these locations, the 4-ring PAH pattern was pyrogenic with the concentration of parent PAHs exceeding the di- and tri-alkyl homologues. The higher levels of 2- and 3ring PAHs, pyrogenic 5- and 6-ring PAHs, and diagenetic perylene confirmed the presence of water from the Delaware River.

Less than 1 mile down river, the average PAH composition at the cooling water intake (CWIN) canal (Figure Hg) matched well the average up river background station 10 (Figure Hi). Both samples contained 2- to 3-ring PAHs and pyrogenic 4- to 6-ring PAHs. The levels of pyrogenic (fluoranthene and benz (a) anthracene) and diagenetic (perylene) PAHs were high relative to the WWTP sample. Although the exact origin of the naphthalenes was unknown, possible sources included weathered automotive fuel, marine fuel, or bilge releases from large tankers. These naphthalenes were the dominant PAH constituents at water monitoring stations dominated by background, especially at the average down river station (Figure 11h) and the maximum up river station (Figure 11j). In summary, the Delaware River water background was compositionally distinct from the petrogenic 4-ring PAHs measured consistently in the Motiva WWTP effluent. The Delaware River background samples contained relatively higher levels of 2- to 3-ring PAHs, pyrogenic 4- to 6-ring PAHs, and perylene.

Principal Component Analysis of the water monitoring data was used to compare the PAH assemblages found among the measured samples. As such, the PAH data generated in this investigation were used to create a multivariate PCA model using Pirouette, Version 3.11 (Infometrix, Seattle, Washington). This model captured the compositional signatures of PAHs in water samples representative of the Refinery effluent and several types of background sources.

Forty-three PAH analytes measured in approximately 320 water monitoring samples served as the basis for the PCA water model. In addition, the model included the average and maximum PAH composition of water samples collected at outfalls 601 and 001. These samples represented the effluent from the Refinery wastewater treatment plant (outfall 601) and a dilution of this effluent with non- contact Refinery cooling water from the Delaware River (outfall 001). Two steps were required to minimize concentration effects. First, the PAH concentrations were normalized to the maximum PAH concentration per sample in order to improve comparability among the samples. second, the PAH concentrations were mean-centered and standardized by the variance to improve comparability among the analytes. The resulting model captured 35% and 12% of the PAH variability in principal component factors 1 and 2, respectively. Samples with a distinctive enrichment of petrogenic 4-ring PAHs plotted in the upper left quadrant, parent 4- to 6-ring PAHs toward the right quadrants, and 2- to 3-ring PAHs toward the lower left quadrant (Figure 12). These three primary PAH categories provided the basis for grouping the samples by PAH origin as described later.

In the graphs to follow, the sample IDs were truncated for legibility. Stations within a specific region were colored blue. In regions with more than one monitoring station, the samples from one station were colored blue and samples from the second station were colored green. The PCA loading plot (upper middle plot) indicated that samples plotting in the upper left quadrant contained 4-ring petrogenic PAHs indicative of the Refinery effluent. By contrast, samples plotting to the right (pyrogenic) and lower left (naphthalenes) corresponded to Delaware River background. The Refinery effluent from outfall 601 was consistently dominated by petrogenic 4-ring PAHs as noted previously (Figure 12). The 601 samples that plotted in the lower left quadrant generally contained higher relative abundances of naphthalenes with a characteristic depletion of CO to C2 homologues indicative of less extreme weathering in the WWTP. This depletion of less alkylated naphthalenes was generally absent in the background samples discussed in what follows.

The PAH composition in the effluent canal varied considerably in response to the dynamic mixing of Refinery effluent and Delaware River water. For example, samples from outfall 001 ranged widely throughout the upper left (petrogenic 4-ring PAHs), right (pyrogenic 4- to 6-ring PAHs), and lower left quadrants (2-ring PAHs). Samples from station Ol plotted over a similar range with fewer samples in the extreme upper left corner indicating greater dilution with water from the Delaware River. Indeed, the position of the Ol average and maximum samples in the lower left quadrant indicated the dominant influence of water from the Delaware River.

Figure 12. PCA of PAH in monthly monitoring \samples by geographic group.

The nearfield samples plotted still further toward the Delaware River reference samples (e.g., upriver and downriver stations). Most of the 02 and 07 samples concentrated in the lower left quadrant with the average and maximum sample points. A minority of 02 and 07 samples exhibited low levels of Motiva-derived PAHs as evidenced by the samples in the upper left quadrant. Samples in the intake (CWIN), up river, and down river generally plotted in the lower left (2-ring PAHs) and upper right (pyrogenic 4- to 6-ring PAHs) quadrants. The average and maximum sample points from the CWIN and 10 stations were nearly identical whereas the down river station was more enriched in naphthalenes. Consequently, water samples from CWIN, 06, and 10 stations were attributed to varying Delaware River urban background influences that were independent of the Refinery.

Sediment Survey Study

The project team conducted a sediment survey study from 1999 to 2000 as described in Hall et al. (2004). It provided detailed information about the character of local and regional hydrocarbons throughout approximately 12 miles of the Delaware River around the Refinery. The study focused on the identification of fugitive petroleum from the Refinery and urban background material within the Delaware River sediments. A detailed PAH analysis was conducted on each sediment sample. These data served as the basis for determining the nature and source of the PAHs found in the Delaware River in two ways. First, the distribution of PAH concentrations were evaluated throughout the study and the potential contribution of PAHs by the Refinery was determined. second, the spatial distribution of PAH patterns representative of Refinery and "Urban Background" sources were evaluated. This tiered approach defined the likely Refinery impacted PAH "footprint" in the Delaware River sediments using multiple lines of evidence.

During 1999 and 2000, the field team collected 63 sediment samples from the study area (Figure 13a, b). A listing of these samples and associated PAH results are presented in Hall et al. (2004). Inspection of the PAH results and qualitative gas chromatographic "fingerprints" of total extractable hydrocarbons revealed that the sediment samples collected during the two surveys contained hydrocarbons from both natural and anthropogenic (human) sources. Examples of naturally occurring hydrocarbons included PAH compounds like perylene and plant waxes like normal alkanes with a strong odd to even preference in the 23 to 33 carbon range (Peters and Moldowan 1993). As discussed previously, we attributed these materials to terrestrial plant debris commonly identified in sediments. These naturally occurring compounds dominated the hydrocarbon composition of many samples.

The anthropogenic hydrocarbons found in sediments consisted of hydrocarbons from petroleum (e.g., automotive lubricating oil, marine fuels, and Refinery effluent) and partially combusted organic matter commonly termed soot (e.g., engine exhaust, furnace emissions, asphalt, worn tires, and wood fires). Recognizing that these were at least two potential sources of petroleum-derived hydrocarbons important to this investigation (namely Refinery and urban background), we distinguished the anthropogenic hydrocarbons associated with urban background (e.g., automotive lubricating oil and marine fuels) from those attributable to the Refinery. Specifically, samples collected at remote locations in the Delaware River defined the urban background signature (e.g., down river station 06, up river station 09, and eastern shore station 61). As discussed previously, the PAH pattern attributed to the regional background contained varying levels of 2- and 3-ring PAHs plus pyrogenic 4- through 6-ring PAHs. By contrast, samples collected from the wastewater treatment plant effluent defined the Refinery signature (e.g., outfalls 601 and 001). As discussed previously, the PAH pattern attributed to the Refinery exhibited enriched petrogenic 4-ring PAHs; specifically FP2, FP3, BC2, BC3, and BC4.

Figure 13a. Locations for sediment survey samples collected in 1999 and 2000 (northern panel).

Figure 13b. Locations for sediment survey samples collected in 1999 and 2000 (southern panel).

Spatial Distribution of Sediment Samples

The sediment survey spanned a wide geographic area of over 15 river kilometers (Figure 13). The sampling stations were organized into five geographical groups based on proximity to the Refinery for ease of forensic analysis and discussion. The spatial distribution of PAHs demonstrated the utility of these groupings. The total concentration of 16 EPA Priority Pollutant PAH (EPAPAH) adequately summarized the general PAH concentration gradients in the study area. This convention for summarizing PAH data also compared well with regulatory thresholds and other background PAH studies. In general, the PAH concentrations in the study area (61 to 11,000 g/ kg dry weight) fell within background levels observed in other regions (20,000 g/kg dry weight; see previous discussion). Specific observations were as follows: (1) the highest PAH concentrations appeared in the down river sediments below the C&D canal (Figure 13, EPAPAH 5,000 to 7,500 ppb (g/kg dry weight sediment) as light blue dots and 7,500 to 11,000 as dark red dots); (2) intermediate PAH concentrations were observed at several up river and farfield stations (Figure 13, EPAPAH 2,500 to 5,000 ppb as crossed gray circles), and (3) lower PAH concentrations appeared in the effluent channel, nearfield, and background reference stations (06 down river and 09 up river) (Figure 13, non-detect to 2,500 ppb as open gray circles).

In summary, the low PAH loading from the effluent channel failed to exceed local background conditions. Consequently, the effluent channel and near-field sediments constituted an unlikely source of regional PAH enrichment.

The linear relationship between EPAPAH and a pyrogenic PAH marker for urban background (fluoranthene) indicated that the EPAPAH gradients themselves contain a measure of source identification information (Figure 14). This figure mirrored the geographic presentation of EPAPAH in F

Source: Human and Ecological Risk Assessment

More News in this Category