Jump to Navigation

Chemical Fingerprinting of Gasoline and Diesel Fuel – What Is It? Where Did It Come From? Who Is Responsible?

August 23, 2006

Introduction

The need for identification, delineation, and differentiation of petroleum-derived contaminants resulting from leaking storage tanks, pipelines, or following a release of fuel during trans-shipment of petroleum is a particularly challenging aspect of site investigations where an equitable settlement of the resulting liability and damages is at stake. Significant advances have been made over the last fifteen years with regard to detailed compositional analysis of petroleum in the environment—often referred to as “chemical fingerprinting.” Some of the earliest applications of chemical fingerprinting were related to marine oil spills, e.g., Exxon Valdez grounding, in which knowledge of crude oil or residual fuel geochemistry was applied to identify and differentiate the spilled oil in Prince William Sound and assess its environmental impacts (e.g., Bence et al., 1996). In the past few years, continued developments in the chemical fingerprinting of refined petroleum products, such as gasoline and diesel fuel (Kaplan et al., 1997; Stout et al., 2002), have provided new tools for answering questions surrounding the source and/or age of contamination resulting from fugitive releases of petroleum (Beall et al., 2002; Kaplan, 2003).

Environmental forensic investigations typically address questions aimed at identifying the nature of contamination, its sources, and the timing of its release to determine the responsible parties. Definitive answers to these questions are not always achieved through forensic investigations, but combining chemical fingerprinting with other types of forensic data, including an understanding of the site-specific geologic and hydrogeologic conditions and operational and regulatory histories for the site, can produce highly effective and defensible arguments (Stout et al., 1998).

Here, we describe some of the advances in forensic chemistry that have been developed in the last five years, and we present examples of their application. These new methods are now routinely used in the chemical fingerprinting of automotive gasolines and diesel fuels, and the two abbreviated case studies presented below demonstrate how they have been successfully applied in recent environmental forensic investigations.

Analytical Strategies

Chemical fingerprinting data must provide sufficient specificity to recognize the particular types of petroleum present at a site, characterize the effects of weathering on their chemical composition, and provide the diagnostic information necessary to distinguish and perhaps allocate between multiple sources of petroleum and assess their likely ages. For this reason, a “turnkey” analytical program that utilizes standard, regulatory methods for chemical analysis (e.g., U.S. Environmental Protection Agency (“EPA”) Methods 8015, 8020, 8260 and 8270, etc.) usually does not produce the chemical detail needed to defensibly resolve environmental forensic questions. Conventional lists of chemicals measured to demonstrate regulatory compliance do not include the dominant and important hydrocarbon compounds that make up petroleum. For example, the polycyclic aromatic hydrocarbons (PAH) and benzene, toluene, ethylbenzene, and xylenes (BTEX) compounds measured using standard US EPA methods (i.e., methods 8270 and 8260) typically make up less than five to eight percent of the total PAHs and volatile chemicals in most petroleum products, and as such the data have, at best, limited diagnostic value (Douglas and Uhler, 1993). Instead, methods of chemical analysis suitable for environmental forensics investigations measure a broader suite of compounds in gasoline and diesel fuel that are useful for source identification and differentiation.

For investigations of gasoline or middle-distillate (e.g. heating oils, diesel fuels, kerosene) releases, we recommend the use of a tiered analytical strategy that captures a full spectrum of chemical compositional information (Figure 1). Such a strategy allows for the quantitative measurement of a large number of gasoline-range (volatile) and diesel range (semi-volatile) hydrocarbons and non-hydrocarbons. In gasoline investigations this involves measurement of nearly one hundred of the so-called “PIANO” compounds (paraffins, iosparaffins, aromatics, naphthenes, olefins), oxygenates, alkyl lead additives, halogenated lead scavangers, and volatile sulfur compounds. In diesel fuel investigations this involves the measurement of n-alkanes, acyclic isoprenoids, parent and alkylated PAH, low-boiling biomarkers (e.g., sesquiterpanes), and total sulfur concentrations. Detailed descriptions of the analytical methods used to measure these compounds have been published elsewhere (Stout et al., 2002; Uhler et al., 2003; Douglas et al., 2004).

Gasoline Fingerprinting

Automotive gasolines are complex fuels blended from a variety of intermediate refinery streams, each with different physical and chemical properties (Stout et al., 2001). Historic gasolines were blended primarily to achieve physical specifications for boiling range, vapor pressure, oxidation stability, and octane with the goal being suitable engine performance, such as starting under a range of temperatures, acceleration, knock, and resistance to vapor lock. How these physical specifications were achieved was largely left up to the individual refiners. Consequently, historic gasoline blends were quite variable in chemical composition in both a temporal and spatial sense.

Modern reformulated and oxygenated gasolines must now meet stricter physical and chemical specifications. The latter include restrictions on the content of olefins, sulfur, benzene, total aromatic hydrocarbons, and oxygen. These stricter specifications have reduced the compositional variability that had existed within the gasoline pool. Even so, molecular-level chemical differences between different ‘types’ of gasolines persist depending on the refining process (Beall et al., 2002; Stout et al., 2001). This is exemplified in Figure 2, which shows the normalized PIANO distribution for two premium reformulated gasolines (RFGs) sold in the mid-Atlantic region (an ozone non-attainment area) during the winter of 1999. Both gasolines met the Federal RFG requirements, the American Society for Testing and Materials (“ASTM”) requirements, and the performance requirements, yet each exhibits distinct hydrocarbon distributions. Refiner A achieved the required octane levels primarily from the blending of methyl tertiary butyl ether (“MTBE”) and iso-octane, whereas refiner B achieved octane levels from blending MTBE and toluene. The reformulated gasolines originating from these two refiners can be distinguished on this basis.

The following case study illustrates the effectiveness of this method for fingerprinting gasoline:

Case Study 1 –Identify the Source of Off-Site Contamination

The objective of this investigation was to determine if a non-aqueous phase liquid (NAPL) encountered under a street separating two service stations was correlated to free-phase gasolines on either of the two adjacent service station properties. Detailed analysis was conducted on free-phase gasoline product samples from each property and on the NAPL from beneath the street (Figure 3). The gasolines recovered from each station revealed genetic differences related to refinery blending. Station (Refiner) B’s gasoline contained an abundance of particular iso-paraffins, namely, 2,2,4-, 2,3,4- and 2,3,3-trimethylpentane (Fig. 3), which indicate that Refiner B blended alkylate into their gasolines. Station (Refiner) A apparently did not use alkylate in production of their gasoline. The relative absence of the Station B iso-paraffins in the ‘Street’ NAPL indicated it was not consistent with the gasoline from Station B; therefore, Station A is the likely source of the NAPL.

Diesel Fuel Fingerprinting

Diesel fuel #2 used in on-road vehicles belong to the distillate family of fuels. As their name implies, the production of distillate fuels involves vaporizing and re-condensing, which distinguishes them from the higher boiling, residual fuels (e.g., fuel oil #6). With minor exceptions, diesel fuel #2 generally boils within the range of approximately 100oC to 400oC. The specific characteristic of any given diesel fuel #2 will depend upon: (1) the specific “recipe” by which it was refined and blended (e.g., hydrotreated versus straight-run), (2) the nature of the crude-oil feedstock (e.g., sweet versus sour crude), and (3) the intended market (e.g., on-road versus off-road grade diesel fuel; Stout et al., 2004). Each of these factors can introduce considerable variability in the detailed molecular composition of distillate fuels.

Due to the detrimental effects (corrosion, wear, and deposit build-up) sulfur has on engine and furnace parts, and the implications for deleterious effects on air quality, sulfur content of most distillate fuels has been long specified (Gruse, 1967). The first U.S. specification for diesel fuel #2, dating from 1922, required less than 1.5 %vol sulfur (less than 15,000 parts per million or “ppm;” Gruse, 1967). However, it was quickly learned that the higher the sulfur content, the greater were the maintenance problems encountered in diesel engines. Thus, in practice, most historic diesel fuels contained less than 5000 ppm sulfur. In 1993, due to concerns surrounding air emissions (not engine maintenance), the EPA required that on-road varieties of “low sulfur” diesel fuel contain less than 500 ppm sulfur. Prior to 1993, on-road diesel fuels #2 contained an average of 2,500 ppm sulfur (U.S. EPA, 2000), i.e., five times higher than current limit. This difference in sulfur content with time can prove useful in certain environmental forensic investigations at sites where the “age” of diesel fuel determines liability. Another case study illustrates the potential usefulness of examining sulfur content differences:

Case Study 2 – Determine Timing of NAPL Release

The objective of this study was to determine the ages of NAPL present at a truck stop. When ownership of the truck stop changed in December 1993, there was an agreement that existing contamination was the responsibility of the prior owner and any future contamination would be the responsibility of the new owner. In 1997, the thickness of the NAPL layer increased dramatically despite the prior owner’s on-going and successful efforts to recover the product. The prior owner suspected a more recent (post-sale) release had occurred.

Distinguishing between the different owner and operator fuel sources was a challenge because each operator had received diesel fuel from a variety of sources over their time of operation and there may have been long-term releases of fuel from either operator. The conventional fingerprinting data (e.g., isoprenoid ratios, PAH distributions, and low-boiling biomarkers), which normally recognize distinct types of diesel, yielded ambiguous results. Dating the time of release based upon degrees of fuel biodegradation (Christensen and Larsen, 1993) was inappropriate because the fresh-dispensed diesel fuel was erroneously estimated to be 8-years old by this method. However, when the total sulfur content was measured in the NAPLs and modern dispensers (ASTM D-4294-03) and compared to the historic trend for diesel fuel #2 sold in the northeastern U.S., as compiled from National Institute of Petroleum and Energy Research (NIPER) annual databases, the apparent NAPL ages became clear. Data presented on Figure 4 clearly demonstrated that most of the twenty-five NAPLs (M#) and all eight of the dispensed diesel fuels (D#) from the site contained less than 0.5 percent (<500 ppm) sulfur. This indicated that most of the NAPLs were consistent with low-sulfur diesel fuels produced after the 1993 regulation requiring less than 500 ppm sulfur. Those few NAPLs containing more than 500 ppm total sulfur were likely from an area of the site where pre-1993 diesel fuels were stored in underground storage tanks by the earlier operator. These results demonstrated that the increase in NAPL thickness observed in 1997 was the result of recent releases of diesel fuel, which was the responsibility of the new owner.

Summary

Chemical fingerprinting of gasoline- and diesel-fuel-derived contamination can help resolve environmental forensic questions surrounding the source and age of the contamination as a means of establishing the responsible party at a site. Chemical fingerprinting can be combined with other types of environmental forensic data (e.g., geology/hydrology, refining history, operational history, and regulatory history) to increase the defensibility of any conclusions. At the heart of chemical fingerprinting is the ability to tailor or modify analytical methods to provide sufficient chemical detail to identify and distinguish gasoline and diesel-fuel derived from different sources at petroleum impacted sites.

References

  • Beall, P.W., S.A. Stout, G.S. Douglas, A.D. Uhler. 2002. On the role of process forensics in the characterization of fugitive gasoline. Env. Claims J. 14(4):487-506.
  • Bence, A.E., K.A. Kvenvolden, and M.C. Kennicutt II. 1996. Organic geochemistry applied to environmental assessments of Prince William Sound, Alaska, after the Exxon Valdez oil spill - a review. Org. Geochem. 24(1):7-42.
  • Christensen, L.B. and T.H. Larsen. Fall 1993. Method for determining the age of diesel oil spills in the soil. Ground Water Monitoring and Remediation. Fall Issue: 142-149.
  • Douglas, G.S., W.A. Burns, A.E. Bence, D.S. Page, and P.D. Boehm. 2004. Optimizing detection limits for the analysis of petroleum hydrocarbons in complex environmental samples. Environ. Sci. Technol. 38: 3958-3964.
  • Douglas, G.S. and A.D. Uhler. 1993. Optimizing EPA methods for petroleum-contaminated site assessments. Environ. Testing Analysis. 5:46-53.
  • Gruse, W.A. 1967. Motor Fuels. Performance and Testing. New York: Reinhold Publ. Corp. 280 pp.
  • Kaplan, I.R. 2003. Age dating of environmental organic residues. Environ. Forensics. 4:95-141.
  • Kaplan, I.R., Y. Galperin, S. Lu, and R. Lee. 1997. Forensic environmental geochemistry: differentiation of fuel-types, their sources and release time. Org. Geochem. 27:289-317.
  • Stout, S.A., A.D. Uhler, K.J. McCarthy. 2004. Characterizing the source of fugitive middle distillate fuels A case study involving railroad diesel fuel, Mandan, North Dakota. Environ. Claims J. 16:157-172.
  • Stout, S.A., A.D. Uhler, K.J. McCarthy, S.D. Emsbo-Mattingly. 2002. Chemical fingerprinting of hydrocarbons. In: B.L. Murphy and R.D. Morrison, Eds., Introduction to Environmental Forensics. Academic Press, Boston. pp.137-260.
  • Stout, S.A., A.D. Uhler, K.J. McCarthy, K.J. and S.D. Emsbo-Mattingly. 2001. The influences of refining on petroleum fingerprinting – Part 2. Gasoline blending practices. Contaminated Soil, Sediment & Water, Nov/Dec. Issue: 42-44.
  • Stout, S.A., A.D. Uhler, K.J. McCarthy, T. Naymik. 1998. Environmental forensics. Unraveling site liability. Environ. Sci. Tech., 32: 260 A-264 A.
  • Uhler, R.M., E.M. Healey, K.J. McCarthy, A.D. Uhler, A.D. and S.A. Stout. 2003. Molecular fingerprinting of gasoline by a modified EPA 8260 gas chromatography/mass spectrometry method. Int. J. Environ. Anal. Chem. 83(1): 1-20.
  • U.S. Environmental Protection Agency (2000). Fuel standard feasibility. In: Heavy Duty Standards/Diesel Fuel, RIA EPA420-R-00-026, 122 p.

Figure 1
Figure 1. General analytical approach and inventory of analyses conducted in the chemical fingerprinting of gasoline and diesel fuel. See Stout et al. (2002) for detailed descriptions.

Figure 2
Figure 2. Normalized PIANO distribution for two premium reformulated gasolines (RFGs). Refiner A achieved octane primarily from the blending of MTBE and iso-octane whereas refiner B achieved octane from MTBE and toluene.

Figure 3
Figure 3. Histograms showing the PIANO distribution for NAPLs from three locations in a study area. The presence or absence of trimethylpentane isomers (ISO, 234TMP, and 233TMP) (cross hatch) distinguish the Station A and Station B NAPLs from each other. Note that the distribution of BTEX compounds (grey), as would be measured with conventional EPA Method 8260, does not distinguish these NAPLs from one another.

Figure 4
Figure 4: Histogram showing the concentration of total sulfur (ASTM D4294) measured in 25 NAPLs (M#) and 8 dispensed diesel fuel #2 (D#). Superimposed on the histogram is the historic trend in total sulfur (averaged by year) in gasoline dispensed in the northeastern U.S. showing the significant reduction following the new Federal regulations in 1993. Most NAPLs and all dispensed diesel fuels from this site fall below the 0.5 wt% (500 ppm) limit (horizontal dashed line) established in 1993; these NAPLs were released after 1993.

This article is not a substitute for legal advice. Please consult with your legal counsel for specific advice and/or information. Read our complete legal disclaimer.