Underground Storage Tanks

PIRI's (Partnership in RBCA Implementation) official members include representatives from the major oil companies (i.e., Amoco, BP, Chevron, Exxon, Mobil, and Shell), ASTM, EPA's OUST, and States. The paper presented here represents the points of view of the various PIRI authors and not the companies or states with which they are affiliated. This document is not an EPA document. This document will be updated as changes in technologies and policies require. It is critical that readers check with the implementing agency in their States before acting upon the policies discussed in these papers.

Total Petroleum Hydrocarbon (TPH) analysis is widely used as a general measure of the presence of crude oil or petroleum product in soils. TPH is defined as the measurable amount of petroleum-based hydrocarbon in an environmental media (e.g., soil, water, sediments) and, thus, is dependent on analysis of the medium in which it is found. While providing an overall concentration of petroleum hydrocarbons, TPH itself is not a direct indicator of the risk (i.e., mobility, toxicity, and exposure to human and environmental receptors) posed by petroleum hydrocarbon contamination. Both mobility and toxicity are very dependent upon the relative amounts of individual (or groups or families of) constituents within a hydrocarbon mixture. For example, a crude oil may contain different types and amounts of aromatic compounds than does a gasoline. Other analysis or information in addition to a single TPH number must be used to relate TPH concentrations to risk.

This paper presents an overview of the measurement of TPH and the methods that have been used in the past to estimate the risk from TPH contamination. More recent developments, including those of the Total Petroleum Hydrocarbon Criteria Working Group (TPHCWG), which is a national ad hoc committee that was formed to develop technically defensible risk-based approach to TPH, are also discussed. The paper concludes with examples of how the TPHCWG technology might be used by State regulatory agencies to incorporate TPH into a risk-based corrective action program.

Measurement Of TPH

A TPH method of analysis often used and required by many regulatory agencies is EPA Method 418.1. This method provides a "one number" value of TPH in an environmental media; it does not provide information on the composition (i.e., individual constituents such as benzene) of the hydrocarbon mixture. The amount of TPH measured by this method depends on the ability of the solvent used to extract the hydrocarbon from the environmental media and the absorption of infrared (IR) light by the hydrocarbons in the solvent extract. Method 418.1 is not specific to hydrocarbons; in fact, it can give false positive results when organic matter is extracted from the environmental media for analysis. In other words, TPH measurements do not always indicate petroleum contamination (e.g., humic acid).

Another analytical method commonly used for TPH is EPA Method 8015 Modified. This method reports the concentration of purgeable and extractable hydrocarbons which are also sometimes referred to as gasoline and diesel range organics (i.e., GRO and DRO) because the boiling point ranges of the hydrocarbon in each roughly corresponds to that of gasoline (i.e., C₆ to C_10-12) and diesel fuel (i.e., C_8-12 to C_24-26), respectively. Purgeable hydrocarbons are measured by purge-and-trap gas chromatography (GC) analysis using a flame ionization detector (FID), while the extractable hydrocarbons are analyzed by GC following extraction with a solvent and subsequent concentration of the extract by evaporation. While more detailed information is generated by this method (e.g., GC chromatograms), the results are most frequently reported as single numbers for purgeable and extractable hydrocarbons.

The Massachusetts Department of Environmental Protection (MADEP) has developed a method based on 8015 Modified which gives a measure of the aromatic content of the hydrocarbon in each of several carbon number ranges. In the MADEP method, the lighter hydrocarbon fractions (i.e., carbon numbers from C₆ to C₁₂) are analyzed by purge-and-trap GC analysis using a flame ionization detector (FID) to measure the total hydrocarbons and a photo ionization detector (PID) to measure the aromatics (e.g., benzene) with the aliphatic (e.g., hexane) component of the TPH being found by difference. The aromatic and aliphatic fractions are divided into carbon number fractions based on the normal alkanes (e.g., n-octane) as markers. The heavier hydrocarbons (i.e., C₁₂ to C₂₆) are analyzed using an extraction procedure followed by a column separation using silica gel (Modified EPA Method 3630) of the aromatic and aliphatic groupings or fractions. The two fractions are then analyzed using GC-FID. PAH markers and n-alkane markers are used to divide the heavier aromatic and aliphatic fractions by carbon number, respectively.

The MADEP method is based on standard EPA methods (i.e., 8020/8015 Modified) which allows it to be easily implemented by contract laboratories. However, there are some concerns or issues about the method. One issue is that the PID is not completely selective for aromatics (i.e., it does respond to some aliphatic compounds). Thus, this approach can lead to an overestimate of the more mobile and toxic aromatic content. Another issue with this analytical approach is that the results from the two analyses (i.e., purgeable and extractable hydrocarbons) can overlap in carbon number and thus may not be simply added back together to get a total TPH concentration. The performance of this method on real hydrocarbon products may be limited.

The TPH Criteria Working Group has developed a method for identifying and quantifying the presence of the groups or fractions with similar mobility in soils. The technique is based on EPA Method 3611 (Alumina Column Cleanup and Separation of Petroleum Wastes) and EPA Method 3630 (Silica Gel Cleanup), which are used to fractionate the hydrocarbon into aliphatic and aromatic fractions. A gas chromatograph equipped with a boiling point column (non-polar capillary column) is used to analyze whole soil samples as well as the aliphatic and aromatic fractions to resolve and quantify the fate-and-transport fractions selected by the TPH Criteria Working Group. The method is versatile and performance-based and, therefore, can be modified to accommodate data quality objectives.

Estimates Of Risk For TPH

There are three basic approaches that have been used in the past to estimate potential human health risks posed by TPH contamination. The one most generally applied and most appropriate for evaluation of the carcinogenic risk from TPH is an "Indicator" approach. This approach assumes that the estimated risk from TPH is characterized by a small number of indicator compounds (e.g., BTEX, PAHs). This approach was necessitated by the inability to analyze for the large number of constituents in TPH and the lack of toxicological and other relevant data for many of those constituents that could be individually identified. The indicator approach is generally accepted and used by state regulatory agencies for carcinogenic risk posed by TPH. The use of the indicator approach for determining non-carcinogenic risk has, however, not been fully developed.

Another approach, the "Surrogate" approach, assumes that TPH is not specifically included as an indicator and can be characterized by a single surrogate compound. This approach could overestimate toxicity and mobility because of the compounds typically available for use as surrogates. For example with respect to toxicity, benzene is the surrogate compound for the aromatics; it is also the most carceno- genic. With respect to mobility, benzene is, again, the surrogate compound because it travels at a faster rate than other petroleum constituents. Benzene is, however, the least abundant of the constituents found in petroleum mixtures. A variant of the "Surrogate" approach is the "Whole Product" approach in which the toxicity and mobility of the TPH product are based on that of a whole product of similar character. Neither the "Surrogate" nor the "Whole Product" approach is capable of taking into account the effects of weathering (i.e., changes in composition and toxicity and mobility over time) and the wide range of mobility of the constituents of the typical hydrocarbon product. Because the lighter and more mobile constituents tend to weather faster, weathered crude oils and hydrocarbon products are typically less mobile and, thus, could pose a lower risk (although mobility is only one of several factors used to deter- mine risk). These approaches are similar to the "Indicator" approach in that they use specific knowledge of a single or a few constituents to characterize the many constituents in a hydrocarbon mixture.

More recently, approaches have been developed which are a compromise between the "Indicator" and the "Surrogate" or "Whole Product" approaches. In these approaches, carcinogenic risk is estimated based on indicators (e.g., benzene and the carcinogenic PAHs) while the non-carcinogenic risk from the TPH is estimated based on a relatively small number of groupings or fractions. Each of these groups or fractions is composed of constituents of TPH that have similar toxicity and mobility characteristics. The risk as a result of TPH contamination can be estimated for each of the groupings or fractions as individual contaminants or for the measured TPH (i.e., the sum of all the groups or fractions) by assuming additivity of risk. The estimated risk approach to TPH developed by the Massachusetts Department of Environmental Protection and the Total Petroleum Hydrocarbon Criteria Working Group are examples of this compromise approach.

TPH Criteria Working Group

The TPHCWG approach is a combined indicator and grouping or fraction approach. Note that the non-carcinogenic indicators (i.e., TEX and non-carcinogenic PAHs) are included in the grouping or fraction analysis and do not need to be analyzed as indicators (although, this can be done if desired by backing them out of the TPH analysis). The basic approach is similar to that developed by MADEP in that the TPH is split into a small number of groups or fractions that have similar properties. The main difference between the approaches is that in the TPHCWG approach, the groups or fractions of TPH are defined based on the potential mobility of the hydrocarbons within each group; in the MADEP approach, they are based on the available toxicity data.

The TPHCWG approach is not a surrogate approach in which the physical/chemical and toxicological properties of the grouping or fractions are based on single surrogate compounds. The physical/chemical and toxicological properties of each of the groups or fractions are based on all available data for individual constituents, well defined mixtures, and/or whole products that are representative of each group or fraction. Thus, the TPHCWG approach is an extension of the MADEP approach, based on a more complete database of physical/chemical and toxicological properties and specifically taking into account the variability in the potential mobility of the petroleum hydrocarbon groups or fractions. The approach developed by MADEP does an adequate job of assessing the risk from TPH for direct exposure scenarios, while the approach developed by the TPHCWG is better suited for addressing cross media exposure pathways such as soil leaching to groundwater.

The TPHCWG has completed its effort and final reports are available on the Internet ( http://www.aehs.com/publications/catalog/tph.htm). The working group is still finalizing the document (Volume 5) which will incorporate all of the findings into a sample risk-based decision making framework. The TPHCWG approach is, however, summarized in a technical overview document that is available on the web page and has been presented at numerous conferences and workshops.

Incorporating TPH Into A Risk-Based Corrective Action Program

The petroleum fate-and-transport fractions and toxicity criteria developed by the Working Group can be used in the ASTM/RBCA framework to compare certain fractions obtained from sampling results to previously developed RBSLs or to develop a compound-specific risk-based screening level (RBSL). The soil risk-based screening levels (RBSLs) that are calculated for the indicator compounds and for the fate-and-transport fractions can be used individually as a basis for the management of a contaminated site. In this instance, site remediation will be governed by the most restrictive or lowest soil RBSL. Alternatively, the composition of the total petroleum mixture present at a site can be used to yield a soil RBSL for TPH. The RBSL for TPH is calculated by assuming that the risk for the individual compounds and fate-and- transport fractions can be added, each weighted by their composition in the total petroleum mixture. The Working Group's Volume 5 provides a thorough discussion of how the approach may be used within the RBCA framework, including risk calculations and results from demonstration sites where this new approach has been used.

The approach developed by TPHCWG is intended to provide the technical basis for a broad range of regulatory programs. Thus, while it can be used to develop risk-based analysis for the individual TPH groupings, the approach can also be used to develop a risk assessment or risk-based screening levels for TPH (i.e., the sum of the fractions). In addition, the application of the full analytical method and risk analysis may not be needed for all soil samples collected at a petroleum-contaminated site. Once the petroleum composition has been fully characterized at a site, additional sampling can rely on traditional, less expensive TPH analysis rather than the new Working Group method (if the TPH fingerprint is similar across the site). This simplification can be carried further if process knowledge for a site can be used to characterize the hydrocarbon contamination such that the composition of the petroleum hydrocarbons at the site can be based on non-site specific analysis (e.g., typical jet fuel at an Air Force base).

These simplifications are important for application of the TPHCWG technology at sites where the cost of the more detailed analysis is not justified. For example, the full TPHCWG analysis may not be cost effective at heavy hydrocarbon sites where significant concentrations of the indicator compounds are not anticipated, historical data for TPH does not allow identification of the individual groupings, and/or additional analysis is not cost effective. The full TPHCWG analysis may also not be cost effective for sites such as retail/marketing sites with underground storage tanks where information about the type of hydrocarbon is available (i.e., the contamination is known to be gasoline from an underground storage tank). For these cases, generic risk-based TPH screening levels based on typical composition data (i.e., a typical weather gasoline) can be developed, and less expensive TPH analysis or indicator analysis can be used to characterize the site. The TPHCWG technology can be adapted to fit within and support a broad range of regulatory programs.

One possible use of the TPHCWG technology is to provide the technical basis for the assumption that the risk from TPH other than the indicator compounds (e.g., BTEX and the PAHs) is not significant under certain exposure scenario assumptions. In this case, the risk from TPH contamination would be evaluated based on the analysis of indicator compounds or on some non-risk-based criteria such as the potential mobility of the TPH as a non-aqueous phase liquid or NAPL. Ohio is a good example of a State that has developed a look-up table for TPH based on soil types and the boiling range of the hydrocarbon contamination. The criteria were developed using models to estimate the percent of saturation required for the non-aqueous phase liquid (NAPL) hydrocarbon to be mobile. The States of Hawaii and Louisiana have developed similar policies where the upper limit for TPH screening levels (i.e., 5000 mg/kg and 10,000 mg/kg, respectively) are based on aesthetics or some other non-risk-based criteria. Another possible use of the TPHCWG technology is to develop generic criteria for TPH based on typical compositions of fresh and/or weathered hydrocarbon mixtures. A good example is a recent analysis of crude oils in the State of Michigan. Analyses of five representative crude oils from across the State were used to support cleanup levels of 10,000 mg/kg. At this concentration of TPH (i.e., 10,000 mg/kg), the concentrations of the carcinogenic indicator PAHs and the TPHCWG fate-and-transport fractions are below levels of concern (i.e., concentrations are less than RBSLs). The state regulatory agency is currently adopting 10,000 mg/kg as a generic screening level for TPH at all crude oil contaminated sites.

NOTE: The technical documents of the TPHCWG are in press or in preparation. Some of these documents are now available on the Association for the Environmental Health of Soils' web site ( http://www.aehs.com/). EPA neither endorses nor discourages the TPHCWG approach. EPA recognizes that TPH may be used as a screening tool for the measurement of total hydrocarbon contamination, but cautions states that it is inconsistent with a risk-based approach that focuses on individual chemicals and their risk to human health.

Top of page