Independent Consultants in Environmental and Forensic Chemistry

Volume 3, Issue 1, Winter 1999

RBCA Petroleum Analyses - Sand but No Grit?

During the past five years, the risk based corrective action (RBCA) approach to petroleum release remediation has gained popularity among environmental professionals and regulators. In an attempt to apply RBCA to petroleum releases, new analytical methods have been developed to quantify toxicologically significant hydrocarbon chemicals. In 1995, the Massachusetts Department of Environmental Protection (MADEP) issued draft analytical methods for volatile petroleum hydrocarbons (VPH) and for extractable petroleum hydrocarbons (EPH).

The EPH analytical method is a solvent extraction, gas chromatography with a flame ionization detector (GC/FID) procedure which employs a post-extraction, pre-analysis silica gel/differential solvent fractionation process to separate aliphatic from aromatic compounds. Two separate injections into the GC are made, and three fractions are quantified: C₉ - C₁₈ aliphatics, C₁₉ - C₃₆ aliphatics, and C₁₁ - C₂₂ aromatics.

To test the methods, MADEP conducted blind round robin analyses on a set of samples prepared by an independent organization. Laboratories previously certified by MADEP for analyses of organic analytes in drinking water or wastewater were invited to participate; twenty-nine laboratories responded. After receiving results from 27 laboratories, MADEP addressed a letter to all laboratories identifying problem areas with data generation, manipulation, and reporting and provided the opportunity to modify submissions. Nineteen laboratories then provided modified reports, and another laboratory provided its initial report.

One of the samples of particular interest was a sand containing 1,200 Fg/g (ppm) of fresh #2 fuel oil. No weathering, biodegradation, or aging was allowed to occur. This sample represented a fresh release of #2 fuel oil into a medium that could not absorb or sequester the petroleum hydrocarbons. In other words, this blind sample represented a moderately contaminated soil that should have been relatively easy to analyze but was not a Areal world@ sample. Nonetheless, the average recovery of the petroleum hydrocarbons was slightly less than 80%. The acceptance criterion for laboratory performance was " 50% (increased from 40% after laboratories reported their results) of the mean recovery for each fraction (40 - 117% of the total petroleum hydrocarbons of 1,200 Fg/g). Even with the increased acceptance criterion, only eight (8) of the 28 reporting laboratories passed all three fractions for the EPH analyses. This is less than a 30% success rate.

Besides adding increased analytical complexity, and undoubtedly cost, to the project, the EPH analytical method is technically flawed. The aliphatic/aromatic fractionation step will give different results depending on the amount of silica gel and solvent used and the amount of petroleum hydrocarbons in the sample. If the acceptance criterion of " 50% of the mean reported (not actual amount) gives less than a 30% success rate on a fresh, non-weathered sample, what can be expected from aged and weathered Areal life@ samples? The EPH method approach is no more accurate than the less expensive TPH methods and a regulatory set TPH value.

Are the results using the MADEP EPH method an improvement, in terms of accuracy and precision, over EPA total petroleum hydrocarbon (TPH) methods 418.1 or 8015 modified? Are these the types of data on which to base risk assessments and subsequent corrective actions? We think not.

Nonetheless, the average recovery of the petroleum hydrocarbons was slightly less than 80%. The acceptance criterion for laboratory performance was " 50% (increased from 40% after laboratories reported their results) of the mean recovery for each fraction (40 - 117% of the total petroleum hydrocarbons of 1,200 Fg/g). Even with the increased acceptance criterion, only eight (8) of the 28 reporting laboratories passed all three fractions for the EPH analyses. This is less than a 30% success rate.

Besides adding increased analytical complexity, and undoubtedly cost, to the project, the EPH analytical method is technically flawed. The aliphatic/aromatic fractionation step will give different results depending on the amount of silica gel and solvent used and the amount of petroleum hydrocarbons in the sample. If the acceptance criterion of " 50% of the mean reported (not actual amount) gives less than a 30% success rate on a fresh, non-weathered sample, what can be expected from aged and weathered Areal life@ samples? The EPH method approach is no more accurate than the less expensive TPH methods and a regulatory set TPH value.

Are the results using the MADEP EPH method an improvement, in terms of accuracy and precision, over EPA total petroleum hydrocarbon (TPH) methods 418.1 or 8015 modified? Are these the types of data on which to base risk assessments and subsequent corrective actions? We think not.

What's in a detection limit?  -  Part 2 of 3  -  And then... what?

Appendix IX of 40CFR264 describes practical quantitation limits (PQLs) as "the lowest concentration in ground waters that can be reliably determined within specified limits of precision and accuracy by the individual methods under routine laboratory conditions." Appendix IX states that "PQLs are not a part of the regulation," that "the PQLs listed are generally stated to one significant figure," and that "the PQL values in many cases are based on a general estimate for the method and not on a determination for individual compounds."

No rigorous unified approach for the determination of PQLs for all analytes in all matrices has been established. The EPA document, Report on Minimum Criteria to Assure Data Quality (December 12, 1989), provides the qualitative PQL definition in terms of "the lowest analyte concentration in a given matrix that the Agency believes a competent laboratory can be expected to achieve consistently." EPA characterized the PQL as "analogous to the limit of quantitation (LOQ) as defined by the American Chemical society"

"as the level above which quantitation results may be obtained with a specified degree of confidence" The ACS also defined the limit of detection (LOD) "as the lowest concentration level that can be determined to be statistically different from a blank." The ACS recommended values of 3s and 10s for the LOD and LOQ, respectively, where s is the standard deviation of the measurements for a sample. The ACS stated that "the LOD is numerically equivalent to the MDL."

With the promulgation of SW846 Third Edition Update I (July 1992), the term, PQLs, were replaced with estimated quantitation limits (EQLs) and defined as "the lowest concentrations that can be reliably achieved within specified limits of precision and accuracy during routine laboratory operating conditions." This reference stated that "the EQL is generally 5 to 10 times the MDL" (method detection limit) and continued that "it may be nominally chosen within these guidelines to simplify data reporting." The numerical values reported for the EQLs were the corresponding ones previously reported as PQLs.

To compound the problem with detection limits, other environmental programs regulated by the EPA have established their own acronyms (e.g., contract required quantitation limits (CRQLs) for organics and contract required detection limits (CRDLs) for inorganics in the Contract Laboratory Program (CLP)). These detection limits are requirements of a legal document (the contract between EPA and its CLP laboratories) that depend more on administrative levels of concern than on technical ability or capability. Whereas MDLs and PQLs are dictated by the current state-of-the-art knowledge and techniques, CRDLs and CRQLs can be changed by the administrative stroke of the word processor.

Nonaqueous PQLs or EQLs are determined by a mathematical extrapolation of the corresponding aqueous values adjusting for the difference in the sample preparation quantities. For many semivolatile compounds, the aqueous 10 mg/L (parts per billion or ppb) limit is achieved by extracting a 1,000 mL (approximately 1,000 grams) sample. To achieve a similar detection limit in a solid matrix, a 1,000 gram (or one kilogram) sample would have to be processed. A kilogram of solid material is an unwieldy (although not impossible) quantity to process. Instead, a 30 gram sample is typically used in the analyses of soils. Since a smaller quantity (in terms of weight) of sample is used for the soil analysis, the quantity of contaminant available for extraction and detection is smaller. The MDL, PQL, EQL, etc., for a nonaqueous sample is, consequently, numerically higher than for an aqueous sample. For a 10 ppb aqueous PQL, this is equivalent to a 330 ppb soil PQL due to the fact that the typical 30 gram soil sample is 33 times smaller than the typical 1000 ml aqueous sample. The practice by some environmental professionals to correspond or equate a 10 ppb aqueous result with a 10 ppb nonaqueous result is to follow the teachings of Johnny Orangeseed. Are you confused yet?

PBMS .... what now?

You know how if you wait long enough, those old clothes in the back of your closet come back into style? Like the platform shoes and bell-bottoms that are "in" right now? Well, you can look at PBMS - performance based measurement systems - in just about the same way.

In the early days of environmental analyses, few reference methods and fewer regulations existed to tell you what to use and when to use them. Analysts worked with clients to evaluate problems and figure out the best way to obtain usable data from a sampling and analysis program.

As the environmental field grew, a desire for consistency and continuity grew with it, and prescriptive reference methods, the most familiar of which are the Contract Laboratory Program (CLP) and SW-846, were published. Use of these or similar "cook-book" methods was required by statute, with little or no allowed flexibility.

Now, new instruments and methods come out so quickly that it is difficult to keep pace. Yet, published methods don't allow the use of these improvements, and formal approval of revised methods can take years, by which time what has been approved may already be outdated. What's a laboratory to do?

PBMS is the EPA=s response to the current situation. Defined as a set of processes that specifies the data quality needs of a program and serves as criteria for selecting appropriate and cost-effective methods to meet those needs, PBMS will allow the regulated community to select any appropriate analytical test method to comply with EPA=s regulations. Ideally, laboratories and/or consultants will be encouraged to work with clients to evaluate problems and determine the best way to obtain usable data from a sampling and analysis program ... much like in the early days.

Sounds good! Agency approval will not be required to use new or revised methods, and the regulated community can make its own decisions about what is an acceptable method for their purposes and even what is acceptable performance. What could be better?

But PBMS is not without its pitfalls. With the freedom to make its own decisions about methods and performance, the regulated community accepts responsibility for those decisions as well as for demonstrating that they really work; what constitutes demonstration of acceptable method performance is still one of the biggest sources of contention in the PBMS arena.

Implementation of PBMS by parties who do not understand the environmental problem or the analytical methods can lead to poor quality, even worthless, data that may be used anyway. It would be very easy to abuse a PBMS system, cutting corners and sacrificing data quality to save money. Comparability of data may suffer from use of widely divergent methods, and legal defensibility is still a major question mark.

Many factors still need to be worked out, and we can expect to hear more on PBMS from EPA soon. If there were ever a time to have a solid working relationship with an environmental laboratory you trust, it is now. Caveat emptor!

Book Review:  Dense Chlorinated Solvents in Porous and Fractured Media, Freidrich Schwille, translated by James F. Pankow, Lewis Publishers

Chlorinated hydrocarbons (CHCs) such as trichloroethene (TCE), tetrachloroethene (PCE), and 1,1,1-trichloroethane (111TCA) have been recognized as major environmental contaminants. Considering the volatilities of these compounds, their discovery in groundwater systems was initially surprising, and the transport mechanisms of CHCs to the groundwater were not understood.

ADense Chlorinated Solvents...@ presents the pioneering studies of Dr. Freidrich Schwille, chief groundwater hydrologist at the Federal Institute of Hydrology in Koblenz, West Germany, from 1960 to 1985 on the behavior of CHCs in the environment. The text is amply supplemented by numerous figures and color photographs of the experiments conducted in studying the behavior of CHCs as pure fluids, in solution, and as gases in porous and fractured media in the subsurface environment . The body of the book occupies only 127 pages (excluding appendices), and many of these pages are occupied by the figures and color plates. Concise descriptions and diagrams of laboratory experiments are presented along with the rationales for particular approaches and arrangements. The text reads much like a well narrated laboratory notebook.

The color photographs alone demonstrate how CHCs 1) do NOT migrate unabated through the unsaturated zone from a surface spill (Abecause they are heavy liquids@), 2) do NOT necessarily penetrate the saturated zone (Abecause they are heavier than water@), and 3) do NOT necessarily form a pool at the bottom of an aquifer (Abecause they are immiscible with water@). ADense Chlorinated Solvents...@ contradicts many of the common misconceptions advanced by today=s environmental Aexperts.@ Must reading for the environmental professional!