Independent Consultants in Environmental and Forensic Chemistry
Volume 3, Issue 2, Spring 1999
Environmental Forensic Science: Expertise or Advocacy?
Environmental forensic science is the latest Ain@ field for consultants and experts. Its popularity is evident by the 15 presentations given at the annual American Academy of Forensic Sciences (AAFS) Engineering Sciences section meeting in Orlando, Florida. Two weeks later in Oxnard, California, seven papers and a workshop were presented in environmental forensics at the 9th Annual Meeting of the West Coast Conference on Contaminated Soils and Water. Other papers concerning topics in environmental forensics are presented at numerous other conferences throughout the country.
The best environmental forensic science is a multidisciplinary effort. No single person has all of the needed disciplines to provide technically defensible positions on contaminant sources and ages of releases. The emphasis is on a team concept of experts including aerial photogrametrists, chemists, engineers, historians, hydrogeologists, toxicologists, etc. A team of independent experts from different organizations provides excellent quality control through the process of Adevil=s advocacy@ and by interpreting all of the data and information into a technically sound and consistent final concept.
As with any growing field, numerous claims and opinions are proffered by individuals with little or no verification from independent sources. Oftentimes, these claims and opinions are based on experts= experiences, not on any definable scientific facts or methods. These experts promise to support whatever position is needed or desired. The methods used to arrive at the opinion cannot be described in detail, are intangible, and are, many times, confidential. For example, some individuals tout the ability to age-date a petroleum release based on its hydrocarbon pattern. However, the characteristics of the hydrocarbon pattern that allow the age to be estimated cannot or will not be defined. Since defensible forensic science must be verifiable, the methods used to arrive at the opinions must be definable and open to scientific scrutiny.
Another technique commonly used by experts is to ignore evidence inconsistent with their opinions and to use only those facts that support their current positions. In this manner, an expert can rely on the other facts to support his position when he has to argue the opposite side in other similar situations. Why some facts carry more weight in one situation and less weight in other situations remains a mystery.
The confidentiality and secrecy in methods and the selection of supporting facts allow an expert to fit the outcome of his musings to the desired position of his client rather than to arrive at logical conclusions and reasonable opinions based on all of the available facts. These techniques are contrary to the traditional application of the scientific method. Opinions and conclusions, therefore, do not fall into the realm of reasonable scientific certainty. When an expert is asked AWhat is the basis for your opinion?@ and the answer is Amy experience,@ that answer must be thoroughly explored before it is accepted. When an expert cannot reveal his methods, ignores evidence contrary to his position, or fits the same data to whatever side he is supporting, he is no longer an expert. He is an advocate.
When is a Blank not a Blank?
Recently, soil sampling was conducted at over 1,000 locations on an industrial property. After reviewing the data, the client was surprised to notice the presence of toluene in the VOC results. The toluene seemed to be found randomly throughout the site with respect to both location and depth. The concentrations varied from a few parts per billion (ppb) to a few parts per million (ppm). No source of toluene was known in the site=s history, but petroleum fuel storage and an incinerator had been on the property. What was the probability of a natural source of low quantities of toluene in the soil? Could the toluene be the result of fuel releases from tanks, vehicles, or the incinerator at the facility? What was the source of the toluene?
The soil samples were collected in six inch brass sleeves. The ends of the sleeves were covered with teflon tape (3 inch wide) and capped with polyvinylchloride caps which were secured to the brass sleeves with black electrical tape. Following EPA protocol, rinsate blanks were collected during sampling but no toluene was found in the blanks.
Headspace-gas chromatography of the teflon tape, polyvinylchloride cap, and electrical tape was conducted to determine if these materials might be the source of the toluene. While toluene was not found in the teflon tape or polyvinylchloride caps, the electrical tape yielded considerable quantities. Both teflon tape and PVC are plastics that are permeable to solvents such as toluene. Toluene apparently originated in the electrical tape, penetrated the PVC caps and teflon tape, and contaminated the soil in the brass sleeves.
The source of the toluene contamination may have been determined more easily if a true field blank had been collected. A true field blank exposes a similar matrix to the same conditions that each sample experiences. In this case, soil samples should have been collected from an area known to be free from the analytes of concern and treated in the same manner as the samples including the use of the teflon tape, PVC cap, and electrical tape. Toluene contamination of this type of field blank would have suggested problems in the sampling procedure.
A rinsate blank is not a field blank. A rinsate blank will assist in determining if cross-contamination is occurring due to inadequate cleaning of the sampling equipment between samples. In this case, the rinsate blanks provided no information regarding the source of toluene contamination. The fact that the rinsate blanks were clean did not mean that contamination problems did not exist in the field.
Detection Limits - Part 3 of 3 - You mean there=s more?
We=ve talked about EPA=s infamous MDL as well as PQL and EQL; covered the contract laboratory program terms CRDL and CRQL, and described two contributions from the American Chemical Society (LOD and LOQ). And yes, there=s more.
The American Society of Testing and Materials (ASTM) has their interlaboratory detection estimate or IDE. Published as a standard (D6091-97) in 1997, the IDE is briefly defined as Athe lowest concentration at which reliable detection can occur.@ More specifically, it is Athe lowest concentration at which there is 90% confidence that a single measurement from a [single qualified laboratory] will have a true detection probability of at least 95% and a true nondetection probability of at least 95%.@ The qualified laboratory would be one of those that participated in the interlaboratory study
The IDE differs from the MDL in two key ways: (1) it is calculated by an oversight body (e.g., ASTM or EPA) that is independent of the laboratory and (2) it is based on the results
of an interlaboratory study that incorporates data from a minimum of six laboratories, rather than on results generated by a single laboratory. As a result, the IDE includes interlaboratory variability. IDE also accounts for variations in precision at different concentrations by analyzing a concentration range at each participating laboratory. An interlaboratory precision function, similar to that used to calculate an MDL based on measurements at a single concentration, is also used.
The IDE is probably a more realistic estimate than the MDL of the smallest amount of an analyte that a typical laboratory could measure on a typical day. However, very little multilaboratory data exist from which IDEs can be established, and little incentive currently exists for laboratories to develop this approach. Use of the ASTM standard is not required at this time under any regulations.
Other terms such as IDLs (instrument detection limits), LLDs (lower limits of detection), RLs (reporting limits), LCs (critical levels), and MLs (minimum levels) exist. Each of these terms is defined as something similar to the terms already discussed and often used in the course of defining other terms. This is some serious alphabet soup!
But what do we really need to know? Research and method development needs aside, what most data users need to know is - what is the lowest concentration at which I can safely and reliably say an analyte is not present in my samples? And, of course, is that concentration below the applicable regulatory limit at my site?
Too often, a laboratory will report volatile organic analytes as Aless than the MDL@ (which may be as low as 0.1 ppb). Quite likely, that MDL was established months ago on a different instrument when the associated calibration curve was established using concentrations ranging from 5 ppb to 150 ppb. Since the matrix spike concentration associated with your samples was 25-50 ppb, nothing in your supporting data can demonstrate that, had 0.1 ppb of an analyte been present, it would have been detected on that instrument on the day that your samples were analyzed. However, you are assured that that analyte is not present above this concentration. Don=t buy it!
EPA Method 8000B (12/96) states that AThe lowest concentration calibration standard that is analyzed during an initial calibration establishes the method quantitation limit ...@ It also states that ALinearity through zero (this is the assumption made when the average relative response factor is deemed acceptable for sample calculations) is a statistical assumption and not a rationale for reporting results below the calibration range demonstrated by analysis of the standards.@ The bottom line is, if your regulatory action limit is 0.1 ppb for a given analyte, the lowest initial calibration standard concentration should be 0.1 ppb for that analyte. This way (hopefully), you will have pertinent and reliable data to support the Aless than 0.1 ppb@ result that you reported.
A viable alternative to actually including such a low concentration in the initial calibration is to have the laboratory analyze a matrix spike at the action limit concentration using one of your site samples that has no analyte present as an additional quality control sample. This should not replace the routine matrix spike at the midpoint concentration. If the analyte is not recovered, you at least know you have a problem right away. Or, with good recoveries, you will know that the analyte would have been detectable at the action limit, even if the quantitation at that concentration is estimated.
It=s pretty simple, ... know what you need, make sure the lab knows what you need, and work with them to generate the data you need to support your results.
Did you know... ???
Although commonly used interchangeably, Anot detected (ND)@ and Abelow the detection limit (BDL)@ do not necessarily mean the same thing. AND@ implies that the analyte of interest was not observed. ABDL@ implies the analyte may or may not have been observed but, if it were observed, its concentration was below the claimed detection limit. If your results contain AND@s, you need to know how much has to be present to be detected. If your results contain ABDL@s, you need to know what detection limit the lab uses and how the detection limit was determined.
Book Review: GC-MS Guide to Ignitable Liquids, Rita Newman, Michael Gilbert and Kevin Lothridge, CRC Press LLC, Boca Raton, Florida
This text is comprised of gas chromatographic-mass spectroscopic (GC-MS) data for over 100 different commercially available ignitable liquids. The liquids include products from adhesive remover to Zippo lighter fluid, and some are analyzed at different evaporation percentages. All liquids are analyzed under identical conditions which are provided.
The data for each liquid are presented in several formats. For example, Page One of the data for Stoddard Solvent consists of two total ion chromatograms. The first chromatogram gives a time scale from 4 to 24 minutes for comparison with any of the other products. This chromatogram includes a list of four or five known chemicals in the substance such as nonane, decane, undecane, and dodecane in the Stoddard Solvent. The second total ion chromatogram expands the 5 to 15 minute range for a comparison with products having similar boiling temperature ranges. Page Two has four chromatograms from selected ion monitoring (SIM) for alkanes, aromatics, cycloparaffins/alkenes, and naphthalenes. Pages Three and Four give the individual SIM chromatograms for the above four categories. The ions monitored are 43, 57, 71, and 85 for alkanes; 91, 105, and 119, for aromatics; 55, 69, and 83, for cycloparaffins/alkenes; and 128, 142, and 156, for napthalenes.
Although published for those interested in analyzing fire debris, we have found this text to be an excellent reference for the identification of petroleum products released into the environment. AGC-MS Guide to Ignitable Liquids@ is an excellent reference for elucidating the composition of petroleum products that may be found in environmental samples. Weathering of these products can be estimated from the delineation of their composition and from a knowledge of the weathering properties of each class of chemicals that compose these products. We recommend this book to anyone concerned with the analysis and characterization of hydrocarbon releases into the environment.