Home_____________________Forensic Techniques


Forensic Techniques for Establishing the Origin and Timing of a Contaminant Release

submitted by:
Robert D. Morrison, Ph.D.
R. Morrison & Associates, Inc.
201 E. Grand Avenue, 2B
Escondido, CA 92025

Common issues in environmental coverage cases include determination of the origin of the contaminant release, the timing of the release, the distribution of the contamination in the subsurface, and the reasonableness of the alleged remediation costs. Of these issues, the origin and timing of a release are pivotal and frequently the most difficult to resolve. The timing of a release and whether it is sudden and accidental is especially relevant given that it is foundational as to when a particular insurance policy is "triggered". The selection of triggering theories such as "exposure" (when the property was exposed to hazardous waste), "manifestation" (discovery of the contamination), "injury in fact" (the date at which the property damage occurred) and "continuous trigger" (the continuous damage to property throughout all policy periods) is influenced by the ability to identify the timing of a release.

In instances where conclusive evidence is lacking to identify the timing or origin of a release, forensic review of the environmental data may provide this important information. Techniques available include aerial photography interpretation, identification of the date the chemical was commercially available, association of the chemical with a manufacturing process, chemical profiling, chemical degradation models, and groundwater modeling. These techniques individually, or in concert with other information, can provide the evidence needed to identify the timing and origin of a release.

Aerial Photography Interpretation

Interpretation of aerial photographs can provide convincing evidence in establishing the timing of a contaminant release. The successful use of this technique is dependent on acquiring a complete list of coverage dates and retaining an expert in aerial photographic interpretation. The latter is important because the interpretation of aerial photographs is dependent on both the sophistication of the diagnostic equipment as well as the experience of the interpreter.

Acquisition of a complete list of aerial photographs requires a firm that specializes in such acquisitions. Public sources of aerial photographs include the Summary Record System maintained by the United States Geological Survey, the United States Department of Agriculture, the National Archives, the United States Forest Service, the National Oceanic and Atmospheric Administration, the Army Corps of Engineers and local state Highway Departments. Specialized firms are aware of and can access private collections and brokers of photography (Chinese and Russian satellite imagery) which can provide the crucial evidence for identifying a surface feature that is consistent with a release.

Wherever possible, acquisition of stereo pairs is required as it allows features such as stream channels, barrels, disposal areas, pits, above ground tanks, drainage ditches, lagoons, or settling ponds to be examined three dimensionally. The presence of dark staining in areas of known contamination is also useful in establishing a causal relationship for arguing that a release occurred within a time period established with the aerial photographs.

Commercial Availability of a Chemical

If the chemical of concern is known, the date of its commercial availability can identify the earliest time period that it could have been present. Chlorinated hydrocarbons, chemicals associated with fuel additives, pesticides, herbicides, fungicides, and insecticides are especially amenable to this type of dating analysis. Knowledge of the synomyns and trade names of the chemical is necessary to perform this analysis because a chemical can have numerous trade names with no similarity to the chemical name. Numerous texts are available for this examination (1). Exhibit 1 summarizes when selected chemicals became commerically available in the United States.

Exhibit 1 Commercial Availability of Selected Chemicals
Chemical Date
Dibromochloropropane 1955
DDT 1942
Aldrin 1948
Bromacil 1963
Dieldrin 1948
Dinoseb 1945
Parathion 1947
Phorate 1954
Trifluralin 1960
Carbon tetrachloride 1907
Trichloroethylene (TCE) 1908
Perchloroethylene (PCE) 1925
Chlordane 1947
Toxaphene 1947
Chloroform 1922
1,2 dichloroethane 1922
1,1,1 trichloroethane (1,1,1 TCA) 1946
1,1,2 trichloroethane 1941-3 1941-3
1,2 -dibromo-3-chloropropane 1955
Chemicals Unique to a Manufacturing Process

An understanding of a site's manufacturing processes and material - handling systems can provide insight into probable locations of a release. For a semiconductor manufacturing site, likely locations where chlorinated solvents can enter the subsurface include neutralization sumps, corroded sewer and transfer piping, and chemical storage areas. The chemical distribution of chlorinated solvents in the subsurface in relation to these features can then assist in developing a causal relationship between these features and the observed contamination.

The spatial association of a particular chemical with various unit processes at the facility is used to further define the origin of the release. Chemicals, such as chlorinated solvents, uniquely associated with particular equipment can provide insight into probable source locations. An example is the use of chlorinated solvents in degreasing operations. In 1970, trichloroethylene (TCE) accounted for 82 percent of all chlorinated solvents used in vapor degreasing; by 1976, its share had declined to 42 percent. (2). A degreaser is also designed to handle solvents only within a certain boiling range; obtaining the manufacturer's operating manual for the degreaser can provide information concerning the inclusion or exclusion of chemicals used by the degreaser. This information, coupled with a knowledge of when the degreaser commenced operation and its historical location, can help identify where and when a release of a particular chlorinated solvent occurred.

Using information on a chemical's application to identify the timing of a release is also helpful for soil contamination by polychlorinated biphenyls (PCBs). PCBs are listed on laboratory reports with a numbered designation, such as PCB-1254; 12 refers to the number of carbon atoms and 54 refers to the number of chlorines. Different carbon and chlorine combinations are manufactured for specific uses during discrete time periods. For example, PCB-1016 was manufactured between 1971 and 1976 and used as an insulator fluid for electric condensers and as an additive in high-pressure lubricants. PCB-1254 was used as a secondary plasticizer in the manufacture of polyvinyl chloride and in capacitors; it was produced from 1957 to 1977 (3). The date of manufacture of PCBs can therefore be combined with its' particular use to bracket the date at which it was available and identify the location of the equipment at the facility that would use a particular PCB formulation.

Chemical Fingerprinting

Chemical fingerprinting is used to describe the ability to distinguish the age and often the origin of a chemical. It is most commonly used in hydrocarbon contamination cases. In its simplest form, chemical fingerprinting identifies the type of hydrocarbon (e.g., diesel, gasoline, jet fuels, kerosene, mineral oil, Stoddard solvent) as a means to identify the source and timing of a release. More sophisticated techniques to identify the age of the hydrocarbon include the examination of the composition of fuels as a function of the time of formulation or the identification of additives associated with a discrete time period. These techniques can often determine whether a chemical release was a single event, a series of events, or a continuous release.

Techniques used to identify the age of a hydrocarbon release include analysis of proprietary additives, the composition of antiknock formulations, trace metals analysis, hydrocarbon profiling, physical characteristics, and degradation models. Some are described below.

Proprietary Additives

Proprietary additives are compounds added to hydrocarbon products for specific purposes. Additives often have discrete time intervals during which they were introduced into a formulation. The use of additives for hydrocarbon fingerprinting requires a prior knowledge of the additive package and the ability to detect a unique additive that is not masked by other chemicals or obscured by degradation.

Gasoline additives include antiknocks, antioxidants, metal deactivators, corrosion inhibitors, anti-icers, lead scavengers, and oxygenates. Exhibit 2 lists examples of each.

Exhibit 2: Examples of Gasoline Additives
Type of Additive Examples
Antiknocks Alkyl leads
Antioxidants p-phenylenediamine; alkyl-substituted phenols
Metal Deactivators Disalicylpropanediamine
Corrosion Inhibitors Carboxylic acids and diimides
Anti-icers Short-chained n-alcohols (freeze point depressants); amines and ethoxylated alcohols with long hydrocarbon chains.
Oxygenates Methanol, methyl-tertiary butyl ether (MTBE)
Lead Scavengers Ethylene dibromide
The composition of additive packages vary with time. For example, a typical additive package for gasolines formulated in the 1980s was 62 percent tetraethyl lead, 18 percent ethylene dibromide, and 2 percent inactive ingredients such as stability improvers, dyes, and antioxidants.[4]

Diesel fuels also contain additive packages. Diesel often contains quality-enhancing additives such as diesel ignition and stability improvers, corrosion inhibitors and surfactants that can similarly be associated with discrete periods of time (5).

Antiknock Additives

Alkyl leads are the most frequently encountered antiknock additive. The first reported use of a lead antiknock additive was in 1923 by the General Motors Development Company in Dayton, Ohio, which used tetraethyl lead. By 1950, most gasolines in the United States were leaded; in 1960, tetramethyl lead was marketed. Consumption of all lead alkyls peaked in 1969 (6). For premium-grade gasolines, these concentrations were as much as 2.9 grams per gallon. Subsequent reductions in lead concentrations in gasoline occurred from the late 1970s to 1985 due to regulatory concerns. Tetraethyl lead was reportedly the only alkyl lead added to leaded fuels after 1980. Only tetraethyl lead is currently used as an additive in leaded gasoline in amounts up to two orders of magnitude less than that added before 1980. The history of lead additives to gasoline frequently provides a basis for bracketing the age of the gasoline in the soil or groundwater.

The most common organic lead alkyl additives are tetraethyl, triethylmethyl, methyldiethyl and tetramethyl lead. Redistribution reactions of equimolar amounts of tetraethyl and tetramethyl leads can also produce trimethyl, trimethylethyl, dimethyldiethyl, and methyltriethyl lead [7] . Older gasolines included these lead compounds, along with lead scavengers such as ethylene dibromide and 1,2-dichloroethane; these additives were first introduced in 1928. Exhibit 3 summarizes the concentration of leads in a leaded gasoline(8).

Exhibit 3 Concentrations of Lead in Leaded Gasoline
Leaded Gasoline Additive Concentration(mg/l)
Tetraethyl lead 600
Tetramethyl lead 5
Dichloroethane 210
Dibromomethane 190
The concentration of organic lead in a sample has been argued as a means to determine when the fuel was released into the subsurface. In 1982, the allowable levels of lead in leaded gasoline was 4.2 grams per gallon, which was changed by the Environmental Protection Agency (EPA) in 1984 to a maximum of 0.1 gram per gallon. This concentration applies to the average quarterly production from a refinery or "pool standard." The pool standard is based on the total grams of lead used by a refinery in a given time period divided by the total amount of gasoline manufactured by the refinery in that time frame. As a result, individual batches of gasoline can contain 4.2 grams per gallon per EPA requirements and 0.8 grams per gallon in California. It has been argued that these guidelines can be used to predict the time frame during which the product was manufactured. A challenge to this argument is that the lead content of an individual sample is not conclusive because the lead content for any given time frame is based on the pool standard. The consequence of this practice is that individual gasoline samples can vary from batch to batch and therefore cannot be used to date the year of manufacture.

A recent technique forwarded to age-date is the examination of lead isotope ratios in gasoline-impacted soil or water samples. It is based on the premise that the use of one particular source of lead with distinctive high isotopic ratios increased progressively from the late 1960s to the late 1980s. High-precision lead isotope ratio analysis is used to calibrate these changes as a function of time. It is reported that this technique allows one to established the time of formulation to within one to five years. Examples of the degree of resolution for different time periods using this analysis is: one to two years for 1969-1982, two to three years for 1982-1990, and three to five years for 1955-1968. [9]

Other antiknock ingredients used in fuels include iron-based compounds. The most well-known metal compound is iron pentacarbonyl, which was marketed in the 1930s. Another is the manganese additive, methyl cyclopentadienyl manganese tricarbonyl (MMT), which was introduced in the United States in 1957 as an antiknock and lead alkyl supplement and later commercialized as a supplement to tetraethyl lead[10]. Although MMT can be age- diagnostic, its absence in gasoline does not necessarily indicate a basis for age-dating because it was not routinely added by all manufacturers. It is currently an additive in Canadian gasolines.

Oxygenates are added to gasoline to increase the oxygen content and reduce carbon monoxide emissions. Oxygenated compounds used for this purpose include methanol, methyl tertiary butyl ether (MTBE), ethanol and tertiary butyl alcohol, tertiary methyl ether (TAME), and ethyl tertiary butyl ether (ETBE). MTBE is used to improve the octane rating and/or improve the tolerance for moisture in gasoline. MTBE has been an additive in gasoline since about 1979; current unleaded gasolines contain as much as 15 percent by volume. MTBE was introduced into East Coast, Gulf and Midwest gasolines after 1980 and into West Coast gasolines after 1990 [11].

The changing formulation of fuels over time also provides opportunities for dating a release. Prior to 1975, diesel was primarily straight-chained, while post-1975 diesel was thermally cracked. This distinction can be determined analytical with a small volume of product, thereby providing a time frame for when the product was available.

Trace Metals

Trace metal analysis may be useful in the confirmation of waste oil, which is more likely to contain lead, zinc, chromium, copper, or aluminum from the abrasion of an engine than from a fuel.. Metals can also be additives to a hydrocarbon fuel; the addition of barium and zinc to motor oil is an example. Vanadium and nickel are also associated with fuels; concentrations can range from 100 to 400 parts per million. While not a trace metal, boron was a common gasoline additive in use from 1956 to 1981 and, if detected with gasoline, may provide an indicator of a pre-1981 formulation. Another example is the addition of borates to gasoline in the 1960s by ARCO.

Chemical Profiling

Chemical profiling is used to identify the individual components of a soil or liquid sample. This information is then used to estimate how long a chemical has been in the subsurface. The presence of gaseous hydrocarbons, isobutane, n-butane, iso-pentane, and n-pentane in a sample contaminated with gasoline can be used to determine whether a gasoline released into the subsurface is "fresh" or "weathered." Fresh gasolines normally contain n-hexane and n-heptane in higher concentrations than methylcyclohexane (MCH) and n-octane. After the gasoline weathers in the subsurface, the MCH concentration increases relative to n-hexane and n-heptane and carbon seven (C7) normal paraffins. The difficulty with this analysis and testimony is that "fresh" and "weathered" are relative terms that allow a wide range of interpretations by forensic geochemists.

Chemical Degradation Models

Chemical degradation models are used to estimate the length of time that a hydrocarbon or chlorinated solvent has been in the subsurface as a means for developing allocation models in environmental insurance cases. The following are commonly used degradation models used.


The degradation of specific petroleum fractions in a fuel has been proposed as a means to age-date a hydrocarbon [12]. One approach relies on the volatilization and dissolution of BTEX (benzene, toluene, xylene, and ethylbenzene) compounds in groundwater as an qualitative indicator of how long the product has been in the subsurface. The normal sequence of loss of these compounds in groundwater is that benzene diffuses most rapidly out of free- phase gasoline and partitions into groundwater followed by toluene, ethylbenzene, and xylene. The reverse occurs with BTEX in soils. Toluene, ethylbenzene and xylenes are preferentially retained by soil relative to benzene; ethylbenzene and xylenes are also more resistant to degradation than benzene or toluene. Calculating ratios between these four compounds has been argued in environmental insurance cases as a means to identify the relative age of a hydrocarbon release.

In BTEX impacted soils, the concentration ratio of benzene plus toluene to ethylbenzene plus xylene changes from about 0.8 in the original fuel to about 0.4 in five years. For BTEX-impacted groundwater, if the ratio between benzene plus toluene to ethylbenzene plus xylene is between 1.5 to 6.0 near a suspected source, the release probably occurred within the last five years. This ratio decreases exponentially with time because of the preferential transport of benzene and toluene, which increases the less soluble ethylbenzene and xylene concentrations. The degradation of benzene and toluene with time also results in a reduction of the BTEX ratio.

Significant uncertainty is inherent using these ratios for age-dating due to the variation of BTEX compounds in the original gasoline as well as environmental alterations to the product. For example, under anaerobic conditions, toluene may be more rapidly degraded than benzene. These uncertainties can result in a wide range of ratios for identically aged spills, especially in different soils. A knowledge of the range of BTEX in fresh and weathered gasolines illustrates the uncertainties associated with such ratio analyses. Exhibit 4 summarizes BTEX ranges in different grades of dispensed gasoline and ranges of BTEX hydrocarbons in free product, water, and soils. [13]

Exhibit 4. BTEX Ranges (mg/ml) in Gasolines
Gasoline Grade Benzene Toluene Ethylbenzene Xylene
Leaded Gasoline 6.6-14.8 18.6-64.4 6.2-14 32.1-77.4
Regular Unleaded 5.0-20.0 17.9-44.3 5.8-12.0 27.1-48.6
Unleaded Plus 7.1-17.3 23.9-42.6 7.7-10.0 37.5-50.5
Super Unleaded 6.6-23.0 22.4-81.0 6.6-16.0 33.4-65.8
BTEX Ranges in Environmentally Altered Gasolines
Gasoline Grade Benzene Toluene Ethyl benzene Xylene
Free Product(mg/ml) 0.16-24.0 0.39-100 2.1-29 9.1-98
Water (g/ml) 0.02-30.3 0.002-38.3 0.01-5.8 0.005-29.6
Soil (g/g) 0.01-10 0.01-77.4 0.02-50.9 0.01-220
A more quantitative approach is reliance on the biodegradation half-life of hydrocarbon compounds in the soil or groundwater. The estimated half-life is the time required for one half of the compound to biodegrade. Exhibit 5 summarizes the biodegradation rates of selected BTEX and polynuclear aromatic hydrocarbons (PAHs) compounds measured at 25 degrees Celsius:

Exhibit 5. Biodegradation Half-Lifes for Selected Hydrocarbons

Compound Biodegradation Half-Life (hours)
Compound Soil Groundwater
Benzene 120-384 240-17,280
Toluene 96-528 168-672
Ethylbenzene 72-240 144-5472
o, m, p-xylene 168-672 336-8640
Acenaphthene 299-2,448 590-4896
Anthracene 1,200-11,040 2,400-22,080
Benzo(a)pyrene 1,368-12,720 2,736-25,440
Chrysene 8,904-24,000 17,808-48,000
Fluoranthene 3,360-10,560 6,720-21,120
Fluorene 768-1,440 1536-2880
Naphthalene 398-1,152 24-6192
Phenathrene 384-4,800 768-9600
Pyrene 5,040-45,600 10,080-91,200

The degradation of diesel in the subsurface over time can also provide a qualitative indicator as to whether the diesel is "fresh" or "aged". For example, examination of chromatograms from fresh and aged diesel samples shows that n-alkane compound peaks associated with fresh diesel are diminished or non-existent in weathered diesel. Similarily, while the n-alkane compounds in diesel readily degrade, other compounds such as pristane and phytane remain in weathered samples. Consequently, the concentration of these compounds increases as the diesel ages.

Chlorinated Solvents

The presence of chlorinated solvents and their breakdown products has been proposed as a means of identifying how long a chemical has been in the subsurface. This argument is usually based on laboratory-measured degradation rates between chlorinated solvents, primarily for the parent compounds tetrachloroethene (PCE), trichloroethane (TCA), and carbon tetrachloride (PCM) The presence or absence of a particular breakdown product is argued as evidence that the parent compound was present for a particular period of time. For example, the compound 1,1-dichloroethene is a breakdown product of both TCA and PCE, while chloroethane is a degradation product of only TCA or 1,2-dichloroethane. The presence of chloroform can indicate the presence of carbon tetrachloride; it is not an associated product of either PCE or TCA. A variation to this dating approach is that if the original concentration of a chlorinated solvent introduced into the subsurface is known along with measured values in the soil or groundwater, a knowledge of the half-life for that chemical can provide a basis for estimating when the chemical was released into the subsurface.

Several arguments to these approaches are available. One is that site-specific information is rarely available to compare to laboratory-derived degradation rates. A wide range of half-lives are also reported in the published literature. For example, published half-lives for PCE at 100 to 250 C range from 0.7 to 1.3 X 106 years. [14] For 1,1,1-trichloroethane, published half-lives at 100 to 250 C range from 1.1 to 12 years. [15]. Furthermore, the heterogeneity of the physical and chemical systems at a site introduces tremendous uncertainty into degradation rates. A simple example is whether site-specific information is available to determine whether soils are anaerobic or aerobic; this determination is significant as to whether the chlorinated solvent can be degraded.

The presence of these compounds in the subsurface can also inhibit the microbial degradation that may otherwise occur. For example, the anaerobic degradation of carbon tetrachloride is reported to be inhibited at concentrations between 80 and 250 g/L; for TCE anaerobic inhibition was observed at concentrations of 150 g/L. These types of complications introduce further uncertainty in the use of degradation rates as an indicator of the timing of a release. [16]

Impurities in the original product can similarily introduce uncertainty with the use of degradation rates for chlorinated solvents as an age-dating technique. It is not uncommon for industrial-grade PCE to contain impurities of one to three percent which can include the degradation products of PCE. Degradation can also occur prior to a release into the subsurface (e.g., degradation within a clarifier or sewer pipe).

Groundwater Modeling

Groundwater modeling is a frequently used technique to establish the timing and origin of a release. Reverse modeling is a type of modeling used specifically for this purpose and is commonly encountered in insurance cases.

Reverse Modeling

Reverse modeling uses the direction and velocity of groundwater to predict, in reverse, when and where a contaminant entered a groundwater system. This technique is called "reverse" or "backward extrapolation groundwater modeling"[17]. In its simplest application, the observed length of a contaminant plume and a representative groundwater velocity are used to estimate the timing of the release.

Reverse modeling relies on various hydrogeologic and contaminant characteristics to mathematically simulate the origin and timing of a release. Hydrogeologic variables used in reverse modeling include the groundwater gradient, soil porosity, velocity, and horizontal and transverse dispersivity. The reliability of these values is due in part to whether they are measured in the field or laboratory or are obtained from published numbers. For example, it is generally recognized that the most representative measurements for determining the hydraulic conductivity of a formation are from a pump test, while less reliable values are associated with slug tests, sieve analyses, and laboratory measurements of soil cores. Given that variables such as the gradient of the groundwater table and changes in soil porosity differ significantly with soil texture, a scientifically defensible approach is to use a range of values for each parameter and to develop a corresponding range in the groundwater velocity.

Dispersivity is a hydrogeologic characteristic used in reverse modeling. It is the three- dimensional spreading of a contaminant plume in groundwater and is primarily dependent on the aquifer characteristics and the geometry of the contaminant source area. Dispersivity is a fitted parameter that allows considerable latitude in calibrating or adjusting a reverse model to fit a prescribed result. Dispersivity values are used in both a longitudinal and transverse direction relative to the center of mass of a contaminant plume. In reverse modeling, it is used to fit the shape of the plume to the observed plume at one point in time. Longitudinal dispersivity values used with solute transport models are commonly in the range of 90 to 300 feet, while horizontal dispersivity values can be as much as 150 feet. There is little physical evidence for using such large numbers other than the fact that groundwater models are able to simulate contaminant concentrations that compare favorably with observed values in monitoring wells. In cases where no data exist to estimate dispersivities, EPA recommends multiplying the length of the plume by 0.1 to estimate the horizontal dispersivity [18].

Contaminant properties used in modeling affect the modeled transport of a contaminant and include variables such contaminant density and viscosity, retardation, and biodegradation. The retardation of a selected chemical in groundwater is an example of a fitted parameter that is commonly used in litigation where reverse modeling is employed [19]. If groundwater flow is six feet per day and a selected retardation coefficient for PCE (perchloroethylene) is two, then PCE is assumed to be transported at a rate of three feet per day. Given that published values for the retardation of PCE in sand and gravel aquifers is between one (no retardation) and five,[20] one can vary the retardation rate to fit the observed detection of PCE at a monitoring well to fit a selected time of release. While retardation values for TCE (trichloroethylene) are reported as less than ten and usually between one and two and a half [21] these values can similarily be adjusted to correspond to a prescribed time of release of TCE.

An argument against the use of reverse groundwater models is that the direction of groundwater flow and velocity at time periods prior to the first measurement are unknown. Similarily, the rate of transport of a contaminant as compared to groundwater is unknown with certainty. These factors introduce significant uncertainty and variability in terms of when the contaminant entered groundwater. As parameters such as the retardation of a chemical and the dispersivity of the groundwater are usually not measured but rather fitted parameters, these values can significantly bias the modeling results. A successful challenge to reverse modeling usually is based on on the representativeness of these hydrogeologic and chemical parameters and the lack of information to verify the accuracy of the results given that the data are not available at this earlier point in time.

A Case Study

Selected forensic techniques in identifying the origin and date of release for allocation purposes was used in T. H. Agriculture & Nutrition Co. Inc., v. Aceto Chem. Co., Inc., et al.[20]. The case dealt with a pesticide manufacturing facility located in Fresno, California. A number of parties had formulated a variety of pesticides, herbicides, fungicides and defoliants from 1951 to 1982. Building materials, soil, and groundwater were contaminated with numerous types of chemicals throughout site. Allocation of discrete areas of contamination to responsible parties was the focus of the technical case.

Techniques used for identifying the origin and timing of releases by both parties included aerial photography interpretation, determining when a chemical was first commercially available, association of chemicals with a formulation process, chemical degradation models, and reverse groundwater modeling.

Aerial photographic interpretation was key in identifying certain activities and areas of contamination with specific time periods. For example, the presence of several landfills, numerous above ground storage tanks, buildings, railroad spur lines, cisterns, and a concrete sump associated with soil and groundwater contamination provided a means to associate these features and subsurface contamination with a particular time period. The location of specific activities, such as the preparation of defoliants, was confirmed by aerial photography by identifying the associated equipment used in these formulations in a particular area.

Hundreds of chemicals were used at the site. An analysis of when these chemicals first became commercially available in the United States provided the basis for identifying the earliest time that the chemical could have been present. For example, Diphenamid and Bromacil were available after 1960 and 1963, respectively, thereby providing the basis for the association of these chemicals with a particular operator. Other chemicals, such as DDT (available in 1942), were used throughout the operational history.

Examination of degradation models and the resulting by-products were examined to identify patterns that could be associated with a particular feature or chemical that was known to be used by a particular owner. For example, chloroform was detected in groundwater about one mile downgradient of the site and was a chemical of concern. Chloroform was allegedly used as a laboratory solvent and discharged into a brick cistern. Chloroform is also a degradation product by hydrolysis and bacteria of carbon tetrachloride; analysis was therefore performed to distinguish whether the chloroform detected in groundwater was associated with carbon tetrachloride degradation. This issue was also relevant because chemicals such as DDT, DDE, toxaphene, dieldrin, and other pesticides preferentially dissolve into chloroform. This phenomena is known as cosolvation. The chloroform thereby transports these chemicals to depth in the soil column or into the groundwater; these chemicals otherwise are strongly sorbed by soil and remain shallowly distributed in the soil.

Cosolvation was similarily argued to be responsible for the distribution of DDT and dieldrin at depth in soil near a railroad loading dock due to the release of xylene onto contaminated soil. The volume of DDT contaminated soil at depth, resulting from the release of xylene and subsequent cosolvation of the DDT, depth was argued to be allocated to the operator responsible for the xylene spillage.

Another technique used was association of a chemical or group of chemicals with a specific activity. For example, the chemical ethylene dichloride (1,2-DCA) was a chemical detected in downgradient monitoring wells. It is a lead scavaneger in unleaded gasoline as well as a component in food fumigants. The detection of 1,2-DCA in samples also contaminated with hydrocarbons or fumigants provided the basis for associating the ethylene dichloride to a particular activity and time period.

Reverse modeling was employed to link groundwater contamination with particular source areas, such as cisterns and drainage lines, as well as a time period. An assumption of the model is that the direction of groundwater flow and velocity has remained generally unchanged for approximately 45 years. Challenges to the approach included the inability to model groundwater for a 30 year period for which no groundwater velocity values were available. Changes in the direction and velocity of groundwater flow over time due to the influences of a nearby irrigation canal and a high capacity irrigation well introduced uncertainity. The shape of the chloroform plume was also more consistent with a known source of chloroform from a cistern than with leakage from a sewer line which would produce a more diffused plume shape. The time required for the chloroform to reach the groundwater after discharge into the cistern was also unknown thereby introducing additional uncertainty regarding the accuracy of the model.

When the testimonial and historical information was combined with the results from these dating techniques, the ability to allocate many of the contaminated buildings and soil to a particular time period was possible.


Absent direct testimony or other conclusive evidence, various forensic techniques are available to identify the timing and origin of a contaminant release. When used in concert with other evidence, they can provide convincing information regarding the origin and timing of a release that forms the basis for allocation. Techniques such as reverse groundwater modeling and the use of BTEX ratios should be scrutinized to determine whether they are scientifically defensible.


1. IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, Halogenated Hydrocarbons, Volume 20, International Agency for Research on Cancer, October 1979, Switzerland.

United States Environmental Protection Agency, 1971, EPA Compendium of Registered Pesticides, Volume III, Washington, D.C., United States Government Printing Office.

Groundwater Chemicals Field Guide., 1995. J. Montgomery. Lewis Publishers, Chelsea, MI.

2. J., Blackford, J., 1975, "Trichloroethylene," Chemical Economics Handbook, Stanford Research Institute, Menlo Park, CA, 697.301A-697.302Y.

3. Monsanto, Polychlorinated Biphenyls, A report on Uses, Environmental and Health Effects and Disposal, White Paper, p. 18.

4. I. Kaplan, and Y. Galperin, 1996, How To Recognize a Hydrocarbon Fuel in the Environment and Estimate Its Age of Release. Groundwater and Soil Contamination: Technical Preparation and Litigation Management, Thomas Bois and Bernard Luther, eds. John Wiley & Sons Inc, New York, pp. 145-200.

5. L. Lee et al., 1992, Partitioning of Polycyclic Aromatic Hydrocarbons from Diesel Fuel into Water, Environmental Science and Technology, 26:2104-2110.

Global Geochemistry Corp., 1991, Characterizing petroleum contaminants in soil and water and determining source of pollutants, p. 37.

6. R. Rhue et al., 1992, "The Fate and Behavior of Lead Alkyls in the Environment: A Review," Critical Reviews in Environmental Control. 22(3/4): 169-193.

7. P. Cline et al., 1991, "Partitioning of Aromatic Constituents into Water from Gasoline and other Complex Solvent Mixtures." Environmental Science and Technology, 25:914-920.

8. L. Gibbs, 1993, "How Gasoline has Changed," SAE Technical Paper Series (Fuels and Lubricants Meeting and Exposition, Philadelphia, PA, (October 18-21, 1993).

9. R. Hurst et al., 1996, "The Lead Fingerprints of Gasoline Contamination," Environmental Science and Technology,Vol. 30, No. 6., p. 304-307.

10. T. Russell, 1988, "Petrol and Diesel Additives". Petroleum Review, (October 1988), The Institute of Petroleum, pp. 35-42.

J. Faggan et al., 1975, An evaluation of manganese as an antiknock in unleaded gasoline. SAE Automobile Engineering Meeting, Detroit, MI, (October 13-17, 1975), p. 21.

11. J. Walker et al., 1976. Biodegradation Rates of Components of Petroleum. Canadian Journal of Microbiology, 22:1209-1213.
12. R., Raymond et al., 1976. Oil Degradation in Soil. Applied and Environmental Microbiology, 31(4) 522-535.

I. Kaplan et al., 1995, Pattern of Chemical Changes in Fugitive Hydrocarbon Fuels in the Environment. Society of Petroleum Engineers, SPE 29754, SPE/EPA Exploration & Production Environmental Conference, Houston TX (March 27-29, 1995).

13. W., Haag and T. Mill, 1988, "Effect of Subsurface Sediment on Hydrolysis of Haloalkanes and Epoxides", Environmental Science and Technology, 22:658-663.

14. P. Jeffers et al., 1989, "Homogeneous Hydrolysis Rate Constants for Selected Chlorinated Methanes, Ethanes, Ethenes, and Propanes". Environmental Science and Technology. 23(8):965-969.

P. Cline and J. Delfino, 1989, "Transformation Kinetics of 1,1,1-trichloroethane to the Stable Product 1,1-dichloroethene," Biohazards of drinking water treatment. (R.A. Larson, ed.) Chelsea, MI: Lewis Publishers, Inc., pp. 47-56.

15. W. Dilling et al., 1975, "Evaporation Rates and Reactivities of Methylene Chloride, Chloroform, 1,1,1, Trichloroethane, Trichloroethylene, Tetrachloroethylene, and other Chlorinated Compounds in Dilute Aqueous Solutions". Environmental Science and Technology, 9:833-838.

P. Howard, et al., Handbook of Environmental Degradation Rates. Lewis Publishers, Chelsea, MI.

16. M. Alexander, 1985, Biodegradation of Organic Chemicals. Enviornmental Science and Technology, 18:106-111.

17. A. Kezsbom and A. Goldman, 1991, "The Boundaries of Groundwater Modeling under the Law: Standards for Excluding Speculative Expert Testimony," Environmental Claims Journal, Vol 4, No. 1 (Autumn 1991).

I. Kornfeld, 1992. "Comment to the Boundaries of Groundwater Modeling under the Law: Standards for Excluding Speculative Expert Testimony," Tort and Insurance Law Journal, Vol. 28, No. 1. (Fall 1992).

R. Morrison and R. Erickson, 1995, Chapter 7, "Groundwater Investigations". Environmental Reports and Remediation Plans: Forensic and Legal Review. John Wiley & Sons, Inc., New York, p. 155.

18. J. Wilson, et al., 1981, "Transport and Fate of Selected Organic Pollutants in a Sandy Soil", Journal of Environmental Quality, 10:501-506.

A. Lallemand-Barres and P. Peaudecerf, 1978, Recherche de reltions entre les valeurs mesurees de la dispersivite macroscopique d'un milieu aquifere, ses autres caracteristiques et les conditions de mesure. Etude bibliographique. Bull. Bur. Rech. Geol. Min (BRGM), Ser. 2, Sec. III, No. 4, pp. 277-284.

19. U.S. Environmental Protection Agency, 1985, Water Quality Assessment: A screening Procedure for Toxic and Conventional Pollutants in Surface and Ground-water, Parts I and II (revised 1985), EPA/600/6-85-002a (Part I, 609 p.), EPA/600/6-85/002b (Part II, 444 pp) Environmental Research Laboratory, Athens, GA.

Carrier Corp. v. Detrex Corp., No. C-703625, Cal. Super. Ct. Los Angeles, County, (1996).

20. T. H. Agriculture & Nutrition Co. Inc., v. Aceto Chem. Co., Inc., et al., No. CV-F-93-5404 OWW/DLB, E.D. Cal., (1996).

R., Schwarzenbach, et al., "Behavior of Organic Compounds during Infiltration of River Water to Groundwater. Field Studies." Environmental Science and Technology, 17, (1983), p. 472-479.

21. L. Barber, et al., "Long-term Fate of Organic Micropollutants in Sewage Contaminated Groundwater, " Environmental Science and Technology, 22 (1988), 205-211.

D., Mackay, "Characterization of the Distribution and Behavior of Contaminants in the Subsurface", Ground Water and Soil Contamination Remediation: Toward Compatible Science, Policy, and Public Perception, Washington, DC, National Academy Press, (1990), p. 70-90.

M. Mehran, et al., "Distribution Coefficient of Trichloroethylene in Soil-Water Systems," Ground Water, 25, (1987) 275-282.

If you have any questions contact Dr. Morrison or E-Mail at:
R Morrison & Associates [now part of DPRA]
Soil Physicists, Expert Witnesses, Forensic Analyses
Robert D. Morrison, Ph.D.
100 E. San Marcos Blvd.
Suite 308
San Marcos, CA 92069
760-752-8342 voice
760-752-8377 fax
Website:  http://www.dpra.com

Return To Equipment PagesReturn To Home PageRequest More Information From EDP