Technical Issues in Litigation Involving Chlorinated Solvent Releases

Susan Tighe Litherland, PE
Robert I. Chapin
Weston Solutions, Inc.

Cases involving chlorinated solvents pose unique technical challenges during litigation. The typical questions asked of an expert witness in most contamination cases include: what material was released, what and where was the source of the release, how much was released, when was it released, and how was it released. Due to the ways in which chlorinated solvents have been used, the chemical nature of the chlorinated solvents, and their behavior in the environment, these questions may be difficult to answer, even in situations where a large amount of environmental information has been collected.

Background Information

An overview of the chemistry of chlorinated solvents and their behavior in the environment may be helpful to understand some of the technical difficulties associated with cases involving releases of these chemicals.

Chlorinated solvents include a class of compounds with short carbon chains and one or more chlorine atoms. A variety of chlorinated solvents have historically been used for a number of different commercial and industrial purposes, including the dry cleaning of clothes, degreasing and cleaning of parts and machinery, and as a feed material for the formation of other chemicals. The two chlorinated solvents that appear to be the most prevalent in environmental litigation are tetrachloroethene and trichloroethene. This is due to their widespread use and their persistence in the environment.
Tetrachloroethene is also known as: tetrachloroethlyene, perchloroethylene, PCE or "perc." All of these names refer to the same chemical; an ethene (2-carbon) molecule with four chlorine atoms. PCE is the most common dry cleaning fluid used today, and has (and is) also used as an industrial degreaser and cleaner. Trichloroethene is also known as trichloroethylene or TCE. TCE is an ethene molecule with three chlorine atoms attached and has (and is) used primarily for industrial degreasing and parts cleaning. For the sake of clarity and focus, this article uses the term chlorinated solvent to specifically refer to PCE and TCE since these are the chlorinated solvents most commonly encountered in environmental litigation.

Density. Chlorinated solvents are heavier than water and petroleum products. For comparison, a gallon of water weighs about 8.3 pounds, a gallon of gasoline weights about 6.2 pounds, a gallon of PCE weighs about 13.5 pounds, and a gallon of TCE weighs about 12.0 pounds. Unlike common petroleum products that will float on the water's surface, chlorinated solvents like PCE and TCE will sink in water until they reach a natural or artificial barrier. Due to the density and low solubility, these chlorinated solvents are also referred to as Dense Non-Aqueous Phase Liquids (DNAPL). As will be discussed later, the presence or absence of DNAPL is also a significant area of discussion in litigation matters involving chlorinated solvents.

Solubility. PCE and TCE are sometimes listed from a chemical point of view as "non-soluble" because they do not readily dissolve in water. However, from an environmental standpoint, sufficient chlorinated solvent will dissolve in groundwater to create a potential environmental problem. The solubility of PCE in water is around 200 milligrams per liter (mg/L), and the solubility of TCE in water is around 1,100 mg/L. For liquids, mg/L is essentially the same as parts per million (or ppm). For comparison, the drinking water standard for both chemicals is 0.005 mg/L. This value is the same as 5 micrograms per liter (ug/L) or 5 parts per billion (ppb). To put this concentration in perspective, a nice-sized residential swimming pool contains approximately 10,000 gallons of water. To mix up a batch of 5 ppb PCE solution, you would add slightly less than 190 milligrams, or less than 5 drops of the solvent to the swimming pool.

Viscosity. Chlorinated solvents are less viscous than water. Viscosity is an indication of the resistance of a fluid to flow under certain conditions. At ambient pressures and temperatures, the viscosity of water is about 1.0 cP (centipoise), the viscosity of gasoline is about 0.5 cP, the viscosity of PCE is about 0.8 cP, and the viscosity of TCE is about 0.5 cP. Fluids with high densities and low viscosities, like chlorinated solvents, will flow more quickly than water through a porous media. As an example, if a glass of water, a glass of gasoline, and a glass of TCE were poured out on a surface such as concrete, the gasoline and TCE will move faster through the concrete, resulting in a greater mass of the liquid reaching the soils below the concrete. If enough liquid was released it would move completely through the soil to the groundwater. When the gasoline reached the groundwater table it would float and spread on the water's surface. In the case of PCE and other chlorinated solvents (DNAPLs), the groundwater does not act as a barrier to further vertical migration. The PCE (or other chlorinated solvent) may be temporarily retarded at the groundwater table, but would then continue migrating downward until the quantity was exhausted, or the PCE reached a layer that was relatively impermeable to the PCE.

Volatility. Chlorinated solvents are considered volatile organic compounds, or VOCs, indicating they have relatively high vapor pressures at ambient temperatures. As the vapor pressure of a substance increases, the amount of material that passes from a liquid or solid phase to a gas phase increases. Compounds with vapor pressures that exceed 1 millimeters of mercury (mmHG) can be found in relatively high concentrations in the vapor phase in the vicinity of the liquid chemical. The vapor pressures of PCE and TCE are 14 mmHG at 20ºC, and 57.8 mmHG, respectively. The volatile nature of the chlorinated solvents will result in elevated concentrations of the solvent in the soil vapors. In some cases these vapors can be used to locate sources of chlorinated solvents, but also can migrate in the soils and along subsurface utilities, in some cases a significant distance from the original source.

Partitioning. The partition coefficient measures the tendency of a chemical to be distributed between a solution and a solid phase. In the environment, the solid phase that "attracts" the chemical is usually organic materials, such as decayed plant residues that are incorporated into the soil. A high partition coefficient indicates a strong attraction to other organic-containing materials, including soils. In the case of chlorinated solvents that are released in the environment, partitioning results in a distribution between the soil and groundwater. Additionally, PCE and TCE will bind tighter to soil with higher natural organic content, e.g., clays and silts. The bond is less strong in sands, which typically have lower natural organic content. This partitioning results in a mass of the chlorinated solvent being "stuck" on the soil. The mass associated with the soil is typically much greater that the mass dissolved in the groundwater. The solvent that is associated with the soil continues to be released into the groundwater over a long period of time. The partitioning, combined with the DNAPL character of the chlorinated solvents, makes remediation of chlorinated solvent sites more difficult than the remediation of sites contaminated with other types of chemicals.

What Was Released?

This is typically the easiest question to answer with cases involving chlorinated solvents. Unlike petroleum products, which generally contain a number of different individual chemicals (i.e. benzene, toluene, ethylbenzene, etc.), chlorinated solvent products primarily contain just one chemical compound. For example, if PCE was being used at a facility, the material that was purchased and subsequently released to the environment was most likely pure PCE. Typical Material Safety Data Sheets for PCE list the percentage of PCE in the product as 100% or >99%. Although PCE may contain trace levels of TCE, the presence of PCE at concentrations in excess of TCE is generally a good indication that the released material was PCE. Although there are products which contain both PCE and TCE, from an environmental standpoint, the products which most frequently result in contamination with PCE and TCE are generally pure products. Based on this, sample results from soil and/or groundwater taken from the source area can generally be used to determine with a high degree of confidence what chlorinated solvent was released.

What Was the Source of the Release?

While the legal definition of a release is quite lengthy, in "non-regulatory" terms, a "release" has occurred when a chemical is no longer in the container that it was supposed to be in. For the purpose of this discussion, a release would include a spill (accidental or otherwise) to a porous surface or soil, or an escape of the liquid or solution from a sewer line or other subsurface structure. Although releases of chlorinated solvents to air occur, air issues are not being considered in this discussion.

The term "source of release" is often used in two different manners. First, as the general location or property where the release of the chemical occurred (e.g., "the source of the release was the dry cleaners on Main Street"), and second, as the specific location where the chemical escaped into the environment (e.g., "the source of the release was a crack in the sewer line leading from the dry cleaners on Main Street"). For the purpose of this discussion, we will use the term "source of the release" to refer to the specific location where the chemical escaped into the environment.

In virtually every situation, the highest concentrations of a chlorinated solvent will be found at the location of the release of the chlorinated solvent. At first glance, this would seem to make defining the source location relatively easy. Unfortunately, practical factors often complicate this issue.

Types of Samples. There are three general types of samples that can be used to investigate the presence of chemicals in the subsurface environment: soil, groundwater, and soil vapor. Because of the solubility and volatility of chlorinated solvents, groundwater and/or soil vapor data are generally the most efficient means to locate the source or sources of chlorinated solvents. With sufficient data, the source or sources of the chlorinated solvents can be highlighted by finding the highest concentrations of chlorinated solvents in the groundwater and/or soil vapor.

Issues with Soil Samples. The absence of chlorinated solvent concentrations reported in soil samples from potential source areas has frequently been used in an attempt to demonstrate that a certain area or property was not a source of identified groundwater contamination. Migration of a chlorinated solvent, in a liquid phase (DNAPL), results in a path of residual chlorinated solvent on the soils where it migrates. Even when the material released is a solution of the chlorinated solvent in water, due to partitioning, soil in the vicinity of the release and along the migration path would be expected to have elevated concentrations of the chlorinated solvent. Although the lack of measurable chlorinated solvent in soils is a useful fact in an attempt to show that a certain property was not a source, it cannot be used alone to eliminate a property or area as a source. Conversely, the presence of measurable solvent in soil samples is a good indicator of a source location. In practice, soil samples frequently do not provide accurate data regarding the source location. Frequently, we see non-detectable levels in the soils and cannot locate a soil source, even when the general source of the contamination is fairly obvious based on groundwater data. There are several reasons for this apparent inconsistency:

• The area of affected soil may be very small. For example, a spill of DNAPL on gravel or sand may not spread out very much. Finding a 4-square foot-size area where someone supposedly dumped a bucket of TCE within a one-half acre (20,000+ square feet) property may be extremely difficult.
  
  The source of the release may be under a building or other permanent structure. Releases that have migrated through the concrete floor, or through a joint in a sewer line, may not be particularly accessible to sampling equipment.
• Sample collection sometimes involves drilling or pushing a tool into the ground, which may create heat that can drive off any volatile chlorinated solvent that was there.
• Laboratory preparation of soil samples sometimes results in release of volatiles. Using traditional methods, a soil sample from the field is placed into a glass jar and placed in a cooler for shipment to the laboratory. The laboratory then opens the jar and obtains a portion of the sample for testing. This procedure can result in another opportunity for loss of volatiles from the soil sample. There are fewer opportunities for loss of volatiles from water samples.

DNAPL. Experiments have shown that even small changes in the permeability of a layer of soil can divert chlorinated solvent migrating as a DNAPL. Even below the water table, DNAPL will not be greatly influenced by the groundwater flow direction due to the differences in density of the two fluids, but will follow the sloping contour on top of the lower permeability layers. Slight variations in subsurface stratigraphy can influence the flow path in ways that are difficult to predict without a large amount of information. That said, in general, soil layers slope in the same direction as the groundwater flow. Because the DNAPL is being pulled down under the influence of gravity, it typically does not move significant distances upgradient from the source through the soils.

Multiple Sources and Release Mechanisms. There are typically multiple release events and release mechanisms at facilities which have used chlorinated solvents. In addition to accidental spills, releases often occur during transfer of the solvent from one container to another, through overflow or washing into a drain or sink, and from improper disposal of wastes. This type of pattern results in multiple "sources" associated with one facility and at times confusing environmental data. A release through a sanitary or storm sewer can result in a source which appears to be upgradient of the site of the chlorinated solvent usage. In addition to multiple release events and/or mechanisms from a facility, there are frequently also multiple facilities in a particular area. More than one dry cleaner in a strip shopping center or multiple dry cleaners in nearby shopping centers are not unusual.

How Much Was Released?

Theoretically, one gallon of TCE could result in contamination of close to 300 million gallons of groundwater at the drinking water limit of 0.005 mg/L. In the environment it would not be possible for one gallon of TCE to contaminate this much groundwater for a number of reasons including "losses" to the soils (due to adsorption and partitioning), and vapor (due to volatilization). As discussed above, it is difficult to obtain accurate information regarding the concentration of the solvent in the soils. In general, it is not possible to calculate accurately the volume of solvent that was released at a site. With sufficient information and assumptions regarding partitioning and other factors, it is sometimes possible to provide an estimation or range of the volume of solvent released for a particular site.

When Was it Released?

Practitioners often would like to tie the presence of chlorinated solvents to a release that occurred during a particular time period when their client was not involved in the operations. The time of the release can be one of the most difficult questions to answer for a number of reasons. Without some unusual information, it is difficult to ascertain exactly when a release began and when it ended. Some of the technical issues associated with evaluating the time of the release are discussed below:

Lack of Documented Release. Although use of a product at a facility can be documented through purchase records, inventories, and interviews with workers, documentation of releases are not always available. One hundred percent of the dry cleaners that we have ever investigated have had chemical evidence of a release (PCE in soil and/or groundwater samples); however, probably less than one third of these have had any documented spill. This discrepancy probably does not arise from the dry cleaner operators' selective memories, but more likely because the releases that caused the contamination were not in their minds "spills," or did not occur where they could be observed.

For example, even drips of PCE to concrete over even a relatively short period of time can result in PCE reaching the soil. Anecdotal reports from dry cleaner workers suggest that drips were almost unavoidable during some activities, such as moving loads of clothes from one machine to the other (for transfer-type operations), or changing the filters used to filter the PCE prior to reuse. One of the major sources of releases through the sanitary sewer lines resulted from overflows of water containing PCE and even pure PCE from the PCE/water separator. At a facility that used TCE for degreasing, the parts were rinsed into the sink after being cleaned with TCE. Even this apparently minor amount of TCE being released to the sanitary sewer line resulted in an area of groundwater containing low levels of TCE that was almost 800 feet long. These historically routine methods of operation would not have been considered a spill or release by the operators, but often did result in contamination of soil and groundwater.

"Fingerprinting" Difficulties. Fingerprinting of petroleum products can sometimes provide significant information regarding the time period when a release occurred. By looking at ratios of various chemicals within a petroleum product or crude oil, the source of the material can sometimes be ascertained. In addition, the relative degradation can be evaluated, and the age of the spill can sometimes be estimated fairly accurately. Additives (such as tetraethyl lead and methyl tert-butyl alcohol) have historically been used to enhance the performance of petroleum products. Determining the presence of these additives at the location of a petroleum product spill can provide useful information regarding the time frame of the release, because these chemicals were often added only over specific periods of time.

Conversely, chlorinated solvents are essentially pure products and typically do not contain additives or impurities that could be used to evaluate the period of a release. Very low concentrations of stabilizers, inhibitors, and binders are sometimes used in chlorinated solvents, but these are not usually identified in environmental samples and/or may not be specific to certain periods of production.

Degradation. The closest thing to "fingerprinting" for releases of pure chlorinated solvents is the presence and concentration of degradation products. The primary path of degradation of PCE and TCE in the environment is due to the subsequent removal of chlorines from the ethene molecule leading to the formation of TCE (if the source material was PCE); then to cis-1,2-DCE; then to vinyl chloride. Subsequent degradation can result in formation of ethene, carbon dioxide, and water.

Unlike petroleum products, which degrade fairly predictably and relatively rapidly in an aerobic (oxygenated) environment, the specific subsurface conditions can significantly affect the degradation rate of chlorinated solvents. PCE and TCE degrade more efficiently under anaerobic (low or no oxygen present) conditions. Most water table aquifers have some dissolved oxygen and do not present the ideal conditions for degradation of PCE and TCE. The rate of degradation will depend on a variety of factors that may change over time, including the redox potential of the aquifer, the amount of dissolved oxygen, the presence of anaerobic bacteria, and the amount of organic matter and nutrients that are available to those bacteria. For example, during a recent site assessment the authors investigated a release from dry cleaner that had commingled with a release of gasoline from an adjacent service station. The bacterial degradation of the organic material in the gasoline created anaerobic conditions and served as an alternate carbon source for the chlorinated solvent degrading bacteria. And, as a result the PCE was degraded in this area at a faster rate. The result was that the concentrations of the PCE were much lower in the immediate vicinity of the gasoline release than farther downgradient where less organic material and higher oxygen levels were documented.

Some information regarding the period of release can be gained from a review of the ratios of the degradation products to the original material. The presence or absence of a particular breakdown product can indicate that a particular parent product was present. Depending on the site-specific information, this evaluation of the parent product and degradation compounds can provide useful information regarding the age of the release. However, in many cases this information may result in qualitative conclusions such as "it looks fairly recent," or "it looks fairly old." In many cases, this is useful. In a case where an operator was in the middle of three or more operators and/or the time periods of interest are fairly short (or close together), the information may not be adequate to confirm or refute participation in the contamination of a certain party.

Therefore, estimating the age of a release based on the current concentrations and a constant rate of degradation is usually not a straightforward calculation. Under ideal laboratory conditions, the solvents may degrade fairly rapidly, but in the field we have observed releases that are in excess of 10 years old with few degradation products.

Distance from source. One of the most helpful pieces of information to assist in identifying the time period of a release is the distance from the release point. With this information, confirmation that there are not preferential pathways (such as utility lines), and information about the hydrogeology, some conclusions can generally be drawn. The time of the release can be estimated based upon the average flow rate of the groundwater from the approximate source area to the downgradient edge of the plume. However, the groundwater flow equations are based on the chemical and physical properties of water and do not account for specific contaminants. The outcome of this type of investigation will likely be an opinion that the releases had to have started so many years ago due to the distance from the site, or that releases could not have occurred so many years ago because contamination is not found at a certain distance. This type of evaluation, in addition to good groundwater quality information, requires information regarding the affected groundwater aquifer.

Conclusions

A number of technical issues can make it difficult to answer the basic questions in litigation involving chlorinated solvents. As with most technical issues, more data generally provide a higher degree of confidence in the conclusions that can be drawn. In addition to operational and environmental information, information regarding subsurface structures is critical in the evaluation. Rarely in these cases are the answers to the questions going to be straight forward or will "proof" of a certain position be found in the available information. More likely, the environmental practitioner will need to look at all of the pieces of the puzzle to provide opinions regarding the various issues.

Susan Litherland is a Professional Engineer and a vice-president in the Austin office of Weston Solutions, Inc. Robert Chapin is a senior hydrogeologist and Client Service Manager in the Austin office of Weston Solutions, Inc.