Dedham Water Co. v. Cumberland Farms, Inc., 689 F. Supp. 1223 (D. Mass. 1988)
July 14, 1988
CUMBERLAND FARMS, INC., Defendant.
United States District Court, D. Massachusetts.
Thomas F. Holt, Jr., DiCara, Selig, Sawyer & Holt, David Chaffin, Boston, Mass. *1224 (John R. Cope, Bracewell & Patterson, Washington, D.C., of counsel), for plaintiffs.
Allan Van Gestel, Goodwin, Procter & Hoar, Boston, Mass., Christopher Davis, for defendant.
TAURO, District Judge.
The plaintiff Dedham Water Company ("DWC") is a Massachusetts corporation that, until December 17, 1986, held a public franchise to supply drinking water to the towns of Dedham and Westwood. The plaintiff Dedham-Westwood Water District ("the District") was created by statute in July 1985. In December 1986, the District purchased the assets of DWC and, thereafter, has provided drinking water to the residents of Dedham and Westwood. Because their interests in this litigation are the same, both plaintiffs will be referred to in this opinion as DWC.
The defendant Cumberland Farms Inc. ("Cumberland") is a Delaware corporation engaged in the business of processing and selling dairy products and other food items.
In May 1979, two wells in DWC's White Lodge Well Field, numbered 3 and 4, were found to be contaminated with certain volatile organic compounds ("VOCs") primarily 1,1,1-trichloroethane (1,1,1-TCA). These wells were removed from service, and steps were taken to prevent contamination of the other two wells. Later that year, DWC retained engineering consultants to design a plant that would eliminate the VOC problem and also remove iron and manganese from the groundwater supply. The plant was completed and began operating in early 1987.
The central allegation of DWC's 15-count Consolidated Complaint is that chemical discharges from Cumberland's property caused the groundwater contamination in the White Lodge Well Field. Cumberland denies any culpability, and so the issue as to who caused the contamination of the White Lodge groundwater is joined.
The key issue in this case is causation. When a plaintiff alleges that chemicals have migrated underground from another site, the plaintiff must establish that the second site was in fact the source of the pollutants in question.
Section 107 of the Comprehensive Environmental Response, Compensation and Liability Act of 1980, 42 U.S.C. § 9607 (1982) ("CERCLA") explicitly requires proof of this causal connection. The statute imposes liability only for a release or threatened release of a hazardous substance "which causes the incurrence of response costs." 42 U.S.C. § 9607(a) (4) (emphasis supplied).
Proof of a causal link between a defendant's release and the plaintiff's response thus forms one of the "basic elements" of a plaintiff's prima facie case under CERCLA. United States v. Ottati & Goss, Inc., 630 F. Supp. 1361, 1401-02 (D.N.H.1985) (CERCLA plaintiff must show that "[t]he release or threatened release causes the incurrence of response costs").Accord: State of New York v. Shore Realty Corp., 759 F.2d *1225 1032, 1043 & n. 16 (2d Cir. 1985); Artesian Water Co. v. New Castle County, 659 F. Supp. 1269, 1278 (D.Del.1987); United States v. Bliss, 667 F. Supp. 1298, 1304 (E.D.Mo.1987); United States v. Tyson, 10 Chem. Waste Lit. Rep. 872, 888 (E.D.Pa. 1986) [available on WESTLAW, 1986 WL 9250]; United States v. South Carolina Recycling and Disposal, Inc., 653 F. Supp. 984, 992 (D.S.C.1984); United States v. Wade, 577 F. Supp. 1326, 1333 (E.D.Pa. 1983); United States v. Reilly Tar & Chemical Corp. 546 F. Supp. 1100, 1115 (D.Miss.1982).
DWC asserts that Cumberland is strictly liable under CERCLA as long as there has been a release from its site, without regard to causation. DWC's argument confuses the principle of strict liability for the effects of a release, with the antecedent question of whether the release has any effects at all. While the two questions may merge in some traditional CERCLA cases, in the instant case they remain distinct.
Most CERCLA cases have concerned the clean-up of a single hazardous waste dumpsite. See, e.g., Violet v. Picillo, 648 F. Supp. 1283 (D.R.I.1986); United States v. Ottati & Goss, Inc., supra; United States v. South Carolina Recycling and Disposal, Inc., supra; United States v. Wade, supra. In these "one-site" cases, plaintiffs had expended substantial amounts of money to clean up the sites where defendants had disposed of hazardous waste. CERCLA allowed plaintiffs to recover such response costs from the original waste generators.
In a one-site case, defendants who have dumped waste at the site are strictly liable under CERCLA for any necessary response costs. Violet v. Picillo, 648 F. Supp. 1283, 1290 (D.R.I.1986) (citing cases).
Some defendants in one-site cases have argued that the plaintiff must show that a particular defendant's waste caused the release in question, or escaped during that release. But courts hold that, because "scientific technique has not advanced to a point that the identity of the generator of a specific quantity of waste can be stated with certainty," the CERCLA plaintiff in a one-site case is not required to "fingerprint" releases as belonging to a particular defendant. United States v. Wade, 577 F. Supp. at 1332. Rather, "upon a showing that the defendant's wastes were delivered to a site," the burden shifts to the defendant, which "bears the risk that its wastes will become unidentifiable." United States v. Bliss, 667 F. Supp. 1298, 1310 (E.D.Mo. 1987). This is the nature of the "strict liability" that CERCLA imposes upon generators. See Violet v. Picillo, 648 F.Supp. at 1290-92; United States v. Wade, 577 F.Supp. at 1331-34; Developments in the *1226 LawToxic Waste Litigation, 99 Harv.L. Rev. 1458, 1520-24 (1986).
In the context of the traditional one-site case, then, "courts seem to be satisfied that if a generator used a disposal site, there is a substantial likelihood that its wastes were part of the release." Developments in the LawToxic Waste Litigation, 99 Harv.L.Rev. at 1524 (footnote omitted). When plaintiff must clean up a mess in one location, and defendant's waste is admittedly part of that mess, the basic causal connection is presumed.
But this is not a one-site case. Rather, this is a classic "two-site" case. DWC alleges that Cumberland released hazardous wastes from its truck maintenance facility, and that these wastes traveled underground to DWC's well field. Two-site cases raise a different causation question: whether the defendant's releases had any effect at all upon the plaintiff's site.
CERCLA's strict liability provisions do not come into play until this antecedent question has been resolved. As the court explained in Artesian Water Co. v. New Castle County, supra, a two-site case in which a water company sought to recover the costs of response to groundwater pollution in its well field:
CERCLA's strict liability scheme does not diminish the necessity of demonstrating a causal connection between a release or threatened release and the incurrence of costs by a section 107 plaintiff. Plaintiff must therefore show that it incurred costs as a result of the release or threatened release of hazardous substances from the [defendant's] site.
659 F. Supp. at 1282 (footnote and citations omitted). Accord: State of New York v. Shore Realty Corp., 759 F.2d 1032, 1044 n. 17 (2d Cir.1985) (even though § 9607 imposes strict liability, plaintiff must still prove causation); State of Idaho v. Bunker Hill Co., 635 F. Supp. 665, 674 (D.Idaho 1986) ("The plaintiff has argued that since there is admittedly strict liability under the statute, there is no need for causation. However, strict liability does not abrogate the necessity of showing causation, but merely displaces any necessity for showing some degree of culpability by the actor. In other words, under strict liability the mental state of the defendant is irrelevant, but the damage for which recovery is sought must still be causally linked to the act of the defendant.") (emphasis added); 2 Law of Hazardous Waste: Management, Cleanup, Liability, and Litigation § 14.01 [a] (S.Cooke ed. 1988) (principle of strict liability becomes relevant only "[a]fter identifying a responsible party and a release that causes the incurrence of response costs as the two major prerequisites for a cost recovery suit"). Cf. id. at § 14.01[d] ("There are potential statutory and equitable problems with holding defendants associated with one site liable for an independent set of environmental harms that by coincidence occurred nearby.")
DWC's causes of action under Massachusetts statutory and common law likewise *1227 require proof of a causal nexus between defendant's conduct and plaintiffs' injury. Consequently, the court's factual findings will be dispositive on these issues as well.
White Lodge Well Field is located in Westwood, Massachusetts. Cumberland has its corporate office, a dairy processing plant, and a truck depot in the neighboring town of Canton. Between the Cumberland facility and White Lodge are the Neponset river, a railroad line, a sewer line, and some undeveloped property. Also in the area is an industrial park whose several tenants include the Shield Packaging Company ("Shield").
White Lodge Well Field is one source from which DWC provides water for the towns of Dedham and Westwood. The four White Lodge wells (WL-1, WL-2, WL-3 and WL-4) are located west of the Neponset River. The southernmost well is WL-3. WL-4, WL-2, and WL-1 lie to the north in that order. All four wells are contained within the University Avenue Industrial Park. East of the wells is a railroad line running south out of Boston. Also in close proximity to these wells are Shield, the New Neponset Valley Sewer ("NNVS") and the Westwood Extension Sewer ("WES").
Of the four White Lodge wells, WL-3 was the most contaminated. It was installed in 1962, WL-4 in 1967, WL-2 in 1958 and WL-1 in 1954. The wells are gravel packed with a depth of about 70 feet. The average pumping volume from the four wells during the 1970's was about 2.4 million gallons per day. The maximum yield from these wells, without depleting the aquifer, would be approximately 3.0 to 3.5 million gallons per day. In the 1970's, WL-3 produced an average of one-half million gallons per day.
The WES was installed just south of the well field in 1960. WL-3 was drilled in 1962, approximately 150 feet north of this municipal sewer and about 700 feet west of the NNVS, which had been installed in the 1930's. The University Avenue Industrial Park, together with the roads and sewers servicing it, was built during the 1960s.
Since 1960, DWC has experienced some water quality problems due to naturally occurring inorganic materials, iron and manganese. These problems included discolored water and sludge accumulating in pipes.
In March 1979, DWC tested the White Lodge wells for VOC contamination. Elevated levels of several VOCs, predominantly 1,1,1-TCA, were found in WL-3 and WL-4. VOC concentrations in excess of 1,000 parts per billion were found between 1979 and 1982 in these wells. But, VOC concentrations in WL-3 and WL-4 have steadily declined since 1983 to the point where they have been at or near federal drinking water standards.
On the Cumberland site in Canton is a well approximately 1,500 feet southeast of WL-3. Between the Cumberland facility *1228 and WL-3 is the NNVS, the Amtrak railroad tracks, land owned by the Canton Conservation Commission, and the Neponset River. Corporate activities at the Cumberland site include the processing and packaging of dairy products and the maintenance of vehicles that service Cumberland's retail outlets. The dairy plant and the production well were constructed in 1962. The garage facility was built in 1966, and the office/warehouse building was completed in the mid 1970s.
The truck garage had the capacity to service a fleet of 100 trucks and 40 cars. Everything from oil changes to engine rebuilding was done there. By 1982 the garage employed approximately 30 mechanics. In addition to the maintenance areas, the garage had a paint shop, a grease pit, a refrigeration maintenance area, a body shop and an outdoor truck wash.
For most of the period between 1962 and early 1982, the Cumberland production well was the main source of water for all uses at Cumberland's Canton facility. The well is 46 feet deep and has a capacity of 792,000 gallons per day, although its typical output averaged about 270,000 gallons per day. Output varied according to production demand. Cumberland was also connected to the Canton public water system.
The Massachusetts Department of Environmental Quality Engineering ("DEQE") instructed Cumberland to stop using its well in 1979, because it was found to be contaminated. In early September 1979, after testing showed low levels of contaminants, DEQE authorized Cumberland to resume using the well. Cumberland, thereafter, used the well for non-food production purposes, until it shut the well down permanently in 1982.
Various petroleum and other chemical products were used at the Cumberland truck garage to fuel and service the company's vehicles. The products used included diesel fuel, lubricating oils, petroleum distillates and solvents. The major solvent used was "Safety Kleen", a product comprised primarily of mineral spirits. Other solvents used by Cumberland mechanics contained VOCs. Cumberland also used the refrigerant Freon-12 for the maintenance of its truck refrigeration units.
Under the Cumberland parking lots is a system of drains that are connected to catch basins which collect storm water runoff. Most of the runoff is discharged through the northwest storm sewer outfall ("NWSSO") located at the north central boundary of the Cumberland site. The NWSSO discharges into a drainage ditch that crosses the Canton Conservation Commission property and merges with Pecunit Brook. The runoff ultimately discharges into the Neponset River situated about 800 feet downstream of the NWSSO.
Liquids that would spill on the Cumberland property would either evaporate or would be collected by the storm drain system and discharged into the drainage ditch. The volume of flow from the NWSSO to the drainage ditch would vary according to weather conditions and site operations.
Conditions in and around the Cumberland truck garage were sloppy. Fuel and solvents commonly were spilled on the garage floor or the parking lot. Mechanics routinely dumped excess solvents on the parking lot or in a storm drain. Most of the chemical products that were spilled or discharged on the Cumberland site, however, were oils and non-chlorinated petroleum distillates. Spillage of chlorinated solvents such as 1,1,1-TCA was minimal in comparison.
Groundwater generally flows downhill or downgradient. The rate and direction of groundwater flow are subject to the permeability of the aquifer and the degree of the gradient.
The Upper Neponset Valley Aquifer transmits millions of gallons of water per day through the valley where the Neponset River and the White Lodge Well Field are located. Groundwater typically flows in parallel flow lines that follow land surface topography, with some discharge into the river from both sides.
Typical groundwater flow is also affected by the interaction of multiple pumping *1229 wells. In the vicinity of the White Lodge Well Field, seven large pumping wells have operated at various relevant times. During the 1970s, at least six were active. These included WL-1, WL-2, WL-3 and WL-4; Cumberland's production well; and Canton Well C-3. Each pumping well has a cone of influence and zone of capture that eventually draws groundwater into the well.
The pumping activity of nearby wells influences the size and shape of a capture zone. In essence, adjacent wells are in a tug-of-war competition for groundwater. This was the case in the vicinity of the White Lodge Well Field, where all of the several active pumping wells affected the capture zones of their adjacent wells. Each of these wells had its own capture zone from which groundwater and any groundwater contaminants would be drawn into the well. The boundaries of the capture zones shifted as pumping rates and conditions changed. The effect of such change can be simulated or predicted by computer modeling techniques.
During the 1970s, the Cumberland well had a capture zone that included all of the Cumberland site (other than its extreme northwest corner), the NWSSO, the upper portion of the drainage ditch, and upgradient areas south and west of the Cumberland site. These upgradient areas included the Shield Chemical property and the site of a break in the NNVS near Dedham Street.
The Cumberland capture zone was influenced by Canton well C-3. When C-3 was pumping, the Cumberland well captured most of the Shield property as well as the area of the Dedham Street sewer break. When C-3 was not pumping, the Cumberland well captured only the eastern portion of the Shield property and a small portion of the sewer break area. Moreover, when C-3 was inactive, WL-3 would capture much of the Shield property and sewer break area. After the Cumberland well was permanently shut down in 1982, what had been its capture zone (the Cumberland site, Shield and the NNVS), fell within the capture zone of WL-3.
When VOCs such as 1,1,1-TCA are spilled or discharged, they are subject to the processes of evaporation and absorption as they penetrate the land surface and work their way toward the groundwater. It is only when VOCs actually enter the groundwater that they are transported in the direction of the groundwater flow.
As they flow downgradient in groundwater, contaminants spread out both horizontally and vertically, forming a plume that increases in width and depth. As a plume enlarges, the concentration of contaminants within the plume decreases, eventually to non-detectable levels.
Here, when WL-3 is pumping, it causes groundwater to flow under the Neponset River from areas to the south and east, including the vicinity of the Shield property and the NNVS on the Canton side of the river. WL-3 also draws a significant amount of water from the river itself in this area. But when the Cumberland well was pumping, WL-3 drew almost no groundwater from the Cumberland site or the upper drainage ditch. These areas were within the capture zone of the Cumberland well.
The issue of causation involves two distinct questions: whether groundwater from the Cumberland side of the Neponset is drawn under the river into WL-3; and whether contaminants from the Cumberland site entered the groundwater and then were drawn under the Neponset to WL-3. The parties' experts, Mr. Miller and Mr. Robertson, agreed that groundwater flows under the river. The critical battle point at trial was whether contaminants from the Cumberland site ever reached the groundwater, *1230 and thereafter found their way to WL-3.
VOCs in groundwater are measured by parts-per-billion (PPB) concentration levels. They are measured by a testing procedure known as "EPA Method 624," employing the gas chromatography/mass spectrometry technique.
WL-3, the Cumberland well, and various monitoring wells in the area have consistently demonstrated contamination by six VOC solvents: 1,1,1-TCA; tetrachloroethylene (PCE); trichloroethylene (TCE); 1,1-dichloroethane; 1,1-dichloroethene; and 1,2-dichloroethene. A compound called trichlorofluoromethane (Freon-11 or R-11) was also found in WL-3 and its upgradient groundwater. The predominant groundwater contaminant, with the highest PPB concentrations, was 1,1,1-TCA, a solvent with numerous domestic and industrial uses. R-11 is used primarily as an aerosol propellant. The other chlorinated VOCs1,1-dichloroethane, 1,1-dichloroethene, and 1,2-dichloroetheneare products of the biodegradation of 1,1,1-TCA and PCE.
Biodegradation is a process by which bacteria transform one chemical compound into another. Biodegradation occurs either aerobically (with oxygen) or anaerobically (without oxygen). Anaerobic, as opposed to aerobic, biodegradation can transform 1,1,1-TCA and PCE into the four VOCs found in WL-3. The characteristics necessary for anaerobic biodegradation are lack of oxygen, a sufficient supply of nutrients, anaerobic bacteria, and sufficient time for the process to take place.
This court is persuaded that WL-3 and the upgradient monitoring wells were affected by a plume of contamination caused by the biodegradation of the parent compounds 1,1,1-TCA and PCE. The court also finds that these parent compounds, plus R-11, were discharged directly into the groundwater at the source or sources of contamination.
The source of groundwater contamination can be distinguished or "fingerprinted" by identifying the pattern and ratio of chemical constituents that consistently appear in groundwater monitoring or production wells within the plume. Plumes from various sources can also be distinguished by the use of chemical tracers. Tracers are chemicals consistently used and discharged in significant quantities by a particular source. They travel in-groundwater in a way similar to the VOCs in issue. Here, chemical tracers from Cumberland included methylene chloride; the non-chlorinated VOC solvents benzene, toluene and xylene; and petroleum distillates. Chemical tracers for Shield include the aerosol propellant R-11 and the solvent trichlorotrifluoroethane, also known as Freon TF (R-113). Chemical tracers are reasonably unique to a given source.
Groundwater monitoring wells are used for the purpose of sampling and measuring the groundwater, to determine whether a plume of contaminants is present and, if so, to identify the characteristics of that contamination. Monitoring wells are used routinely by hydrogeological consultants to determine the source and extent of groundwater contamination.
Plaintiffs' hydrogeological consultants Geraghty & Miller ("G & M") installed 13 groundwater monitoring wells in 1981-82 (GM 1-13). Six of them were placed along the length of and immediately adjacent to the Cumberland drainage ditch for the prime purpose of determining whether the ditch was causing the groundwater contamination that affected WL-3.
These wells had the capacity to detect groundwater contamination coming from the Cumberland site or ditch and traveling toward WL-3. It is important to note that groundwater samples taken from these monitoring wells on several occasions in 1982 were entirely free of 1,1,1-TCA, PCE and TCE. On one occasion, a minor amount of the breakdown product 1,1-dichloroethane was found in GM-7. But, significantly, GM-7 was located adjacent to the NNVS. Otherwise, the 1982 study by *1231 G & M found no evidence of a plume of groundwater contamination emanating from Cumberland and traveling toward WL-3.
A similar monitoring process was undertaken in 1987 as part of the DEQE Neponset Valley Aquifer study. SEA Consultants, Inc. ("SEA") and its subcontractor, Pine & Swallow ("P & S") installed monitoring wells (PS 13-16) at a point between the Cumberland site and WL-3 to determine whether any groundwater contamination was emanating from the Cumberland site that could affect WL-3. The testing result showed these wells to be clean and free of the VOCs in question here.
In addition to its downgradient monitoring wells, in 1982 G & M installed several monitoring wells upgradient from Cumberland (GM 10-13). This testing revealed significant levels of the suite of chlorinated VOCs at issue here. Thereafter, G & M sampled the Town of Canton's monitoring well (C2-2), also upgradient of the Cumberland site. Significantly, it was determined to have the highest level of 1,1,1-TCA found in any well. This court is persuaded that the contamination found in these upgradient monitoring wells (GM-12 and C2-2) demonstrates the existence of contamination sources upgradient from the Cumberland site.
In 1987, SEA/P & S also installed 12 monitoring wells (T1 through 12) at the Shield site, upgradient from Cumberland. Those wells on the downgradient (northern) side of Shield showed high levels of the chlorinated VOCs found in WL-3. The groundwater contamination in the T wells, particularly T1-T3, increased with depth, indicating a deep plume of groundwater contamination leaving Shield in the direction of WL-3.
Later, SEA/P & S installed additional monitoring wells on the Canton Water Department property, southwest and upgradient of the Cumberland site and downgradient of the Shield Chemical property (PS 17-23). Several of these wells were located near GM-12 and the Dedham Street sewer break. A number of these wells demonstrated significant levels of the chlorinated VOCs at issue here, particularly at the deepest levels.
The only monitoring well between Cumberland and WL-3 that showed any 1,1,1-TCA was PS 14, located at a point north of the Cumberland site between the NNVS, the railroad and the Neponset river. Of significance is the fact that PS 14 was located downgradient of both Shield and the Dedham Street sewer break, within the capture zone of WL-3. It was downgradient of only the northwestern corner of the Cumberland site.
The suite and ratios of the chlorinated VOCs consistently found in numerous upgradient monitoring wells, the Cumberland well, and WL-3 contribute to persuading this court that the contamination found in WL-3 emanates from a single plume with an upgradient source or sources, the likeliest being Shield and/or the sewer break.
The test results of the various monitoring wells are essential in resolving the issue of causation. The fact that there may be contamination at a particular point in surface water does not necessarily indicate the probability of groundwater contamination, or that surface water is a source of groundwater contamination. Groundwater contamination must be established by analysis of samplings from monitoring or production wells.
*1232 For example, the G & M geological logs indicate that the area of the Cumberland drainage ditch is over a layer of peat at least 3 feet thick. Peat consists of natural organic matter. VOCs tend to stick to peat, and are essentially filtered from surface water as it flows through peat. Peat also has a low permeability. This means that a layer of peat would naturally impede the seepage of water from the ditch downward to the groundwater.
Oil and other petroleum distillates are lighter than water and, therefore, also attract VOCs. These products, when discharged from the Cumberland drainage system to the ditch, would have attracted VOCs, floating on the surface water of the ditch to the Neponset River. VOCs tend to evaporate when exposed to the air. VOCs in the surface waters of the drainage ditch would, therefore, evaporate as the water flowed downstream, a process that would further reduce the concentrations of VOCs in the ditch water.
The combined impact of the peat, petroleum distillates and evaporation significantly reduced the opportunity for VOCs in the ditch to reach the groundwater. This confluence of phenomena serves to explain why VOCs periodically found in the ditch surface water were not detected by the groundwater monitoring wells along the ditch.
The VOC levels in water samples taken from the drainage ditch since 1982 have been quite low, and generally within drinking water standards. The water discharge from the NWSSO to the ditch has been essentially uncontaminated stormwater runoff, and an unlikely source of the contamination in WL-3.
Also highly probative is the fact that the contaminants found in the NWSSO and the ditch, and those found in WL-3 and the upgradient groundwater monitoring wells were an inconclusive match. For example, chemicals and petroleum distillates that were found on multiple occasions in the NWSSO and drainage ditch, were not found in the groundwater. Because these chemicals had the same general characteristics as the VOCs in issue here, their absence from the groundwater further indicates that VOCs from the ditch did not reach the groundwater.
Additionally, in January 1983, G & M analysis of soil samples from four locations in the drainage ditch showed no 1,1,1-TCA. This G & M study indicates that VOCs never passed through the soil under the ditch, making the ditch an unlikely source of groundwater contamination.
A 1982 series of samplings of six groundwater monitoring wells installed along the drainage ditch by G & M showed clean groundwater with no detectable levels of 1,1,1-TCA. As noted above, the same result was obtained in 1987 from a monitoring well (PS-14) installed by SEA/P & S. Such analyses are inconsistent with the plaintiff's contention that the ditch was a source of groundwater contamination, and the court therefore finds that the ditch was not a source of contamination.
As has been pointed out, the group of chemicals used at the Cumberland site does not match or correlate with the suite of VOCs consistently found in WL-3 and the upgradient groundwater. Moreover, for most of the twenty-year period between 1962 and 1982, the Cumberland production well was in consistent use and would have captured any groundwater contamination caused by any releases on the Cumberland site or from the NWSSO and upper ditch. With a single exception during that period, the VOC concentrations in the Cumberland well were significantly lower than those found in WL-3. The court is persuaded that if the Cumberland site were the source of contamination, the VOC concentrations in the Cumberland well would have to have been significantly higher than those at WL-3, given the substantial dilution that would take place over the 1500 foot distance between the two wells.
The groundwater downgradient of the Cumberland site, and between it and WL-3, has been determined to have been consistently clean, according to the monitoring well studies conducted by G & M and SEA. Of particular significance is the fact that no 1,1,1-TCA has ever been detected by any monitoring well downgradient of the Cumberland site. Yet groundwater tests *1233 upgradient of the Cumberland site have consistently demonstrated high levels of the same suite and ratios of VOCs found at WL-3 since monitoring began in 1982. This data clearly points to the existence of upgradient sources of contamination that affected both the Cumberland well and WL-3 when both were pumping.
Of further significance is the fact that chemicals spilled at the Cumberland site would not have had time to biodegrade into the same suite of anaerobic contaminants found in WL-3. This is due to the short travel time necessary for any groundwater under the Cumberland site to reach the well.
In summary, therefore, three significant factors tended to prevent VOCs discharged into the Cumberland storm drainage system from reaching groundwater: (1) the capture of VOCs by the petroleum distillates dumped on the ground or into the drainage system; (2) the layer of peat underlying the ditch; and (3) the process of volatilization or evaporation of VOCs from the site's surfaces, as well as from the surface waters of the ditch.
OTHER POTENTIAL CONTAMINANT SOURCES
Indeed, the evidence establishes that the Cumberland well was itself contaminated by one or more upgradient sources that also affected WL-3. The Shield Company and the sewer leak are likely upgradient sources that impacted both Cumberland and WL-3.
Since 1962, Shield has operated a manufacturing facility in two buildings in Canton, east of the Neponset River and southwest of the Cumberland facility. Until the early 1980s, Shield was engaged in the aerosol packaging business, utilizing a wide variety of chemical products. The Shield plant is located upgradient of WL-3, the Dedham Street sewer break, and the Cumberland site. When both WL-3 and the Cumberland production well were in operation, and Canton well CF-3 was off, portions of the Shield property were included in the capture zones of both WL-3 and the Cumberland well. All of the Shield property has been within the capture zone of WL-3 since 1982, when the Cumberland and Canton wells were shut down.
Shield used many VOCs, both in its aerosol products and as cleanup solvents. The chemicals to be injected in the aerosol cans were mixed in kettles that were later washed out with water or various solvents, including 1,1,1-TCA. Shield used very large quantities of the tracer chemical R-11, until R-11 was banned by the EPA in 1978.
The Shield facility is connected directly to the NNVS. A number of its floor drains discharged into the sewer. Until late 1979, when it was ordered to stop the practice by the Metropolitan District Commission ("MDC"), Shield routinely dumped substantial amounts of solvent wastes, including R-11 and 1,1,1-TCA, into the MDC sewer.
Several groundwater monitoring wells installed for the DEQE's Neponset River Aquifer study found significant levels of chlorinated VOC contamination downgradient from Shield with the same suite and ratios found in WL-3, but in larger amounts. The highest levels of contamination detected by these monitoring wells were deep within the aquifer. The data from these monitoring wells supports a finding that a plume of VOC contamination emanates from the Shield property and moves northward in the direction of WL-3.
The chemical contaminants recently found in the SEA/P & S monitoring wells on and downgradient of the Shield site have a chemical fingerprint identical to that found in nearby monitoring wells GM-12 and C2-2 in 1982 and in WL-3 and the Cumberland well in 1979.
Given these findings, this court concludes that the Shield facility is a probable source of the VOCs entering both the NNVS and the groundwater in areas upgradient of both the Cumberland site and WL-3. Shield must, therefore, be regarded as having been a probable source of contamination of both the Cumberland well and WL-3 from 1979 to the present.
*1234 B. NNVS
The NNVS is a major interceptor sewer that carries the wastewater from a number of communities south of Boston. It was formerly owned and operated by the MDC, and is now under the jurisdiction of the Massachusetts Water Resources Authority. The NNVS runs generally from south to north along the Neponset River in the area of the White Lodge Well Field. It crosses the western end of the Cumberland site north of Dedham Street.
The NNVS, constructed by the MDC in the early 1930s, is made of concrete pipe with a vertical diameter of 60 inches. An 8-inch underdrain was constructed under the NNVS that allows the transmittal of wastewater contamination that may leak from the sewer.
WL-3 was installed approximately 700 feet to the west of the NNVS. The Westwood Extension Sewer, installed in 1960, runs less than 200 feet south of WL-3. The WES and NNVS intersect at a point approximately 700 feet east of WL-3. The WES collects the wastewater from the University Avenue Industrial Park that surrounds the White Lodge well field. The NNVS conveys industrial wastewater, including that of Shield and the towns of Canton, Norwood, Stoughton, Walpole and Westwood, through the White Lodge well field to Boston Harbor.
The NNVS has a number of hydraulic problems, including flow constriction and excessive loading of wastewater. It periodically surcharges and discharges wastewater through manholes in the vicinity of the White Lodge well field.
Samples taken by DEQE in 1981 from the NNVS in the vicinity of the White Lodge well field showed elevated levels of the VOCs in question. Sewage is a source of the anaerobic bacteria and nutrients necessary for anaerobic biodegradation of the chlorinated VOCs in question.
At a point approximately 1,000 feet upgradient from WL-3 and south of the Cumberland site, a 100-foot section of the NNVS settled and the concrete pipe broke. The MDC attempted to repair the sewer in 1960, but the underdrain was not repaired at all. Investigations in 1983 and 1984 demonstrate that the MDC's 1960 repair effort was unsuccessful, and that the sewer contained serious structural deficiencies.
A 1984 consultant's report notes a severe 30-inch "vertical mis-alignment" of the sewer pipe, with broken and cracked pipe and visible tree roots at the Dedham Street location. The collapse and attempted repair resulted in a constriction that causes the sewer to run full and under pressure there. The location of the collapse is upstream and upgradient from the Cumberland site. Sewage discharges from the Cumberland facility could not flow upstream to this point.
The NNVS lies within the capture zone of WL-3 when the well is pumping. The Dedham Street sewer break is above the water table. These facts, together with the overloading of the sewer in the Dedham Street area, lead to the conclusion that sewage has leaked into the surrounding soil and groundwater.
The chlorinated VOCs, such as 1,1,1-TCA and R-11, that were discharged into the NNVS by Shield upstream of WL-3 and the Cumberland site, are heavier than water and would tend to collect and settle to the bottom of the sewer and provide a consistent supply of VOCs in the sewage. SEA sampled several monitoring wells near the Dedham Street sewer break and between the WES and WL-3. These samplings disclosed indicators of sewage pollution. Elevated levels of the sewer indicator pollutants ammonia and chloride have also been found in WL-3. Groundwater monitoring wells installed by SEA/P & S adjacent to the Dedham Street sewer break, and also nearby well GM-12, have shown very high levels of the chlorinated VOCs in question, including 1,1,1-TCA. Sewage leaking from the NNVS at the Dedham Street sewer break is likely to run along the 8-inch underdrain or gravel bedding constructed under the sewer, thereby forming a conduit by which escaped contaminants may travel to locations downgradient from the source.
All of this evidence leads the court to the conclusion that the NNVS has leaked and is leaking VOC-contaminated sewage to the *1235 soils and groundwater at the Dedham Street sewer break. As a result, the NNVS has been a contributing source of the chlorinated VOCs in the groundwater that have contaminated WL-3.
The Cumberland well and plaintiffs' WL-3 were contaminated in 1979, and thereafter, by a source or sources of VOCs upgradient from them. The Cumberland site and the drainage ditch are not a source of the contamination found in WL-3. Both Shield and the NNVS sewer, however, are probable continuing causes of the contamination of WL-3. Plaintiffs having failed to prove by a preponderance of the evidence that Cumberland Farms was responsible for any groundwater contamination that affected WL-3, judgment will enter for the defendant.
An order will issue.NOTES
 Cumberland has assumed the assets of two other defendant corporations.
 In its Consolidated Complaint, DWC seeks injunctive relief, response costs, and damages pursuant to: (1) The Comprehensive Environmental Response, Compensation and Liability Act of 1980 ("CERCLA"), 42 U.S.C. §§ 9601 et seq. (1983 & Supp. 1988); (2) the Clean Water Act, 33 U.S.C. §§ 1251 et seq. (1986 & Supp.1988); (3) The Resource Conservation and Recovery Act ("RCRA"), 42 U.S.C. §§ 6901 et seq. (1983 & Supp.1988); (4) Massachusetts common law of nuisance; (5) Massachusetts common law of negligence; (6) Massachusetts common law of strict liability; (7) Massachusetts common law of trespass; (8) the Dedham Water Company Charter, 1876 Mass.Acts ch. 138, § 8, as supplemented; (9) the Dedham-Westwood Water District Charter, 1985 Mass.Statutes, ch. 193, § 9, as supplemented; (10) Sections 4 and 5(a) of the Massachusetts Oil and Hazardous Material Release Prevention and Response Act, Mass.Gen.L. ch. 21E, § 4 and § 5(a) (Supp.1987); and (11) the Massachusetts Clean Waters Act, Mass.Gen. L. ch. 21, § 27 (1981), as amended by 1983 Mass. Statutes ch. 7, §§ 2, 9.
 Courts are in general agreement on the elements of a CERCLA action. According to one frequently-quoted formulation:
Stripping away the excess language, the statute appears to impose liability on a generator who has (1) disposed of its hazardous substances (2) at a facility which now contains hazardous substances of the sort disposed of by the generator (3) if there is a release of that or some other type of hazardous substance (4) which causes the incurrence of response costs.
United States v. Wade, 577 F. Supp. 1326, 1333 (E.D.Pa.1983).
 Note that "[s]trict liability under CERCLA, however, is not absolute," State of New York v. Shore Realty Corp., 759 F.2d 1032, 1042 (2d Cir.1985). Section 107(b) of CERCLA provides affirmative defenses as follows:
There shall be no liability under subsection (a) of this section for a person otherwise liable who can establish by a preponderance of the evidence that the release or threat of release of a hazardous substance and the damages resulting therefrom were caused solely by
(1) an act of God;
(2) an act of war;
(3) an act or omission of a third party ...; or
(4) any combination of the foregoing paragraphs.
42 U.S.C. § 9607(b). Plaintiffs argue that this section precludes defendant's causation argument, since the affirmative defenses listed are exclusive. See 42 U.S.C. § 9607(a) (liability under section is "subject only to the defenses set forth in subsection (b) of this section"). But plaintiffs have misconceived defendant's argument. Under CERCLA, causation is an element of plaintiff's case, and a defendant may always escape liability by negating an essential element of the case against it. See United States v. B.R. Mackay, 13 Chem. Waste Lit. Rep. 253, 254 (N.D.Ill.1986) [available on WESTLAW, 1986 WL 13717] ("the only defenses to liability under CERCLA, other than those which negate a required element of proof under Section 107(a), are those set forth in Section 107(b)") (emphasis supplied).
 Even in a one-site case, causation may become an issue. For example, a defendant could argue that its wastes were not in fact disposed of at a given site. See United States v. Wade, 653 F. Supp. 11, 21 E.R.C. 1346, 1348 (E.D.Pa.1984) (proof of actual disposal at site necessary to establish liability). Or, a defendant might prove that, subsequent to disposal, its wastes were completely removed from the subject site. See United States v. Ottati & Goss, Inc., 630 F. Supp. 1361, 1402-03 (D.N.H.1985); United States v. South Carolina Recycling and Disposal, Inc., 653 F. Supp. 984, 993 n. 6 (D.S.C.1984); 2 Law of Hazardous Waste: Management, Cleanup, Liability, and Litigation § 14.01[c][ii][E] (S. Cooke ed. 1988).
 Section 8 of the Dedham Water Company Charter, 1876 Mass.Acts ch. 138, imposes liability upon defendants who "wantonly or maliciously divert the water [taken by DWC under its statutory authority] ... or corrupt the same or render it impure." Similarly, § 9 of the Dedham-Westwood Water District Charter, 1985 Mass.Acts Ch. 193, fixes liability upon a defendant who "wilfully or wantonly corrupts, pollutes or diverts any water" supplied under the act. A finding for plaintiff under either Charter would require proof that Cumberland's discharges actually harmed DWC's water supply. The Massachusetts Oil and Hazardous Material Prevention and Response Act, Mass.Gen.L. ch. 21E, § 5(a), allows a plaintiff to recover property damages "incurred or suffered as a result of" a defendant's releases (emphasis supplied). Section 4 of the Act awards response costs to "any person threatened or damaged by" hazardous releases. The Massachusetts Clean Waters Act, Mass.Gen.L. ch. 21, § 27(14), makes "the person responsible for causing" hazardous discharges liable for resulting property damage.
 A defendant is liable for private nuisance "if, but only if, his conduct is a legal cause of an invasion of another's interest in the private use and enjoyment of land ..." Restatement (Second) of Torts § 822 (1979). A defendant is liable for negligence if its conduct is "a legal cause of harm to another," a test that requires the negligent conduct to be "a substantial factor in bringing about the harm." Restatement (Second) of Torts § 430 (1965). Under the doctrine of strict liability for ultrahazardous conditions, "[o]ne who carries on an abnormally dangerous activity is subject to liability for harm ... resulting from the activity." Clark-Aiken Co. v. Cromwell-Wright Co., Inc., 367 Mass. 70, 323 N.E.2d 876, 887 (1975), citing Restatement (Second) of Torts § 519 (Tent. Draft no. 10, 1964) (emphasis supplied). Finally, liability in trespass is imposed where the defendant "enters land in the possession of the other, or causes a thing ... to do so." Restatement (Second) of Torts § 158 (1965) (emphasis supplied).
 DWC seeks injunctive relief and the exaction of civil penalties, based upon the "citizen suit" provisions of the federal Clean Water Act, 33 U.S.C. § 1365(a) and RCRA, 42 U.S.C. §§ 6972(a) (1) (b) and 6973, as well as the permit requirements and generator standards of RCRA, 42 U.S.C. §§ 6922, 6924 and 6925. A necessary predicate to injunctive relief and civil penalties would be a finding that Cumberland was the source of the groundwater pollution at issue.
 Computerized groundwater modeling is a technique that takes into account all the key influences on groundwater flow patterns under varying pumping conditions. Groundwater modeling has been a standard tool for studying groundwater problems since the mid-1970s. It enables the interested observer to understand what is actually happening at a given time and location within the aquifer. Cumberland's expert hydrogeologist, Mr. Robertson, employed computerized groundwater modeling techniques. DWC's expert Mr. Miller did not. The 1982 study by Geraghty & Miller, upon which DWC relied, focused on the capture zone of WL-3, which it assumed was unaffected by the other neighboring pumping wells.
 The only chemist to testify at trial was Cumberland's witness, Dr. James Smith, who impressed the court as having considerable experience in groundwater contamination investigations. The court, therefore, accepts his testimony and relies on it in making these findings.
 Data appended to the 1982 G & M study reported high VOC levels in the NNVS sewer in 1981. The text did not discuss the significance of this finding.
 The Massachusetts Legislature in 1985 authorized this study of the sources, nature and extent of groundwater contamination in the Upper Neponset Valley Aquifer. SEA was selected by DEQE to perform the study. The recently completed study concludes, based on the Phase II monitoring wells installed by P & S, that both Shield and the NNVS were probable contamination sources that affected WL-3. The SEA study did not conclude that Cumberland was a probable source of contamination of the White Lodge Well Field.
 In particular, G & M found high levels of contaminants in GM-12, upgradient of the Cumberland site near the Dedham Street sewer break. G & M concluded that there was a probable source of contamination near GM-12, and recommended further investigation. No investigation was undertaken.
 Because the issues in dispute were technical and complex, this case necessarily consisted largely of the presentation of expert evidence. That evidence was presented in an extremely thorough and highly professional manner by both lead counselMr. Thomas F. Holt, Jr. for the plaintiffs and Mr. Allan van Gestel for the defendant. Although the expert testimony for both parties was impressive, the court has been persuaded by the defendant's team of experts. Consequently, the court accepts and adopts the opinions and findings of the defendant's experts, and rejects the findings and opinions of plaintiffs' experts insofar as they may be in conflict with those of the defendant's experts.