Citations
- 131 F. Supp. 3d 802
Full opinion text
OPINION AND ORDER
BARBARA B. CRABB, District Judge.
From 1885.until 1947, a manufactured gas plant operated in Ashland, Wisconsin, on a bluff above Lake Superior, generating gas for heating and lighting. For that entire time, the. plant discharged tar wastes into the bluff. In 1987, plaintiff Northern States Power Company acquired the plant site and shortly thereafter discovered contamination. It notified the Wisconsin Department of Natural Resources, which began an investigation of the site in 1994.
In 2002, the entire site, which includes the bluff on which the plant stood, the land below the bluff, known generally as Kreher Park,.Chequamegon Bay of Lake Superior and the deepwater in the Copper Falls aquifer underlying the site, was added to the National Priorities List and the Environmental Protection Agency took over the state’s work. Ten years later, the agency entered into a consent decree with plaintiff, which agreed to remediate the site.
Now plaintiff is pursuing claims against defendants City of Ashland and Ashland County for contribution to the costs of the cleanup under 42 U.S.C. § 9613. Although plaintiff has accepted responsibility for the remediation of the Upper Bluff, it denies that the discharges are the sole source or even the major source of the contamination that has been found :in Kreher Park and in Lake Superior’s Chequamegon Bay. Instead, plaintiff contends that both Ashland County and the City of Ashland bear responsibility-for all or most of the contamination below the bluff.
As to defendant Ashland County, plaintiff allegés that in the 'three years from 1939 to 1942 in which it owned the area, it failed to clean up the wood treating chemicals left by the former owner, Schroeder Lumber Company, leaving these chemicals to sink into the ground and eventually create a waste pond. In addition, plaintiff alleges that while the County' owned the land, it demolishéd the company’s smokestacks and refuse burner, adding contaminants to the soil. As to the City of Ash-land, plaintiff alleges that this defendant has owned the land since 1942 and that it too failed to clean up the wood treating chemical waste. In addition, plaintiff alleges, the City dispersed contaminants through construction projects in the área, drained the waste dump into the bay, sanctioned the dumping of refuse and the changing of oil for city vehicles and built a city sewer through the park that emptied into the bay. Plaintiffs claims' are asserted under the Comprehensive Environmental Response, Compensation, and Liability Act, or CERCLA, 42 U.S.C. §§ 9601-75.
■ From the evidence adduced during the eight-day trial to the court, I find that defendant County is not covered under CERCLA and cannot be held liable for any hazardous substances in the land below the bluff or in the bay. Although the County gained title to the Schroeder Company property after the company failed to pay its property taxes in the 1930s, it did so only as a result of the company’s tax delinquency and it did not cause or contribute to the release or threatened release of hazardous substances from the facility.
Unlike the County, defendant City has potential liability under CERCLA because it operated Kreher Park. However, plaintiff has not shown that the City was responsible for the addition of any hazardous substances to the site. At most, it may have dispersed some of the existing contamination discharged from the manufactured gas plant. Even assuming that it did, the, equitable allocation of response costs is a discretionary decision for the court to make, taking into consideration such factors as the relative contributions of hazardous waste, the plaintiffs settlements with other parties, the defendant’s level of voluntary cooperation with the cleanup efforts, the extent of the parties’ financial resources and the benefit of the cleanup to the potential contributor. In this case, the factors, bearing on that determination weigh in the City’s favor and support an allocation of 0%.
For the purpose of deciding this case, I find the following facts from the evidence adduced at trial.....
FACTS
A. Background
Plaintiff Northern State's Power first discovered tar waste in the bluff when it began construction work on the plánt site in 1987, after acquiring the site through merger of its subsidiary, NSP-Wisconsin, with Lake Superior District Power. Two years later, the City of Ashland discovered soil contamination in Kreher Park, a city park that lies between the bluff and Chequamegon Bay. (It is not clear whether the boundaries of the park and the area beneath the bluff and within the remediation area are congruent, but because it makes no difference to the outcome of the case, I will assume they are.) The park occupies a portion of an area that was open water in the 1880’s and now -lies on a substratum of wood waste dumped into the bay by the lumber companies that lined the shoreline in the late 1880s and early 1900s.
In 1994, when the Wisconsin Department of Natural Resources began investigating the site and undertaking cleanup efforts, it discovered that the contamination of the site was more widespread than it had previously believed, with the principal contamination being two substances. The first was polycyclic aromatic hydrocarbons or PAHs, a category of chemicals often found together in groups of two or more. They “are created when products like coal, oil, gas and garbage are burned but the burning process is not complete.” Polycyclic Aromatic Hydrocarbons (PAHs) Fact Sheet, EPA, Jan. 2008. The other was non-aqueous phase liquids, or NAPLS. Non-aqueous phase liquids are either light non-aqueous phase liquids, that is, one of a group of organic substances, including petroleum chemicals, that are relatively insoluble in water but less dense than water, whereas dense non-aqueous phase liquids are also relatively insoluble in water but are denser than water and tend not to mix with it. http://toxics.usgs.gov/definitions/ napls.html, visited Sept. 9, 2015. In 2002, the site was added to the National Priorities List, qualifying it for federal funding for cleanup activities. With this action, the Environmental Protection Agency took over the state’s work, entering into an agreement with plaintiff for a study and significant cleanup of the site.
In 2012, plaintiff entered into a consent decree with the EPA covering the remediation of the site. Plaintiff then brought suit against the City of Ashland, Ashland County, two railroads that operated at the base of the bluff at one time (the Soo Line Railroad and the Wisconsin Central Railroad, Ltd.), and L.E. Myers Company (which allegedly played a role in the operation of the manufactured gas plant between 1917 and 1922), in an effort to obtain contribution from other potentially responsible parties for the anticipated costs of restoring Kreher Park and Chequamegon Bay. Plaintiff asserted claims under both state law and CERCLA for contribution, cost recovery, declaratory relief and damages. Both railroads settled with plaintiff before trial and L.E. Myers reached a settlement shortly after the start of trial. The state law claims were asserted only against the railroads, leaving only CERCLA claims against the City and County.
B. The Site
The site at issue covers about 40 acres of land in the city and county of Ashland, along the southern shore of Chequamegon Bay. For remedial and investigative purposes, the site has been divided into four parts, referred to by the government agencies as “areas of concern”: (1) the soils and groundwater of the Upper Bluff, which is where the manufactured gas plant was located; (2) the Kreher Park soils and shallow groundwater; (3) the sediments in the bay; and (4) deep groundwater in the Copper Falls Aquifer underlying the site. Only the first three areas play any part in this case.
The manufactured gas plant was built on the Upper Bluff in 1885, with about one-fifth.of the plant straddling a ravine that ran down to the bay. Within the next few years,-the ravine was filled in completely with a mix of materials, including the nearly impenetrable native soil known as “Miller Creek.” (The exact date in which the ravine was filled completely is unknown. All of the witnesses on the subject and the EPA agree that it was filled by the early 1900s.).
In 1885, the Ashland area was riding a short-lived economic boom fueled by the harvesting and sale of timber from Wisconsin’s extensive forests of eastern white pine. ■ Lumber companies had established sites at the harbor and .rafts apd ships were plying the lake with cargoes of harvest timber and lumber products. Railroads moved in and built, tracks along the base of the bluff, running in a generally east-west direction along the waterfront.
Within a short time, the waterfront became so full of timber waste that it could support- buildings. From 1901 until the early 1930s, the John Schroeder Lumber Company owned and occupied a good part of the area at the- base of the bluff below the gas plant. It stopped processing lumber sometime around 1930; in 1939, it lost any remaining ownership rights it had in the land when it failed to pay its property taxes apd the property reverted to defendant Ashland County.
In 1942, the City took over the County’s ownership rights in the area, which is now called Kreher Park. In 1951, it began construction of a wastewater treatment plant in the northeastern part of the park area. It expanded the facility in the early 1970s and considered further expansion in 1989 until early explorations of the expansion site disclosed contaminated soil and groundwater. The City notified the Wisconsin Department of Natural Resources and began building a new plant a few miles away.
C. The Manufactured Gas Plant
1. Gas production
During the 62 years that the manufactured gas .plant was operating, it used three different processes to manufacture gas, starting with the Patton Process, which vaporized naphtha oil. In 1899, the plant switched to a carbureted water-gas process and used it until 1947> when it stopped- manufacturing gas altogether. For a short period around 1917, the plant also used a coal gasification process that did not use petroleum for carburization.
Manufacturers of gas employ large “holders,” both for. containing the gas before it is sent .out to. customers and for relief purposes when the production exceeds immediate demand. The Ashland gas plant began with one relief holder, known as Holder No. 1, which was used in the original Patton Process and possibly in the initial period of the carbureted water-gas process. At some time, that first holder was replaced by a second holder, No. 2, that was used as a production holder for the early years of carbureted water-gas production until it too was replaced by two new holders, Nos. 3 and 4, which were used until production ended at the site. Also on site at one time or another were tanks for storing gas oil (a heavier version of crude oil) and naphtha. The naphtha tanks were built above ground at first, but were replaced by underground tanks at some time before 1923.
Each of the gas holders had a metal superstructure on the outside. Within this superstructure, a structure resembling an inverted cup moved up and down over the gas as it was produced and then released. A water tank on the bottom formed a water seal so that the gas did not escape, but rode up and down with the inverted cup structure.
In producing carbureted water-gas, the plant operators started with a gas that was enriched with a petroleum product, put it through a superheater to insure that the added petroleum vapor would stay in vapor form and then sent the gas through a “washbox,” which took out about 30 percent of the tar from the gas stream. From there, the gas continued into a “relief holder,” after which a variety of condensers and tar extractors took the remainder of the tar out of the gas stream in preparation for the production holder. At this point, very little tar remained in the gas, as shown by the near absence of contamination in the ground that has been excavated under holder no. 4, which was used as a production holder.
So long as the plant was in operation, the washbox and relief holder would have released tars and wastewater. The relief holder alone would have accumulated about 70% of the tars from the process, all of which had to overflow at some point. Although the plant had a tar well on site by 1923, it, too, would have overflowed regularly as fluids were added to it. Inevitably, tar wastes overflowed into the ravine, unencapsulated, or later, in pipes, after 1902, when the city enacted an ordinance requiring gas plant waste to be discharged only through underground sewers.
The precise; nature of the emulsions and of the tar changed, depending on the feedstock. If, for example, anthracite eoal was used as the carbon source, the composition of the tars and waste products would be different from the composition of the waste produced by bituminous coal. Other differences would result if the company used naphtha, gas oil or heavier oils. Records from approximately 1908 to 1947 show that changes in the BTU value of the gas during the time period, indicating changes in the materials used.
To meet the requirements of the 1902 city ordinance, the manufactured gas plant installed vitrified clay tiles in various diameters for discharge of tar waste into the ravine. These clay pipes were made of pieces joined end-to-end-in a simple friction fit for use in conveying drainage or waste.
During the entire period in which it operated, the gas company submitted annual reports to the Wisconsin Public Service Commission and to the Railroad Commission. ' The reports showed that the gas plant produced approximately 1.4 billion cubic feet "of gas over its lifetime. This level of production would have generated 700.000 to 1,000,000 galldns of tar and tar emulsions. Düring this same period, the company reported selling about 126,400 gallons of tar and using 11,200 gallons of tar for boiler fuel. In addition, as of early 2014, 14,000 gallons of tar had been recovered from the Copper Falls Aquifer. If these 151,600 gallons of known tar releases are subtracted from the amount generated over the years, approximately 450,000 to 840.000 gallons of tar releases are unaccounted for, without counting the 1.5 to 3 millions gallons of wastewater containing tar emulsions that would have had to go somewhere.
.Neither, the company’s own records nor those of the two commissions with which the company filed annual -reports show that the plant ever engaged in tar sales of any size. Nothing in any record indicates that the gas plant had the equipment that would be expected to be found in a plant that was undertaking significant tar recovery for resale, such as separator, tar dehydrator or other tar handling equipment.
2. . Excavation findings
During the excavation of the Upper Bluff, investigators found a number of the vitrified clay tiles within the gas plant site, in the filled ravine and running east and west along the base of the bluff where they would have been connected to an open ditch running to the bay. Many of these clay pipes discharged tar and tar emulsions directly to the ravine. From there, the waste would have migrated down to Kreher Park, which is more than 20 feet lower in elevation than the gas plant site.
During excavations in 2014, investigators found black viscous, tarry material within many, of the clay pipes, one of which was a 12-inch diameter pipe at the very bottom of the ravine, running under. St. Claire Street (an east-west street located at the base of the bluff). As the investigators encountered the pipes and broke them open, tar flowed out, demonstrating its continuing mobility, in the environment. Originally, the 12-inch pipe was connected to an east-west pipe that carried the tar and other waste to an open ditch ending at the bay, but the connection failed between the two pipes at some time, presumably before 1939, when the open ditch was filled in. The loss of the connection allowed the direct discharge of! the 12-inch pipé into the wood fill material in the old lake bed that is now the park. Some of the tar continues to migrate through Kreher' Park out toward the bay through the wood fill.
The excavations revealed three different pathways from the gas plant: (1) before 1902, the tar flowed down through the ravine under the gas plant and into bay; (2) in 1902, after the city enacted its ordinance on waste discharges, the plant began discharging its tar waste through clay tile pipes down through the filled ravine, then along the base of the bluff to the west to the open ditch that took the waste out into the bay; and (3) after the east-west pipes clogged up, the tar waste flowed through the filled ravine in clay tiles into the wood waste area and into the seep.
The relatively impermeable clay-rich, reddish-colored soil in the Ashland area has protected the Copper Falls Aquifer underlying the remediation site by not allowing downward movement of the contaminants in that area. However, the relatively loose ■ materials that filled the ravine allowed tar to travel down the ravine and eventually into the wood waste underlying Kreher Park,' which allowed the tar to spread underground.
The 2014 excavations. of the gas plant site uncovered a tarry mass that had been in gas holder No. 2 at the plant site. Other tar samples were obtained, from the area in which the plant had been located, near the boiler house and the coal carbonization location (nearly 100 years after the coal carbonization experiment had ended at the gas plant) and from the bottom of gas holder No. 1, which went out of service in the early 1900s. Sampling of other pipes in the Upper Bluff showed similar tarry material.
Excavations in 2002 revealed wood slab and wood debris under an area known as the seep,-which lies at the base of the bluff and extends in a generally northward direction toward a waste tar dump. The excavations also unearthed very black, tarry material that was determined to be carbureted water gas waste, as were all other samples taken from this area.
In 2005, an investigation of the waste tar dump area showed that contaminated wood waste was present in the dump area. The contamination in the park flowed below the fill material in the ravine and through the waste'wood layer, which variés in depth from 2 feet to 17 feet below the surface.
D. The Schroeder Lumber Company
1. : The site
In 1901, the John Schroeder Lumber Company, a Milwaukee corporation, purchased an existing mill on the waterfront in what is now Kreher Park and set up operations. It. produced wooden building products, shingles, boards, pilings, railroad ties and timber for framing and received and shipped lumber by rail.
As the depression took hold in 1930-31, shipments of lumber into the Ashland harbor ceased entirely and Schroeder Lumber neither received nor sent out lumber shipments. In 1936, the Ashland Daily Press reported that the mill had been closed for almost six years and that the Schroeder docks were being dismantled.
2. Wood treating activity at Schroeder Lumber
The Schroeder Lumber Company’s 1920 Amended Articles of Incorporation stated that one of the corporation’s purposes was to “manufacture and deal in preservative chemicals,” but the evidence does not show that the company had a wood treating facility or wood treating equipment at its Ashland site for which it would have needed preservative chemicals. No such facility or equipment is shown in either of the two existing historical aerial photographs of the lumber company site, one of which was taken between 1920 and 1925, and the second, in June 1939. Neither photograph shows any structure that would have been used in wood treatment, such as a boiler for heating the oil, tanks for the mixture of chemicals, an open tank filled with treating chemicals for immersing wood products or stacks of treated railroad ties or other wood products. No such structure was ever mentioned in Sanborn insurance maps of the site. (“[Sanborn] maps were designed to assist fire insurance agents in determining the degree of hazard associated with a particular property and therefore show the size, shape, and construction of dwellings, commercial buildings, and factories as well as fire walls, locations of windows and doors, sprinkler systems, and types of roofs.” (http://www.loc.gov/rr/ geogmap/sanborn/, visited Sept. 9, 2015)). The 1909 Sanborn maps describe the area that plaintiff alleges was used- for wood treatment as “filled ground,” “scattered lathe piles,” “shingles”, and “wood piles.” The 1923 and 1936 Sanborn maps do not show any wood treating equipment or structures, such as tanks for wood treating chemicals, hoists or drying platforms.
Nevertheless, two of plaintiffs experts, Walter Shields, a specialist in the reconstruction of historical contamination at industrial sites, and James Ebert, a forensic scientist, testified that such a facility could have existed. Each said that the 1939 photograph showed a “dark rectangle” near the mill that he thought was the size and shape of a wood-treating tank, with lighter-colored ground in the same area, suggesting a loss of vegetation in the area that might been the result of leakage of the wood-treating chemicals. Tr. trans., dkt. # 536, 2-A-37; dkt. # 549, 3-A-121-22. However, the Wisconsin Department of Natural Resources used ground-penetrating radar to look for the remnants of structures on the Schroeder Lumber site, along with electronic’ resistivity, which quantifies how strongly a given material opposes the flow of electronic current, and other tests. The department found no evidence of any'contaminants other than curbureted water-gas tar and' distillate oils, the same contaminants attributable to the gas plant discharges.
The primary trade association for companies engaged in wood preservation, the American Wood-Preserver’s Association, published annual lists of wood-treating plants, but never listed the Schroeder Lumber plant in Ashland. A publication of the association’s annual «proceedings includes maps.and tables that indicate the location of wood treating plants in the United States dating back to 1899. None of these list Schroeder Lumber as a wood treater in Ashland.
Schroeder never advertised wood treatment as being available in Ashland. Its 1915 list of lumber prices did not include any treated wood products. The Ashland City Directory did not list Schroeder Lumber in the classification of wood-treater. No records have been found showing any sale of treated wood products from Schroeder Lumber to a third party, any purchases of wood-treating chemicals by Schroeder Lumber or any purchase , of wood tar for treating purposes. The company produced railroad ties, but did not advertise them as dipped. (This is not surprising; according to the American Wood Preservers Association’s 1928 report, 99.95% of all railroad ties were pressure treated rather than dipped.) The Minneapolis, St. Paul and Sault St. Marie Railroad ran through Ashland and could have purchased railroad ties from Schroeder had they been available, but the railroad purchased its treated railroad ties from a facility in New Brighton, Minnesota.
In 1909, the lumber mill suffered a large fire. (Neither plaintiff nor the EPA.has identified soot or wood from this fire as contaminants of concern on the site.). In its coverage of the fire, The Ashland Daily Press made no mention of any potential loss of wood treating facilities or danger from wood treating chemicals. In 1911 and 1912, Schroeder Lumber representatives made a detailed report of the mill’s assets to the City of Ashland Board of Review. The report listed such items as postage stamps and one blind horse, along with planing and saw mills, but said nothing about any wóod treatment facilities or chemicals. In 1939, when the county was negotiating with Schroeder Lumber over the sale of most of the property on the premises, it prepared an inventory that did not include any equipment typically used in connection with a wood treating operation.
Defendants’ expert, Eugene Wengert, testified about the kind of equipment that would' be required for wood treating, such as chemical storage tanks, a connection to the electrical generator to heat the chemicals in the tank, a handling system for getting the wood out of the tank and onto railroad cars, an incising machine for putting holes into the wood to allow the chemicals to penetrate, a place to cut the tie down to the right length and an “S iron” to produce a big piece of metal in the shape of an S to be pounded into the ends of the ties to prevent them from splitting. Wengert saw no evidence of any equipment of this kind, with the exception of a timber saw that could be used to cut ties to the right length. In his opinion, the climate in Ashland did not suggest that a lumber company would want to treat wood there becausé the wet and’ cold climate would extend the process for drying the wood before it could be treated.
When asked about the particular area in the lumber yard that others had thought might be the site of the treating tank, Wengert said it would be “a pretty terrible location in many respects,” tr. trans., dkt. # 553, at 7-P-92. The area was crowded; the tank would have been near the convey- or; and the proximity of the tank and conveyor would have made it more difficult to load the lumber onto rail cars. Also, it would not have been safe to build a conveyor carrying lumber close to wood treating chemicals, which are flammable.
Plaintiff’s expert, Thomas Beck, a person with extensiye experience in the forest-products industry, was unable to say where a wood treating tank might have been located on the Schroeder site. He testified that neither the cold temperatures nor moist climate of Ashland would be a hindrance to wood treating, as defendant’s witness had testified, because wood treating goes on in the state of Washington-and it too has a moist climate. It was Beck’s opinion that Schroeder did treat wood, but he based this on three possible factors, not actual ones: (1) the company had a motivation to treat wood to increase its sales and profits;, (2) the industry recognized the advantages of open tank treating at the time; and (3) wood treating would have been consistent'with the company’s operations, because the company made products that could have been treated. Beck never saw any photographs of a wood treating operation at Schroeder and he had no evidence that any railroad was purchasing railroad ties that had been hand-dipped rather than pressure-treated at the time Schroeder was in operation.
Neither the Wisconsin Department of Natural Resources nor the EPA has found any evidence, of wood treating at the former Schroeder Lumber Company site in Ashland.
E. Ashland-Comity
When the Schroeder Lumber Company failed to pay any property taxes between at least 1931 and 1938, Ashland County filed a tax deed for the property in June 1939. After some court proceedings, Schroeder reached an agreement with the County in November 1939 that allowed the company to buy all the property on the premises, including a refuse burner, retaining for the county the land, all docks, pilings and subsurface materials. In return, the company paid the county $8000 and signed a quitclaim deed to the real property, docks, etc., in December 1939. The County held the land until about 1942, when it turned over the title to the City. During that time,'neither the County nor anyone else operated the mill or the refuse burner.
At some time' before February 1940, Schroeder Lumber Company sold the lumber mill’s refuse burner to Ben Mistroll of the Chicago Tubing’ and Metals Company of Ashland, which arranged for its demolition that month. About a year after the refuse burner was demolished, the remaining Schroeder buildings on the site were razed by a local firm. The County had no role in the demolition of the buildings, burner or smokestacks it had sold back to Schroeder- Lumber. The agencies investigating the Ashland site have not discovered any evidence that the demolition caused any increase in the cost of the remedy required at the site. The refuse burner powered the boilers at the lumber yard; it did not produce wood tar.'
F. City of Ashland
1. The wastewater. treatment plant in , Kreher Park
In 1951, the City of Ashland built a wastewater treatment facility in Kreher Park on the -former Schroeder Lumber property, in close proximity to Chequamegon Bay. It enlarged- the facility in the early 1970s, after thé Wisconsin Department of Natural Resources advised it that improvements were necessary to prevent the inflow of lake' water into the facility and the discharge of effluent that had received oniy primary treatment.
Greeley & Hansen, a Chicago engineering firm specializing in municipal waste-water treatment plants, developed the plans for the treatment plant’s initial construction and for its expansion. Before construction of the .original plant began, during construction and again before construction began for the expansion, the engineers took soil borings.from the area in which the construction was to take place. None of these borings disclosed any tar, oil or petroleum substance. According to Ronald Bizzarri, a Greeley & Hansen employee who worked on the pre-construction plan for the :plant expansion, an encounter with any tar, oil or petroleum substance “would - have, been significant” and “would certainly have been noted.” Bizzarri dep., dkt. # 126, at 113, : 115. Bizzarri was not aware of any contemporaneous report of oil, petroleum, tar dr similar product that -was excavated or moved about during the 1970 expansion, id. at-117. It was his experience that “if-contractors run into substances of that sort [oil, petroleum or tar], they would stop work and wait for direction from the owner as to how to handle that material.” Id. at 117-188. He observed that contractors did this regularly. In fact, their contracts provided extra compensation “because of the difficulties that, would be caused by trying to deal with” such substances. Id. at 117. The engineers involved in the planning for the wastewater plant were aware of the waste tar dump and had aligned the facility’s influent pumps to avoid it. Id. at-112.
Much of the construction of the treatment -plant and - its expansion took place below water level and required the contractors to construct walls to prevent water from flowing in or out of the plant into the lake while the soils were being dewatered, In the original construction, the walls were clay core. They rose above the lake level and rested on the hard clay subsurface under the wood waste and other “unsuitable materials” in the lake bed, such as decayed wood, organic material, topsoil and other earth that would not support the weight of the new construction. For the second round of construction' in the -1970s, the contractor built a wall of steel' sheeting to enclose the excavation for the new tank area and provide a dry’ area in which to construct the new additions to the plant. Until that wall was completed, on at least 89 days, lake water came into the excavatidn area and had to be removed, along with plant effluent, that is, water that had received primary treatment. On one occasion, an existing pipe broke during construction and for less than one day, sewage flowed into the lake before the flow could -be re-routed to the other influent pipe. Gerulat dep., dkt. # 132, at 11617. Daily reports prepared during this time made no reference to encounters with any tar or oily substances. Id. at. 158. Throughout the expansion work, the director of the wastewater treatment plant took regular water quality samples of lake water near the construction site and never notified the resident engineer of any discharges of pollutants. Id. at 158-61.
Although the City moved its treatment plant off site- in .the early • 1990s, because soil borings showed encroaching-contamination, testing conducted in 2014-15 by one of the City’s expert witnesses, Robert Karls, showed no contamination at the treatment plant site itself. Moreover, Karls took samples of the water that had accumulated in the basement of the abandoned wastewater treatment plant after the city turned off the sump pumps that had been removing accumulated water from the basement. Testing of the samples showed no detectable contaminants in the water.
2. The open ditch
Excavations and Sanborn maps show that an open ditch ran ip a generally north-south fropi the base of the bluff to the bay during the first part of the twentieth century, but was not in, evidence by 1939. Discharges from the manufactured gas. plant ran through this ditch into the bay, until the 12-inch clay pipe running from the gas plant lost its connection to the east-west pipes leading to the ditch, presumably before 1939. The location of the ditch suggests that it , was built to bypass the Schroeder Lumber Company to avoid staining any lumber products. Plaintiff introduced no evidence that during, the time the ditch existed the City or the County or any private company other than the gas plant ever used it to discharge any waste products. It appears to have been sited so as to avoid the Schroeder Lumber Company work area to avoid the possibility of staining the company’s lumber products.
3. The city’s sewer system,
The City’s sanitary sewer system never discharged through the park, as confirmed in a 1949 report from Greeley & Hansen, which shows no sanitary sewer discharging through the site between its eastern boundary, Prentice Street, and its western boundary, Ellis Street. The system was never connected to the 12-inch clay pipe installed by the manufactured gas plant to run down the.base of the.ravine. Such.a connection would not have been possible in the absence of a lift station because the clay pipe was 11 feet below the elevation of the sewer and no one has ever found any evidence of a lift station in that area or of any connection between the sewer and any other clay pipe originating at the gas plant. The City did construct a lift station for the sewer system, but it did so for other purposes. The sewer was connected to the gas plant for roof water runoff, but was never intended to take tar waste from the plant.
Neither the construction of the sewer system nor later repairs in Kreher Park contributed to thé need for remediation. Workers involved in the lift construction or the sewer repairs did not move any soil, but replaced it in the construction holes and tests of the areas have shown an absence of any contaminants.
4. Waste and refuse dumping and oil changes
Although the City tolerated allowed the dumping of refuse and waste within the park for some years, no investigator has ever determined that the dumping contributed any hazardous waste or other contamination requiring remediation. This is true as well for the City’s practice of allowing its employees to. use the park for changing oil.
5: The seep and waste tar dump
' The park includes a historically low spot known variously as the waste tar area, the coal tar dump or the waste tar dump. The area was identified officially in 1951 by employees of Greeley & Hansen, the engineering firm employed by the City to design the original wastewater treatment plant. South of the waste tar dump at the base of the bluff is an area referred to as “the seep” because it oozed tar waste before remediation efforts were first taken in 2002. From the excavation of the area, it appears that it was been created by the outflow from a 12-inch clay tile pipe carrying tarry wastes from the gas plant. After that pipe was identified and traced to the gas plant in 2001, plaintiff worked with the Wisconsin Department of Natural Resources on a remediation plan for the area. In the spring of 2002, plaintiff began excavating contaminated soil from the area and putting in clean fill. Its contractors smashed the clay tile piping at a number of places in the filled ravine to reduce its capacity to carry any flow. They also added another extraction well at the exit from the gas plant site.that could be connected to a pump- and treatment plant designed to minimize ongoing emulsions that might still be oozing north from the gas plant site to the seep area. With these actions, the seep dried up.
Beyond the seep to the north is an area the parties call an expansion or transition zone, and beyond that is the waste tar dump, which, as noted, is at an elevation of 600 feet, five to six feet lower than the seep area. The dump is far more heavily contaminated than the seep area, despite being farther from the gas plant. The concentrations in the transition zone are as much as 50 times lower than those in the tar dump, which is in shallow soils, Excavation in 2014 revealed tar contamination extending from the waste tar dump out to the former Pulp Hoist Road at the north end of Kreher Park. At present, the remediation plan for the park is to leave the contamination in place and place a cap on the entire park area.
A drawing prepared by Greeley & Hansen in 1951 shows a 12-inch corrugated pipe extending from the tar dump, under Marina Drive and out to the bay, through a culvert. A 2011 photograph of the area does not show any road culvert in the place shown in the drawings and no other evidence of its existence has been found, although it is possible that one was built and later removed in the intervening 40 years. The dump is at an elevation of 600 feet, MSL (mean sea level), which is lower than the designated target water level of Lake. Superior at 601.7 feet. The culvert was to be built under a road that had an elevation between 604 and 605 feet. Although the bottom of the culvert would have béen at least two feet below the road (taking into consideration the 12-inch diameter of the culvert and the material between it and the road surface), it would still have been two to three feet higher than the 600-foot elevation of the dump. The more likely interpretation of the drawing is that the pipe and culvert were intended for the purpose of diverting storm water from the road and draining it into the bay.
Two existing storm water culverts can be found today, near the original treatment plant. Recent examination of the culverts disclosed no surface contamination, which was not surprising, given the absence of any surface contamination in the area. Constructing these culverts would not have dispersed any contamination.
G. Contaminant Sampling
As part of plaintiffs remedial investigation and other cleanup efforts and in preparation for trial, various scientists took approximately 1,281 samples of the soil, sediment and other materials from 533 locations on the Ashland site to assist them in determining the sources of the polycyclic aromatic hydrocarbons (PAHs) contaminating the Upper Bluff area, Kreher Park and Chequamegon Bay. Paul Boehm, an expert in forensic and environmental chemistry, testified on behalf of plaintiff that the contamination in-the park and bay came 'from a source other than the gas plant discharges found in the Upper Bluff. Using three different scientific techniques, (1) ratio analyses; (2) gas chromatography; and (3) principal component analysis, he concluded that measurable chemical differences could be seen between the PAHs found in the Upper Bluff and those found in Kreher Park. This led him to conclude that the source of the contaminants found in the bay was the park and specifically, the waste tar dump in the’ park. Such a conclusion is hard to square with the evidence that a source of tar contamination other than the manufactured gas plant has never been found in the Upper Bluff, Kreher Park or Chequamegon Bay.
Defendants’ experts identified decisions that Boehm made that could have affected his findings, such as those he made in his double ratio analyses: not using samples below a concentration of 1000 parts per million of PAHs; not using data points from the 2014 and 2015 sampling; and excluding certain low concentration samples. When defendants’ experts used samples with lower concentration cutoffs and the 2014 and 2015 data, they found significant overlap in the PAHs ratios observable among the samples from all three areas: the Upper Bluff, Kreher Park and Chequamegon Bay.
In his principal component analysis, Boehm chose not to use a particular cluster of Upper Bluff samples that amounted to approximately 45, or 41% of the total Upper Bluff samples analyzed. These samples were high concentration samples and were taken from a small area that investigators believed could be the source of the contamination. Testing done by Dennis Helsel, one of defendants’ experts, showed a statistical difference between this cluster and the remainder of the samples from the Upper Bluff and Kreher Park. Leaving out the cluster of Upper Bluff samples eliminated the differences observed by Boehm between the Upper Bluff and Kreher Park.
Another one of defendant’s witnesses, Alan Jeffrey, an expert in geochemistry, found that some of the soil samples showed PAH components inconsistent with carbureted water gas; in his opinion, these could be explained by the different feed-stocks used at the manufactured gas plant, such as the coal used as feedstock in 1917 and 1918. Jeffrey’s testing showed that using a fuller data set for the gas chomatography testing revealed an abundance of petroleum in both areas. Additionally, Jeffrey testified that some of Boehm’s conclusions could be explained by inter-laboratory variability, which results from small variations in laboratory techniques and in technician discretion in categorization.
Another one of defendants’ experts, Guy Patrick, explained that at the Ashland site, the contaminants have left behind residual PAHs and that the phenol concentrations in these PAHs have decreased as the hydrocarbons moved from the gas plant toward the bay. This was shown by a comparison of the phenol concentration to the concentration of benzo[a]pyrene, which is a large molecule that is unlikely to have dissolved out of the PAHs during the decades since they were released into the environment. If, as Patrick found when comparing the samples, the proportion of the phenol concentration to the benzo[a]pyrene concentration is consistent, it is likely that the materials are the same, even if the concentrations themselves have increased or decreased. He found from his examination a continuous path'of PAHs of similar concentration from the Upper Bluff to the -bay, suggesting that the source was in the Upper Bluff. He concluded that the phenols present in Kreher Park can be attributed to the coal gas process used at the gas plant in 1917 and 1918, a process that can produce phenols, sometimes at high rates.
In Boehm’s data set of samples of dense non-aqueous phase liquids, ■ which sink in water, and low density non-aqueous phase liquids, which float in water, he had more low density non-aqueous phase liquid samples from Kreher Park than from the Upper Bluff. 'The samples'-of low density non-aqueous phase liquids were apt to have had nearly constant contact with water, making it likely that although the difference that Boehm observed between samples appeared to be geographic, in fact, it is attributable to the loss of soluble compounds from the sample contaminants.
Boehm also did a chemical mass balancing test that supported his hypothesis that the contaminants had two separate sources. He compared this test to the act of pouring two liquid-filled beakers into one bowl and asking what proportion of liquid comes from each beaker. So long as the two beakers contain different, identifiable liquids, the chemical mass balancing test can be helpful. If, however, the two site areas (or “beakers”) have no statistical difference, as in this case, the test will not identify the amount of contamination from the bay (or bowl) attributable to each source. In. this case, no one has identified a likely second source of contamination. At this point, I must assume that none exists, and that Boehm’s testing was flawed.
Boehm’s testing fails to support a finding that the contamination in the bay has two sources. The work of the City’s experts shows that any variation between the Upper Bluff and Kreher Park is insignificant and that any variation that is present can be explained by the loss of soluble compounds through weathering and by variations in laboratory techniques and technician discretion in categorization.
H. Sources and Transport of Contamination
Kurt Herman, an environmental engineer and registered geologist, testified for plaintiff that the contamination he identified in the bay came primarily -from sources in Kreher Park rather than from the manufactured gas plant. I am placing no weight on his testimony because it was based on assumptions that later testimony showed to be unsupported: first, that wood preservatives had been used by the Schroeder Lumber Company while it occupied the site; second, that the gas plant had either sold or used most of its tar wastes; third, that the City had drained the waste tar dump into, the bay; fourth, that the City had contributed large disposals of PAHs during the construction and expansion of its wastewater treatment plant in the 1950’s and 1970’s; and fifth, that “field observations confirm that most [manufactured gas plant] releases were contained within the Upper Bluff.” Tr. trans., dkt. # 542, at 3-P-19.
In addition, Herman testified that some of the PAHs found in the bay were from contact wastewater, that is, wastewater that had actually come into contact with the raw gas and was to be distinguished from wastewater that had not had such contact. This opinion would have merit only if any evidence showed that gas wastes from the manufactured gas plant had been discharged through ’ the City’s sewer system and there is no evidence to that effect. It was also Herman’s opinion that the gas plant had treated its wastewater in its tar well, which he assumed was used for primary treatment of the waste-water. He estimated that the plant had discharged 1,500,000 million to 15,500,000 gallons of wastewater,, which would have contained about 13,000 gallons of tar. He found this amount of tar much smaller than the total amount of non-aqueous phase liquids in the bay and concluded that there must be other significant sources of contamination. ' His conclusion does not stand up in the face of the testimony and visual evidence of the enormous quantities of tar waste on the site that have either been excavated or. capped.
In Herman’s opinion, the introduction of a new mass of contamination from another source is the only satisfying explanation for the much greater concentration of tar in the waste tar dump than in the seep area, when the dump is farther from the gas plant and the concentrations in the dump are shallow. The problem with this testimony is that Herman’s only candidate for the source of significantly greater concentrations of PAHs is a top-down release of preservatives from a wood treating operation that did not exist. In addition, he listed the low permeability of the Ashland soils as a second reason for thinking that gas plant wastes would not account for all the contamination in the dump, but did not take into consideration the probable permeability ■ of ' the filled ravine ‘ or the highly permeable wood waste underlying Kreher Park.
Finally, Herman estimated the proportion of PAHs that would have been the consequence of urban runoff. In doing so, however, he relied on estimates that had been made for larger urban areas than Ashland’s and on the response filed by Brian Knapp on behalf of the City, which I am not considering, for the reasons set out below.
I. Credibility Determinations
In finding facts, I have rejected certain evidence proposed by plaintiff, starting with the statements in an April 2009 response from then-City Administrator Brian Knapp to a § 104(e) request from the EPA. The City corrected Knapp’s statements in a response authorized by the Ashland City Council on August 11, 2014, Exh. CI5378, after acquiring more information. Among other things, the City determined that “all of the sewage collected by the City in the central part of the City was carried by sewer mains to- a discharge point in Stuntz Avenue,” [east of the Superfund site]; the “area adjacent to the bay had no sewer service”; and “the open ditch was a private facility which carried NSP’s waste to the Bay.” Ltr. to EPA, dkt. # 285-1, at 3. The evidence produced at trial supported the City Council’s version of the facts relating to the sewer system and the wastewater treatment plant construction. In light of that evidence, I find the City Council’s response more credible than that of Knapp, given his far more limited knowledge of the facts in dispute at the time of his response.
I am placing no weight on the deposition testimony of Gordon Parent or Ray Parent about culverts built in the early 30’s or possibly the late 40’s On what was. then known as Pulp Hoist Road. Plaintiff wants to use this testimony to show that-the City tried to drain the waste tar dump into the bay, but it is not sufficiently reliable. The very age of the memories raises questions about their accuracy, as does the manner in which plaintiff obtained the information. Plaintiff sent private investigators into the Ashland area to interview elderly residents about their recollection of certain things, including culverts, lumber treating facilities at the Schroeder Lumber Company and soil disposal in the 1930s, 40s and 50s. It declined' to release most of its results to defendants or to disclose the interviewers’ techniques. As a result, it is not possible to know what techniques of questioning were used, what suggestions were made to the persons being questioned or how many other residents had similar or contradictory memories of the same subject matter. In this case, both Gordon and -Ray Parent had different memories about when the culverts might have been constructed and exactly where they were located, which is hardly surprising given the passage of time. In addition,' it is not possible to tell from Gordon Parent’s testimony whether the City or the Consolidated Pulp Hoist company built the culverts he' remembers, exactly where they-were located-or their purpose. '
Finally, I am giving no weight to the observations by Norman Peterson and City employees David Wos'epka and Robert Klamerus, who testified that they had seen oily, foul-smelling substances and, in Peterson’s case, dead fish in the' water, during various phases of the initial construction or expansion of the wastewater treatment plant or while repairing water mains. Their testimony suggests that they might have been observing tar waste, but it is far from definitive and there is no scientific support for the observation. Moreover, it is of questionable value in light of the repeated soil borings that showed no contamination in the areas in which these men claim to have observed the excavated .soils.
OPINION
A. Contribution Actions and Declaratory Judgments
Plaintiff Northern States Power Company brought this suit in an effort to obtain contribution from defendants for the costs it has incurred and will continue to incur under its 2012 consent decree with the EPA resolving the cleanup of the Upper Bluff and Kreher Park. Plaintiff acknowledges that it alone is responsible -for the cleanup costs for the Bluff, but it seeks contribution for the costs incurred in cleaning up Kreher Park. In addition, it asks for a declaratory judgment covering the.future costs it expects to incur in cleaning up the sediments in ChequamegonBay. '
1. Legal standard
Under CERCLA, parties who have settled their liability with the EPA may bring contribution actions under 42 U.S.C. § 9613(f). In such actions, “the court shall enter a declaratory judgment on liability for response costs or damages that will be binding on any subsequent action or. actions to recover further response costs or damages.” § 9613(g)(2).
To prevail in a contribution action, plaintiff must show that defendants are “covered persons”, under § 9607 and must also prove the appropriate equitable share or response costs that defendants should incur. Environmental Transportation Systems, Inc. v. ENSCO, Inc., 969 F.2d 503, 506 (7th Cir.1992). “Thus, [§ 9613(f)] envisions a two-part inquiry: Fust, the court must determine whether the defendant is ‘liable’ under CERCLA [§ 9607]; Second, the court must allocate response costs among liable parties in an equitable manner____ The party seeking contribution bears the burden of proof under both prongs of the court’s inquiry.” Goodrich Corp. v. Town of Middlebury, 311 F.3d 154, 168 (2d Cir.2002) (citations omitted) (cited approvingly by NCR Corp. v. George A. Whiting Paper Co., 768 F.3d 682, 690 (7th Cir.2014)). See also PCS Nitrogen Inc. v. Ashley II of Charleston LLC, 714 F.3d 161, 176 (4th Cir.2013); Elementis Chromium L.P. v. Coastal States Petroleum Co., 450 F.3d 607, 612 (5th Cir.2006).
Under § 9613(f), plaintiffs initial burden is to establish that either or both defendants are liable or potentially liable under 42 U.S.C. § 9607(a). To do so, plaintiff must prove the following four elements by the preponderance of the evidence: (1). the site in question is a “facility,” as defined by 42 U.S.C. § 9601(9); (2) there has been a release or threatened release of hazardous substances at the “facility”; (3) the defendant is a “potentially responsible person” for the spill as defined by CERCLA; and (4) the plaintiff incurred response costs responding to that release or threatened release. Kalamazoo River Study Group v. Rockwell International, 171 F.3d 1065, 1068 (6th Cir.1999). Some courts add a fifth element, that the response costs conform to the National Contingency Plan. PCS Nitrogen Inc., 714 F.3d at 167-68; ABB Industrial Systems, Inc. v. Prime Tech., Inc., 120 F.3d 351, 356 (2d Cir.1997); Environmental Transportation Systems, 969 F.2d at 506. Defendants do not deny that plaintiff could establish elements (1), (2) and(4), all of which were conceded by plaintiff in its agreement with the EPA, so the only questions are whether defendants are “covered persons” under § 9607 and whether they áre persons responsible or potentially responsible for the contamination at the Ashland site.
If plaintiff can' prove that either or both defendants are ’covered persons and responsible for some part of the contamination, they will be entitled to recover from them under CERCLA. NCR Corp., 768 F.3d at 686. CERCLA provides two ways to pursue contribution, either under § 9607(a) or § 9613(f), depending on the procedural posture of the claim. Id. at 690. Plaintiffs who incur necessary response costs that are not required by court order or settlement may sue to recover them under § 9607. Plaintiffs who incur response, costs that are required under a court order or settlement may seek contribution toward those costs from other responsible parties under § 9613. Id.; Chubb Custom Insurance Co. v. Space Systems/Loral, Inc., 710 F.3d 946, 963-64 (9th Cir.2013), Because plaintiff’s 2012 consent decree- was court-ordered, plaintiff had no choice but to proceed under § 9613(f)(1) when bringing its suit against defendants for contribution. toward the costs of remediating Kreher Park. Id.
The 2012 consent decree requires ‘plaintiff to clean up contamination in the Upper Bluff and Kreher Park as part of Phase 1 of the remediation. Because much of this effort remains to be doné, a declaratory judgment will be éntered with respect to the remaining costs of this phase of the work. United States v. Davis, 261 F.3d 1, 46 (1st Cir.2001) (holding that declaratory judgment appropriate as to future costs under § 9613(f), though discretionary); City of Gary, Indiana v. Shafer, 683 F.Supp.2d 836, 860-62 (ND.Ind.2010) (same); Appleton Papers, Inc. v. George A. Whiting Paper Co., 572 F.Supp.2d 1034, 1046 (E.D.Wis.2008) (same). But see Reichhold Chemicals, Inc. v. Textron, Inc., 888 F.Supp. 1116, 1124 (ND.Fla.1995) (“By its explicit language, [§ 9613(g)(2)] applies only to cost recovery actions [under § 9607].”).
Although the work .on Phase 2 (cleanup of the sediment in Chequamegon Bay) has ■not begun’ and is not part of the consent decree, plaintiff has requested adéclaratory judgment for those cleanup costs as well. In fact, it argues that this declaration is mandatory under § 9613(g)(2), which states that “[i]n any such action described in this subsection [actions under § 9607], the court shall enter a declaratory judgment on liability for response costs or damages that will be binding on any subsequent action or actions to recover further response costs or damages.” Plaintiffs request raises an initial question: under what provision of CERCLA does the request arise? With no court order or settlement as to these costs, the request might arise under § 9607. However, plaintiff requests an “allocation” as to these costs, which is available only under § 9613. Moreover, plaintiff treated the court trial as one under § 9613 alone, without objection from defendants, and its agreements with the EPA clearly contemplate that it will take on the bay’s cleanup, suggesting that the question is inevitably one of contribution. Defendants do not argue that a declaratory judgment is inappropriate or that it should be entered under any provision other than § 9613.
A declaratory judgment is available only when a case or controversy exists, GenGorp, Inc. v. Olin Corp., 390 F.3d 433, 451 (6th Cir.2004) (“Entry of a declaratory judgment, however,- cannot be fully mandatory” because “casé or' controversy” must exist.), and the planned Phase - 2 cleanup is a concrete controversy under 28 U.S.C. § 2201. Cf. New York v. Solvent Chemical Co., 664 F.3d 22, 25 (2d Cir.2011) (“[T]his Circuit has recognized that the ‘proper remedy for future response costs is not a present lump-sum payment of anticipated expenses but instead a declaratory judgment award dividing future response costs among responsible parties.’ ”) (quoting Goodrich Corp., 311 F.3d at 175); Tosco Corp. v. Koch Industries, Inc., 216 F.3d 886, 897 (10th Cir.2000) (“We believe where, as here, a responsible party chooses to go to trial and future response costs are likely to be incurred, but the exact amount remains unknown, a judgment on proportional liability is an appropriate remedy.”). Furthermore, because environmental litigation like .this is “complex, lengthy, and expensive, .... [t]he costs and time involved in relitigating issues as complex as these wh[en] new costs are incurred would be massive and wasteful.” Boeing Co. v. Cascade Corp., 207 F.3d 1177, 1191 (9th Cir.2000). Accordingly, a declaratory judgment will be entered on the question of allocation between plaintiff and defendants under § 9613(f)(1) for both Phase 1 and Phase 2 costs.
2. Covered persons
The obligations imposed by CERCLA apply only to “covered persons,” as defined in § 9607. The term includes any “owner and operator of a vessel or facility”; any person who at the time of disposal of any hazardous substance owned or operated any facility at which such hazardous sub1 stances were disposed of’; and “any person who ... arranged for disposal or treatment, or transport of any hazardous substances ...” § 9607. Plaintiff contends that because the County owned the former Schroeder Lumber Company site (now part of Kreher Park) for a few years in the early 1940s, it was an “owner” of a “facility” “at the time of .disposal of a[ ] hazardous substance.” As to the City, plaintiff contends that it is and has been since the 1940s an “owner or operator of a facility (Kreher Park) or, alternatively, that it owned or operated a facility “at the time of disposal of [a] hazardous substance.”
The p