Environmental Cleanup & Restoration: CASE STUDY

Foundry Cove in the Hudson River, New York

Adapted from: Foundry Cove: History of a Polluted Site and Its Restoration by Jeffrey Levinton and Josepha Kurdziel, SUNY Stony Brook: http://life.bio.sunysb.edu/marinebio/foundryframe.html

History of Metal Pollution in Foundry Cove

Foundry Cove is situated in the Village of Cold Spring, in Putnam County, NY, approximately 54 miles north of Battery Park, NYC (Fig 1). Foundry Cove is a well defined inlet of the Hudson River. This cove contains freshwater marshes and mud flats and is tidally influenced (Fig. 2). A railroad trestle divides the cove into east and west; flow from the Hudson River into West Foundry Cove is unrestricted, while flow into East Foundry Cove is restricted to a 65 foot wide passage under the railroad trestle. The water is generally fresh but the salinity may reach 2-6 parts per thousand in periods of low fresh water flow.

The Marathon Battery Company facility in Cold Spring, NY, was located near Foundry Cove. The plant was constructed in 1952 by the U.S. Army Corps. From 1952 through 1979 this facility manufactured nickel-cadmium (Ni-Cd) batteries, initially for military contracts. The plant was later owned by several private companies (Sonotone Corporation, Clevite Corporation, and later Gould Incorporated) that produced batteries for commercial use.

Fig. 1. Area map showing location of Foundry Cove on the Hudson River.

The battery manufacturing process requires the use of concentrated metal nitrate solutions that result in dilute waste solutions and metal precipitates. Both nickel and cadmium were used in large quantities; for a brief time, cobalt was used as an additive. The plant effluent was a fine suspension of nickel and cadmium hydroxides, in a pH range of 12 - 14, at a volumetric flow rate averaging 50 - 100 gallons/minute. The effluent usually contained from 10 - 100's ml/l suspended Ni and Cd hydroxides, depending upon production values. The total waste water output ranged from 100,000 - 200,000 gallons/day (Klerks 1987).

Fig. 2. Foundry Cove, foreground, with Constitution Marsh Audubon Sanctuary at Center Left. South Cove is at the Rear

Waste water from the manufacturing process was initially discharged into the Hudson River through the Cold Spring sewer system, but approximately 10% was discharged into a bypass system emptying directly into East Foundry Cove. In 1965, the NY State Department of Health concluded that the village of Cold Spring's sewage treatment system could not handle the plant's industrial waste water; the battery company was ordered to disconnect from the sewer system after which all waste was discharged directly into East Foundry Cove. After maximum discharge limits were set in 1971, waste waters were again discharged via the sewer system into the Hudson River. During manufacturing operations, a total of 179,105 kg of cadmium hydroxide was discharged. Of this amount, 51,004 kg of particulate Cd, and 1,569 kg of soluble cadmium were discharged directly into East Foundry cove; the remainder was discharged in the Hudson River (Klerks 1987). This earned Foundry Cove the dubious distinction of being "the most cadmium polluted site in the world".

Fig. 3. Distribution of cadmium in surface sediments in East Foundry Cove in 1974

Fig. 4. Distribution of cadmium in surface sediments in East Foundry Cove in 1983

In 1971, state officials detected high cadmium levels in East Foundry Cove in violation of the Clean Water Act of 1970. A civil law suit filed against Marathon Battery Company resulted in the dredging of all sediment exceeding 900 mg/g Cd based on wet weight. In 1972 - 1973, this dredging removed 10% of Cd released into Foundry Cove. These contaminated sediments (90,000 m3) were buried in a clay-lined, underground vault on the plant property. Extremely high Cd and Ni concentrations were found in sediments in subsequent years, despite the dredging; up to 50,000 and 11,000 g Cd and Ni per g dry weight sediment (Hazen & Kneip 1979, Occhiogrosso et al. 1979). In 1975, about 30% of the cove still had surface Cd levels in excess of 1000 ppm. In 1979, the Marathon company closed the plant and relocated. Merchandise Dynamics purchased the plant in 1980 for use as a book storage facility. In that same year Congress enacted the comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) to address the cleanup of the nation's hazardous waste sites.

Fig 5. Timeline for the history of cadmium pollution of Foundry Cove

Fig 6. Timeline for the cleanup of Foundry Cove

Investigations into the former battery plant began again in 1983 when the NY State Department of Environmental Conservation (DEC) sampled soils on the property as well as marsh sediments in Foundry Cove. At that time, only 8% of the total area of Foundry Cove contained surface sediment Cd levels in excess of 1000 g/g dry weight; the values for Ni and Co were much lower, but spatially strongly correlated with the Cd values. A remediation plan was designed and managed by the U.S. Environmental Protection Agency (EPA). Due to the severity and extent of cadmium contamination, the EPA added Foundry Cove to the Superfund program, a national priorities list of abandoned hazardous waste sites requiring long-term cleanup. In 1986, the EPA divided the site into three geographical areas and following a period of public comment, a remediation plan for the three areas was drafted. Dredging and excavation activities began in 1992 and were completed in 1994. Marsh replanting efforts and the demolishing of plant buildings are ongoing.

Fig. 7. Cadmium profiles with depth in the sediment, as measured in 1983

Unlike some toxic compounds, e.g., PCB's, which can be broken down by natural processes or through remediation techniques, metals like cadmium cannot be degraded. The natural reduction of surface Cd concentrations in Foundry Cove from 1971 to 1983 (prior to EPA's excavation and dredging activities) must be due either to: (1) deposition of new sediment, (2) transport of metals out of the cove, (3) redistribution of sediments within the cove, or (4) some combination of these processes. Depth profiles were established in order to determine if burial had occurred. At the most polluted sites sampled in 1983 (those closest to the battery plant outfall pipe [sites A - E]), there was a subsurface peak in Cd concentration, suggesting that the polluted sediment was being covered by new sediment (Knutson et al. 1987). Movement of Cd out of the cove may also account for a portion of the decreased Cd concentrations. Feeding processes of invertebrates within the cove will also affect metal distributions in surface sediments as well as transfer of metals to other parts of the Hudson River ecosystem.

Animal Feeding and Metal Uptake and Bioconcentration

The benthic oligochaete, Limnodrilus hoffmeisteri (Fig. 8), inhabiting cadmium-, nickel-, and cobalt-polluted Foundry Cove has evolved resistance to these metals (Klerks & Levinton 1989a,b). In survival experiments in which oligochaetes from Foundry Cove and South Cove (control site) were exposed to sediment with highly elevated metal levels, Foundry Cove worms survived the 28 day exposure while control worms did not. Second generation offspring of Foundry Cove worms reared in clean sediment also possess metal resistance.

Fig 8. Cluster of Limnodrilus hoffmeisteri

An increased resistance to a metal can be achieved by a reduced accumulation of the pollutant. Reduced uptake rates have been reported for a number of different organisms, e.g., bacteria, algae, annelids, and fish (references cited in Klerks & Bartholomew 1991). But several studies comparing metal accumulation in populations differing in resistance did not find reduced uptake rates in resistant populations; some authors even found increased uptake rates. If resistant individuals have increased metal uptake rates, then they must possess some physiological mechanism for metal detoxification. The ultimate research goal is to determine the mechanism(s) by which resistance has evolved in Limnodrilus hoffmeisteri inhabiting Foundry Cove.

The first objective towards this ultimate goal was to determine whether resistant worms accumulate less Cd than their sensitive conspecifics from South Cove. Secondly, if metal resistant and sensitive individuals have similar uptake rates, is resistance achieved by an increased cadmium detoxification mechanism? The following experiments determined metal uptake rates in resistant and control populations using the radioisotope 109Cd in water and Foundry Cove sediments.

Limnodrilus hoffmeisteri (Annelida, Oligochaeta, Tubificidae) is a simultaneous hermaphrodite, which reproduces sexually by cross-fertilization. This oligochaete is a deposit-feeder and the most common macrobenthic species at both coves (Klerks & Levinton 1989b). Worms and sediment were collected by Ekman grab from both Foundry and South coves. Worms were then sorted from the >500 m fraction. Laboratory cultures were set up in polystyrene dishes with 1 cm layer of sediment and 9 cm of continuously aerated Hudson River water. Sediments collected from the two coves were sieved to < 500 m, boiled, washed with filtered Hudson River water, then frozen and thawed shortly before use. Ground fish food flakes were added to dishes once per week and cultures were kept at 24C under a 13:11 light:dark cycle.

It is likely that resistance in Limnodrilus in Foundry Cove evolved mainly in response to Cd pollution at this site, rather than Ni or Co pollution. This is based on several observations: these worms accumulate much more Cd than Ni, but they do not accumulate Co at all, and Cd is generally much more toxic than Ni (Khangarot & Ray 1987). This investigation into the mechanism(s) underlying the resistance in Limnodrilus in Foundry Cove thus focused on the fate of Cd in these worms.

Cadmium Uptake from Sediments

Cadmium accumulation in L. hoffmeisteri from Foundry Cove and South Cove was determined by exposing worms to sediment with different metal levels. These exposure experiments were set up as a bioassay for comparison of sediment toxicity among populations (Klerks & Bartholomew 1991). Three replicates of 10 worms each from stock cultures were exposed to 6 different sediment metal levels (ranging from 15 to 34,000 g/ g dry sediment) for 28 days. Worms that survived the exposure and laboratory stock worms were collected for metal analyses (6-10 worms per replicate). These worms were kept in filtered river water for 2 days to exclude gut contents from analyses. The worms were then rinsed in distilled water, pooled by replicate, and frozen.

Worms from each replicate were then thawed, transferred to a beaker and dried. Ultrex grade nitric acid (2 ml) was then added to each beaker, refluxed for 2-4 h at 120C, and evaporated. This procedure was repeated twice, after which each sample was brought to 5ml volume with addition of nitric acid. Cadmium concentrations were determined with a Perkin Elmer 4000 graphite furnace atomic absorption spectrophotometer. Blanks, tissue samples, and standards were run concurrently; the use of the National Bureau of Standards Oyster Tissue resulted in values within the range specified for this reference material.

Cadmium Uptake from Solution

To determine Cd accumulation from water, resistant and sensitive worms were exposed in plastic petri dishes to 8.9 M (= 1 mg/l) Cd in reconstituted fresh water (pH 7.8-8.0) for 6 days (Klerks & Bartholomew 1991). The addition of 109Cd resulted in a radioactivity of 22.2 kBq/ml in the exposure water. Three replicates of 10 worms each from Foundry Cove and South Cove stock cultures were exposed and survivors collected and frozen at -80C. Thawed samples were later homogenized in 50 mM Tris-HCl buffer (pH 7.4) using a 50:1 ratio of buffer to tissue. Cadmium concentrations of each homogenate were determined by gamma counting, using a Beckman 4000 gamma counter with a 3-inch sodium iodide crystal.

Fig. 9. Uptake of Cd by South Cove worms versus Foundry Cove worms. Note that Foundry Cove worms took up and accumulated more Cd.


Cd concentrations of Limnodrilus hoffmeisteri from the sediment toxicity bioassays did not differ among groups exposed to sediment with Cd raging from 5,400 to 34,000 g-Cd/g-dry sediment (p > 0.05, ANOVA) for worms from Foundry as well as South Cove (Fig. 9). These data were thus pooled to compare Cd accumulation of Foundry Cove worms with that of conspecifics from the control area (South Cove). The data provide no evidence for a reduced Cd accumulation in Foundry Cove worms; worms from Foundry Cove accumulated significantly more Cd in both sediment and water bioassays.

The data show that Cd-resistance in Limnodrilus from Foundry Cove is not due to a reduced accumulation of the metal. Worms from Foundry Cove actually accumulate more Cd than their metal-sensitive conspecifics from South Cove. Since Foundry Cove worms accumulate more Cd than their sensitive conspecifics, other mechanisms, such as sequestration (binding up the Cd in a detoxifying compound), must be responsible for the Cd resistance.

Other Animals Are Affected by Cadmium

Muskrats were conspicuously rare in Foundry Cove. Cadmium was measured in kidneys and elevated levels indeed were found compared to muskrats from control areas (Fig. 10a). It was thought that muskrats might have trouble with physiological function so lesions on the liver were counted in Foundry Cove muskrats, as compared to other marshes (Fig. 10b). Clearly, Foundry Cove muskrats were having problems, as evidenced by the high incidence of liver lesions.

Fig. 10. (a) Cd in muskrat kidneys, and (b) lesions observed on muskrat livers.

Mitigation and Recovery

In some localities, toxic materials exist at concentrations high enough to pose potential health risks. These toxic materials contaminate water, bind to sediments, and potentially bioaccumulate in the tissues of plants and animals. Many toxic materials that are detected in the environment are not due to new input, but due to existing compounds recycling between the water, sediments, atmosphere, and organisms.

Fig. 11. Sign at Marathon Battery Factory Site, 1994

Biological uptake patterns may differ among organisms; on the basis of such differences toxic materials are grouped into two categories: non-cumulative and cumulative toxic materials. Non-cumulative toxic materials do not increase in concentration in the body, even if the organism is chronically exposed to the toxin. Conversely, cumulative toxic materials, tend to increase in concentration, and are often associated with a specific tissue, e.g., Cd tends to increase over time in the digestive gland of blue crabs. Such accumulation may lead to food chain magnification, i.e., the magnification of toxic materials concentrations across trophic levels when the prey species possesses a physiological mechanism which concentrates the toxin in a specific tissue and the predator consumes large quantities of this prey.

Consumption of blue crabs by humans in the Hudson River is restricted because Cd can impair kidney function even at relatively low concentrations. Higher Cd accumulation in humans can potentially lead to other serious health risks including bone deformities, cardiovascular and immune system deficiencies, central nervous system disorders, and lung cancer. Beyond the risks to human health, high toxic materials contamination also poses ecological concerns by threatening indigenous wildlife and plant species. Toxic contamination also imposes economic and other societal effects as remediation is often very costly (beyond a local county or state budget) and the individuals involved in policy decisions differ in opinion as to the best course of action in remediation efforts.

Twenty-seven years of Ni-Cd battery production left Foundry Cove contaminated with 179 metric tons of cadmium. In 1983, a remedial investigation of Foundry Cove by the EPA and the NY State DEC, resulted in EPA declaring Foundry Cove a "superfund site" and millions of dollars were allocated to the cleanup.

The Marathon Battery site encompasses wetlands, archaeologically and historically sensitive lands (Indian artifacts and Civil War relics were discovered at the site), a warehouse facility, and contaminated residential areas (yards). The EPA designed specific treatments to address the effects of extensive metal contamination.

Fig. 12. Timeline of Mitigation

Foundry Cove Timeline 1979-1995

Using Superfund authority, and with the advice of scientists and residents, EPA designed the following remediation measures: (1) dredging, draining, and treating contaminated sediments and replanting acres of marshes along Foundry Cove, (2) excavating and treating contaminated soil in an underground vault on the plant property and tearing down plant buildings and processing towers, (3) decontaminating and recycling books stored at the plant, and (4) excavating contaminated soil from residential yards near the site and landscaping these yards. The EPA settled with the former battery plant owners to conduct the cleanup, estimated to cost $91 million. The responsible corporations also agreed to reimburse EPA $13.5 million for past cleanup and future oversight costs.

In 1992, the cleanup of the plant's interior and the recycling of the contaminated books on the property were completed. Starting in 1993, East Foundry Cove was dredged and the contaminated sediments were hauled away and treated.

Fig. 13. Towers Where Foundry Cove Sediment was Dewatered and Processed for Shipment

Prior to excavation a dike was constructed around the marsh to limit transport of sediments into the Hudson. West Foundry Cove was not dredged since the contamination was less severe than in eastern part of the cove and will naturally be covered by sedimentation over time. Constitution Marsh in nearby South Cove was not excavated because it received only low levels of metals and since it is a National Audubon Refuge human intervention was deemed an unnecessary risk.

Fig. 14. Aerial photograph of the site during restoration. Note newly dug creeks to maintain flushing and oxygenation. Also note the ring around the site. This was a large rubber bladder that resembled an inner tube. It protected the site from incursions of water as the cleanup and sediment removal proceeded.

Remediation of metal contaminated sediments has important policy implications. Dredging projects produce contaminated sediments which must be disposed and disposal of dredged materials is often problematic. Often such materials are disposed of in sealed underground vaults, but leaching of toxic materials from such vaults has often occurred, often with drastic consequences, i.e. contamination of the groundwater supply. Since such a remedy was used in Foundry Cove in 1971, groundwater contamination will be monitored for the next 30 years.

Fig 15. Railcars used to transport sediment from Foundry Cove (Cove is in foreground)

A rail spur was also built to haul away treated soil from East Foundry Cove; removal by trucks would have disrupted Cold Spring's historic downtown district and potentially damaged foundations to historic buildings. East Foundry Cove cleanup was completed in 1995.

Fig. 16. Foundry Cove Marsh, after replanting, June 1995

The top layer of contaminated soil was removed from nearby residential yards and re-landscaped. Wetland replanting efforts have just recently been completed, and the battery plant and processing towers are coming down; the site will soon be an empty lot. Wetland recovery will be monitored for a number of years.

The Hudson River serves as an example of how ecosystem contamination may have broad implications for long-term ecological and economic sustainability. Since the large scale release of Cd and other metals into the Hudson River more than 30 years ago, only recently has the issue of how to clean up the Hudson been resolved. PCB's entered the Hudson about 20 years ago and the issue of how to clean this contamination is still unresolved. Toxic contaminants have led to health concerns, a reduced commercial fishery, debate over the appropriate course for cleaning the river, and damage to the Hudson's public image.

Protecting and managing natural resources and promoting economic growth no longer occupy separate interests. Policy-makers are now realizing the interrelatedness of environmental, economic, and social programs, and balancing these concerns is no simple task. Understanding how contaminants get into the environment, what happens to them once they are there, and what can be done to minimize their impacts or remove them will be crucial to policy-makers forced to balance known (and unknown) risks of contaminants with the need for development.