"GM-": THE PROBABLE POISONING OF AN AMERICAN INDIAN NATION AT A SUPERFUND TOXIC WASTE SITE A Critical Review of the Remedial Investigation at GMC-CFD Massena Facility (USEPA Administrative Order on Consent, Index No. II CERCLA-50201)

May, 1988

Stephen Penningroth, Ph.D. Visiting Associate Professor Institute for Comparative and Environmental Toxicology Cornell University Ithaca, 14853 Tel. (607) 255-3099

Permanent Address: Stephen Penningroth, Ph.D. Associate Professor of Pharmacology University of Medicine and Dentistry of School of Osteopathic Medicine 675 Hoes Lane Plscataway, New Jersey 08854 Tel. (201) 463-4466

Review submitted to: Akwesasne Environmental Task Force James Ransom, Chairman St. Regis Mohawk Health Service Community Building Hogansburg, New York 13655

S7KV6 PM'UB ^

fi 10rr)PL£^ C-OPY ^ 77t6

381628 iiiiiiiiiiiiiiuiiiiiiiiiin PREFACE Superfund and Sacred Land A Remedial Investigation (RI) 1s a crucial step In the process of cleaning up a hazardous waste site under the Superfund Amendment and Reauthorization Act (SARA). The RI 1s followed by a Feasibility Study (FS), 1n which alternative technologies and strategies for remediating the contamination at the site are considered. The FS 1s followed by the Record of Decision (ROD), which specifies the technology (ies) to be used and outlines the site-specific remediation strategy. The RI is crucial because it sets the stage for the FS and the ROD by providing an estimate of the nature and the extent of contamination at the site. The quality of the remediation 1s only as good as the quality of the RI. If the RI falls to Identify a hazardous substance that 1s present at the site, then that hazardous substance will neither be considered 1n the FS nor included in the ROD. If the RI Identifies a hazardous substance at the site but fails to identify Its presence in a particular area, then that part of the site will not be considered 1n the FS or in the ROD and therefore will not be cleaned up, even though 1t may be heavily contaminated. Speaking metaphorically, the RI represents the "eyes" of the remediation process. If the RI 1s "blind" to hazardous substances and/or areas of contamination, then these will not be remediated, regardless of how efficient and/or Innovative the cleanup technology may be. It 1s expensive to conduct a proper RI of a Superfund site, for the simple reason that these sites are orders of magnitude larger than typical hazardous waste spills and therefore require a very large number of samples in order to define the nature and extent of contamination. Contamination generally extends deep into the ground, and a three-dimensional sampling grid 1s required. Such costs as are necessary appear justified by SARA, which mandates permanent remediation of Superfund sites. There can be no permanent remediation 1n the absence of a thorough RI. The GM-Massena Superfund site contains large quantities of polychlorynated b1phenyls (PCBs). The amount of PCBs spilled In soil and sediment exceeds by approximately a factor of seven the total quantity of PCB material spilled annually from capacitors and transformers In the .1 The PCBs 1n open on-site waste-water lagoons exceed the annual national PCB material spill by approximately 200-fold.1 This critique focusses on the question: How well does the RI "see" the nature and extent of contamination at the GM-Massena site? It concludes that the data 1n the RI, though Inadequate as a basis for effective remedial action, suggest widespread contamination; the recommendations In the RI, however, minimize the extent of contamination suggested by the data, thereby "blinding" the remediation process to an estimated 50% to 90% of the contamination at the site. To proceed with remediation on the basis of the recommendations 1n this RI would result 1n Incalculable damage to human health and the environment, because up to 90% of the PCBs and other hazardous substances 1n soil, sediment, and waste disposal lagoons, not having been "seen" by the RI due to an Inadequate database and/or to faulty Interpretation of existing data, would not be remediated. The GM-Massena site Is located on the St. Lawrence River. It 1s bordered on two sides by the St. Regis Mohawk , known to Native Americans as the Mohawk Nation of Akwesasne. The Superfund site has already had a profound § Impact on the lifestyle of Mohawks at Akwesasne. It has contributed to the s destruction of the loca.l fishing Industry and, with 1t, the barter economy in foodstuffs which flourished In Akwesasne for centuries.2 It has Impacted the ® 11 ' h-> nearby Akwesasne Freedom School, an alternative elementary school attempting to preserve traditional language and culture by means of a Mohawk Immersion program 1n which all subjects are taught 1n the . Parents' fears for their childrens' health have contributed to a reduction 1n the Akwesasne Freedom School's student body from 96 to 26 students over the past five years, and the school has twice undergone relocation.3 The Akwesasne Freedom School 1s a pivotal component in Native American cultural preservation efforts and has received both national and International attention, as evidenced, for example, by its inclusion 1n a November, 1987, edition of the Good Morning, America! show, and an invitation extended by Soviet Secretary General Gorbachev to meet with teachers and students from the Freedom School during his visit to Washington for the summit meeting with President Reagan in 1987.* The cultural traditions nurtured by the Freedom School have been recognized and affirmed 1n the context of the Bicentennial of the U.S. Constitution: A concurrent resolution acknowledging Iroquois contributions te the U.S. Constitution has been introduced in the U.S. Senate (Appendix 1), and a Mohawk Chief from Akwesasne has been invited to plant a Tree of Peace, a traditional Iroquois symbol of peace and goodwill in the Constitution Garden on the Mall 1n Washington, D.C., in September, 1988.3 Thus, through Its effects on the Akwesasne Freedom School, as well as on fishing, hunting, and agriculture at Akwesasne, the Superfund site poses a serious threat not only to the health of the and the Integrity of their environment, but also to the fabric of their traditional land-based culture, acknowledged by contemporary scholarship to have contributed significantly to the evolution of the American system of government.® The author has learned that to Mohawks, as to other Native Americans, land is sacred, Inhabited by animals and plants who are equal to mankind in the eyes of the Creator. Mohawks experience nature as kin. At the opening and closing of each day at the Akwesasne Freedom School, a child recites the traditional Thanksgiving Address 1n Mohawk, expressing kinship with and reverence for the created world (Appendix 2). In a theological sense, this kinship and reverence amount to a sacramental relationship with nature. This relationship Is voiced In the context of environmental pollution 1n a recent Mohawk government position paper on cleaning up the St. Lawrence River (Appendix 3). While the author has difficulty truly grasping the Native perspective, It nevertheless appears accurate to state that the Mohawks experience the poisoning of their land and water as a form of delclde. The Image of a child at the Akwesasne Freedom School learning that nature is beloved kin while less than a mile away an unremedlated Superfund site threatens to destroy the culture that nurtures that child's loving reverence and makes It possible, captures the Impending tragedy of Akwesasne. In a broader sense, it symbolizes the hazardous waste crisis of our age. We non- Indians share arith Indians a common destiny on the North American continent. We will, in the end, meet the same fate as the Akwesasne Mohawks, whatever that may be. The toxic noose around our necks draws tighter each year. A belief In the sacred quality of the natural world will enable us to see the noose, untie it, and fashion from the.rope a web of life. The Iroquois are Instructed by the Great Law of Peace to weigh the effects of each decision on the next-seven generations of children. In keeping with this Iroquois vision of social responsibility and continuity, the author dedicates his efforts to ensure remediation of the GM-Massena site to children 1n the year 2200.

Ithaca, New York May, 1988 TABLE OF CONTENTS

1. Summary and Conclusions 1 2. Recommendations 3 3. Background . 4

4. Site Classification 5 A. Summary B. GM-Massena site is 1n a "nonrestricted access area" C. Numerical Cleanup Standards 5. Delineation of Spill Boundaries 8 A. Summary B. Areal PCB Contamination of Soil C. Vertical PCB Contamination of Soil D. Areal PCB Contamination of St. Lawrence River/Seaway Sediment E. Vertical PCB Contamination of St. Lawrence River/Seaway Sediment F. PCB contamination of Soil and Sediment 1n Akwesasne 6. Potential Underestimate of PCB Concentratlori 1n Residential Nell Hater on Raquette Point 15 A. Interpretation of Results of RI B. Possible Masking of PCBs by Sulfur 1n Groundwater 7. Failure to Recommend Contaminated Areas for Remediation 16 A. "Nonpolnt Source" Area B. Lagoons and Sludge Pits 8. Evidence for Hazardous Substances In Addition to PCBs 18

9. Human Health Risk Assessment 19 A. Flaws 1n Assessment of Oncogenic Risk Due To Fish Consumption B. Risk Factors Not Included In the Risk Assessment 10. SOP Addendum 23 A. The Additional Studies Included In the SOP Addendum B. Inadequacy of the SOP Addendum

11. Reviewer's Remarks 25 A. The RI as a Scientific Document In the Public Domain B. Glossary of Misleading Technical Terminology 1n the RI C. The RI and the Regulatory Process 12. Recapitulation 28 Acknowl edgements 30

End Notes .31

1v Tables Table 1: Soil PC8 Concentrations at Depths Approximating 20 ft 1n the North Disposal Area Table 2: Soil PCB Concentrations at Depths Approximating 20 ft 1n the East Disposal Area Table 3: Soil PCB Concentrations at Depths Approximating 20 ft in the Industrial Landfill Area Table 4: Locations and Numbers of Surface Sediment Samples Collected From the St. Lawrence River Table 5: Mean PCB Concentrations 1n St. Lawrence River Sediment at Four Sampling Locations Table 6: Variation of PCB Concentration With Depth of Sediment and With Distance From GM Plant Outfall Table 7: PCB Concentrations 1n Surface Soil Samples From Akwesasne Table 8: PCB Concentrations 1n Sediment From Unnamed Tributary 1n Akwesasne Table 9: Lagoons and Sludge Pits Not Recommended For Remediation at the GM-Massena Site Table 10: Factors Elevating Oncogenic Risk From F1sh Consumption Compared to RI

Figures Figure 1: Nap of New York State Showing Location of Akwesasne (St. Regis Reserve) Figure 2: Nap of Akwesasne

Figure 3: At a Distance of 400 ft From the GN Plant Outfall, PCB Concentrations in Shallow and Deep St. Lawrence River Sediment Are Similar

Appendices

Appendix 1: U.S.Senate Concurrent Resolution 76. Appendix 2: Traditional Thanksgiving Address of the Iroquois. ' Q Appendix 3: St. Lawrence River Remedial Action Plan. Joint Position Paper. & Nohawk Council, St. Regis Nohawk Tribal Council, Nohawk Nation Council of .Chiefs. Akwesasne, October 14, 1987. o o V o u> T

Appendix 4: Letter of August 5, 1986, from James Ransom, consultant to St. Regis Tribal Council, to Melvln Hauptman, USEPA, commenting on Draft RI. Appendix 5: Letter of March 10, 1988, from Dr. James W. GUlett, Director, * Institute of Comparative and Environmental Toxicology, Cornell University, to Dr. Stephen Pennlngroth, Visiting Associate Professor, concurring with the major findings of this review. Appendix 6: Curriculum vitae of Dr. Stephen M. Penningroth Appendix 7: Letter of November 17, 1986, from John V. Czapor, Chief, Site Compliance Branch, USEPA, to David P. Fayette, GM Facility Coordinator, responding to comments on Draft RI. Appendix 8: Memo of August 7, 1987, from Lisa Gatton-V1dul1ch, Chemist, USEPA, to Christine Vlcznic, Project Manager, Site Compliance Branch, USEPA, supporting RMT's objection to filtering groundwater samples prior to analysis for PCBs. Appendix 9: General Considerations for Improving the RI.

o 3 3

o o

Vl i—1 —i o 1. SUMMARY AND CONCLUSIONS A "typical" PCB spill area covers about one-fortieth of an acre and extends to a depth of 10 Inches; 1t 1s cleaned up within 24 hours (TSCA Section 40 CFR 761). By contrast, the spill at the GM-Massena site covers between 13 and 250 acres of soil to a depth greater than 20 ft, and between 6 and 300 acres of river sediment to a depth of at least 2.5 ft. The volume of PCB material 1n soil and sediment 1s over seven times greater than the total volume of PCB material spilled annually from capacitors and transformers in the United States. The volume of PCB material in active and inactive waste water lagoons at the GM-Massena site exceeds the annual national PCB spill by a factor of 200 or more. PCBs have accumulated for 30 years at the GM-Massena site, which is located directly on the St. Lawrence River. Following are the salient findings and conclusions of this critical review of the Remedial Investigation (RI) of the GM-Massena site: Summary 1. Soil and sediment samples were collected at approximately 300 locations on GM property and in the St. Lawrence River next to the GM plant outfall. The sampling frequency averaged - very roughly - 1 sample per acre of soil and 1 sample for every four acres of river sediment. Conclusion 1: The data 1n the RI are Insufficient to delineate three- dimensional spill boundaries 1n soil and sediment with 95% confidence. Thousands of additional sampling events are required. Summary 2: The RI does not assess environmental Impact. Data 1n the RI on groundwater, soil and sediment 1n the Mohawk Nation of Akwesasne immediately adjacent to the GM-Massena site, though limited, show that PCB contamination of Akwesasne has occurred.

Conclusion 2: The tiny Indian nation of Akwesasne, which holds its environment sacred, is threatened with the progressive poisoning of Its land, water, animals, fish and plants by the GM-Massena site. Summary 3: The risk assessment In the RI distorts or Ignores available data on levels and pathways of exposure, health risks other than cancer, and toxic substances other than PCBs at the GM-Massena site.

Conclusion 3: The risk assessment In the RI underestimates the health risk to Akwesasne by a factor of at least 100.

Summary 4: The data In the RI Indicate that waste disposal lagoons at the GM- Massena site contain up to 30 times larger quantities of hazardous substances than do soil and sediment at the site. Dlbenzofuran, phenol, substituted phenols, polynuclear aromatic hydrocarbons (PNAs), vinyl chloride, trlchloroethylene, and other hazardous substances, Including possibly dloxln, are present at the site, according td the data In the RI. Conclusion 4: The RI Ignores, 1n Its recommendations and conclusions, the majority of the hazardous substances which 1t documents at the GM-Massena site.

Sunmary 5: Reliable data were not collected on PCB levels' 1n residential wells on Raquette Point nor gn dloxln and dlbenzofuran levels 1n on-site soil and in St. Lawrence River sediment.

1 rI

Conclusion 5: An understanding of key aspects of the GM-Massena site, crucial to a realistic endangerment assessment, was frustrated by technical failures in the laboratory. Sunnary &: The RI, citing TSCA Section 40 CFR 761, assigns the GM-Massena site a classification of •restricted access, or rural area• with a numerical cleanup standard of 25 ppm of PCBs in so;1. Conclusion 6: The classification "restricted access, or rural area• is not found ;n 40 CFR 761. The appHcable deHnition of the GM-Massena site h "nonrestricted access location• with a numerical cleanup standard for uncomplicated PCB spills of 10 ppm. Specific risk factors are described in 40 CFR 761 which call for a more stringent numerical cleanup standard than 10 ppm. Most of these risk factors are present at the GM-Massena site. SU11111ary 7: Conclusions in the RI tend to· be subjective, minimizing the scope of contamination suggested by the data. The reviewer•s independent analysis and interpretation suggest that the data in the RI provide a very rough, order-of-magnitude estimate of PCB contaminat;on at the GM-Massena site. Conclusion 7: The confusing presentation and biased interpretation of data effectively invalidate the RI as a scientific document and as a serious factor in the remediation of the GM-Massena site. ·

' G) :3: :3:

('" -'· 0 0 .I 2 °' ...... ,..... 0 °' I r ., 2. RECQMMENQATIONS 1. Change the classificatfon of the site from •restricted access, or rural area", to "non-restricted access location" • 2. Set the numerical cleanup standard for PCBs in soil and sediment at 1 ppm • (background). 3. Formulate a statistically based, unbiased sampling scheme (or schemes) which is capable of detecting remediable "hot spots" of hazardous substances with 95i confidence in three dimensions. 4. Investigate all potentially contaminated areas of soil, sediment and industrial facilities which may require remedial action: All on-site soil; all on-site buildings; all on-site waste disposal lagoons; soil and sediment in the Raquette Point section of Akwesasne; sediment in the St. Lawrence River upstream, downstream and in the Seaway channel; and sediment in the . 5. Include all hazardous substances within the scope of the remedial action plan, e.g., PCBs, dibenzofuran, phenol, substituted phenols, polynuclear aromatic hydrocarbon (PNAs), trichloroethylene, vinyl chloride, naphthalene, phthalates, and others. 6. Resolve technical problems and perform reliable determinations of dioxins and dibenzofurans. 7. Resolve technical problems and ensure that PCB concentrations in groundwater are not underestimated due to interference from sulfur and retention of PCBs on filters. 8. Monitor the contamination of aquatic and terrestrial biota. 9. Revise the human health.risk assessment, taking into account all available information on exposure, classes of hazardous substances at the GM-Massena site, and pathological effects. 10. Require that the RI conform to customary standards of scientific writing.

Q ~ ~

~ ~ °' 3 ..... ~ ~ • ~ \ {

3. MCQROUNQ The General Motors Corporation, Central Foundry Division (GM), since 1958 has operated an aluminum casting plant at a 250-acre site located ten miles east of Massena, New York, on the St. Lawrence Seaway (Figure 1). The structure housing the aluminum casting plant covers an area of about 25 acres. The plant produced Corvair engines from 1958 to 1968. Since 1968, when it stopped making Corvair engines, the pla9t has cast 4,000 tons of aluminum auto parts per month (48,000 tons· per year). Aluminum is obtained among other places from a Reynolds Metal Company plant situated one-half mile west of GM. GM occupies the western half of a peninsula, known as Raquette Point, which juts out from the southern bank of the St. Lawrence River/Seaway (Figure 2). The eastern half of the Raquette Point peninsula is occupied by the St. Regis Mohawk Indian Reserve, known to Native Americans as the Mohawk Nation of Akwesasne, or simply Akwesasne. Cornwall Island, a part of Akwesasne, lies opposite GM in the St. Lawrence River/Seaway. Consequently, GM is bounded on two sides, the east and the north, by Akwesasne. The prevailing wind direction . is westerly, and groundwater flow is to the east and north; hence, the · Raquette Point section of Akwesasne lies i11111ediately downstream, downwind and downgradient from GM, while the eastern half of the Cornwall Island section of Akwesasne lies downstream and downwind from GM. 8 GM is bordered to the west and the south by New York State. The Canadian province of lies on the far (northern) side of Cornwall Island, and the Canadian province of lies six miles downstream from GM, at the eastern border of Akwesasne (Figures 1, 2). . In 1980, the presence of large quantities of polychlorinated biph.enyls (PCBs) was detected on GM property by the New York State Department of Environmental Conservation (DEC). The PCBs apparently came from hydraulic fluids which had routinely leaked from die-casting machinery in the GM plant and had been bu~ied on-site as well as discharged, untreated, into the St. Lawrence River. Following investigation by the United States Environmental Protection Agency (USEPA), GM was fined $507,000 in 1983 for illegal disposal of PCBs, the highest fine levied under TSCA up to that date. The site was named GM-Massena. The GM-Massena site was placed on the·National Priority List c•superfund•) in September, 1983, and GM was ordered by USEPA to conduct a Remedial Investigation/Feasibility Study (Rl/FS) preparatory to site remediation. A draft Remedial Investigation Report (RI) was completed in May, 1986, by RMT, Inc., a consulting firm contracted by GM. Interested parties were given 60 days to respond 1n writing to the RI. A detailed response was submitted by James Ransom, an ethnic Mohawk Indian and a consultant to the St. Regis Tribal Council government. Mr. R1ns011 objected to •any aspects of the RI/FS. In particular, Mr. RansOll pointed out a) numerous instances of insufficient sampling which resulted in unacceptable uncertainties regarding PCB spill boundaries; b) inadequacy and arbitrariness in the methodology and conclusions of the human health risk assessment; and c) the absence of environmental assessment (Appendix 4)'. Mr. Ranse>11's coanents were made with advice and support from Professor James Gillett, 'Director of the Institute for Comparative and Environmental Toxicology, Cornell University. In response to critical co111111nts submitted by Mr. Ransom and New York State officials, GM was required by USEPA to perform additional investigative work. A Site Operations Plan (SOP) Addendum specifying additional studies requested by USEPA was published by RMT in June, 1987.9 The SOP Addendum ignored most of Mr. Ransom's co111111nts, particularly his calls for additional sampling in order to accurately delineate.spill boundaries. SOP Addendum studies focus primarily 0 0 °' 4 ...... ,...... 0 CX>