Habitability of the Love Canal Area: AnAnalysis of the Technical Basis for the

Decision on the Habitability of theEmergency Declaration Area

June 1983

NTIS order #PB84-114917

Recommended Citation:Habitability of the Love Canal Area: An Analysis of the Technical Basis for the Divisionon the Habitability of the Energency Declaration Area-A Technical Memorandum(Washington, D. C.: U.S. Congress, Office of Technology Assessment, OTA-TM-M-I3,June 1983).

Library of Congress Catalog Card Number 83-600552

For sale by the Superintendent of Documents,U.S. Government Printing Office, Washington, D.C. 20402

Foreword

In December 1982, OTA was requested by Senators Daniel Patrick Moynihan andAlfonse D’Amato of New York to perform a case study of Love Canal. The requestcame at a time when OTA was completing its 3-year study of hazardous wastes. OTA’sreport Technologies and Management Strategies for Hazardous Waste Control was re-leased in March 1983. OTA’s larger study on hazardous wastes formed the basis forthe case study on Love Canal.

The principal goal of the case study was to examine the technical basis for, andvalidity of, the habitability decision for the emergency declaration area near Love Canal.This decision was made by the Department of Health and Human Services, and wasbased principally on the results of a monitoring study conducted by the EnvironmentalProtection Agency. OTA was also asked to examine the current monitoring and cleanupactivities at Love Canal, and the plans for future remedial action.

This technical memorandum consists of three principal sections:

1.

2.

3.

OTA’s findings on the habitability decision, and the four arguments on whichthey are based.An outline of the steps that could be taken to safely achieve incremental rehabita-tion of the emergency declaration area.Implications of the results of this case study for the national Superfund programfor uncontrolled hazardous waste sites, under which cleanup at Love Canal isproceeding.

The timely completion of this study was made possible by the extensive assistanceprovided by many people at several New York State agencies, the Environmental Pro-tection Agency, the Department of Health and Human Services, and the National Bureauof Standards.

,..I l l

.— —

OTA Love Canal Case Study Project Staff

OTA

Lionel S. Johns, Assistant Director, OTAEnergy, Materials, and international Security Division

Audrey Buyrn, Industry, Technology, and Employment Program Manager

Project Staff

Joel S. Hirschhorn, Project Director

Suellen Pirages, Senior Analyst

Michael Gough, Senior Associate

Administrative Staff

Carol A. Drohan Patricia A. Canavan

Contractors

Robert Michaels Katherine Gillman L. Adrienne Cupples

Publishing Staff

John C. Holmes, Publishing Officer

John Bergling Kathie S. Boss Debra M. Datcher Joe Henson

Glenda Lawing Linda Leahy Donna Young

Contents

Page

Acronyms, Abbreviations, and Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

CHAPTER l: Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4The OTA Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Scope of the OTA Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7The Four Arguments for OTA’s Principal Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Possible Steps Toward Rehabitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14Implications for the Superfund Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

l. The ’’How Clean Is Clean?’’ Issue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152. Health Effects Data and Decisions inhabitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163. Technical Guidelines for Monitoring Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164. Selection and Implementation of Remediation Programs . . . . . . . . . . . . . . . . . . . . . . . . 16S. Long-Term Institutional Capabilities and Issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

APPENDIX A: Remediation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Categories of Remedial Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Evaluation of Alternative Technologies at Love Canal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Control Action at the Love Canal Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Uncertainties Associated With the Remedial Action. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

APPENDIX B: Design of the EPA Monitoring Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Scope of the EPA Monitoring Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Evaluation of the Sampling Effort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31Conclusions About the Sampling Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

APPENDIX C: Results of the EPA Study Related to the Habitability Decisions . . . . . . . . . . . . 40Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Basis of the Habitability Decision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Problems With Statistical Comparisons of EPA Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Range of Variability for Reported Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41What If EPA’s Numbers Are Wrong? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44Uncertainties in Potential Health Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45The Special Case of Dioxin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

APPENDIX D: Analysis of the EPA Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54Statistical Analysis of Indicator Substances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Contents—Continued

List

Table

A-1.A-2.A-3.A-4.B-1.B-2.B-3.B-4.B-5.c-1.

c-2.c-3.c-4.

c-5.

C-6.

D-1,D-2,

D-3,

D-4.

D-5.

of TablesNo. Page

Advantages and Disadvantages of Control Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Summary of Lifecycle Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Performance Criteria Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Priority Pollutants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26Criteria Used To Evaluate EPA Sampling Effort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32Number of Sites and Samples for Target Substances 33Approximate Size of EDA Study Regions 34Sampling Effort for Dioxin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Comparison of Dioxin Sampling Effort Between Eastern Missouri and the EDA . . . . . . . 39Number of Samples Required To Detect Actual Differences Betweenthe EDA and Control Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Maximum Values Reported for Organic Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Maximum Values Reported for Dioxin, pub. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Summary of Test Results Available on Health Effects of ChemicalsDisposed in Love Canal and Monitored by EPA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48Comparison of Regulated Exposure Limits to Detected MaximumConcentrations in the EDA: Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Comparison of Regulated Exposure Limits to Detected MaximumConcentrations in the EDA: Air . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Odds Ratio and Detection Rates for l,2-Dichlorobenzene . . . . . . . . . . . . . . . . . . . . . . . . . . . 54Detection of Indicator Substance and Numbers of Samples Analyzedfor Environmental Media in the Control Areas.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55Detection of Indicator Substance and Numbers of Samples Analyzedfor Environmental Media in the EDA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56Detection of Substances Not Found in Control Areas and Numbers ofSamples Analyzed for Environmental Media in the EDA . . . . . . . . . . . . . . . . . . . . . . . . . . . 57Summary of Mantel-Haenszel Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

List

Figure No. Page

B-1.B-2.B-3.B-4.B-5.B-6.B-7.B-8.B-9.B-10.

of Figures

The 10 Subregions of the EDA Included in the EDA Monitoring Study . . . . . . . . . . . . . . 31Distribution of Shallow Well Sampling Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Distribution of Deep Well Sampling Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Distribution of Soil Sampling Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Distribution of Sump Water Sampling Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Disribution of Storm Sewer Sampling Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Surface Water Sampling Sites Along the Black Creek and the Bergholtz Creek . . . . . . . 36Distribution of Air Sampling Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Distribution of Drinking Water Sampling Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Distribution of Dioxin Sampling Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

vi

—

Acronyms, Abbreviations, and Terms

ACGIH

ACSC

CDC

CERCLA

DHHS

EDAEDFEPAFEMA

GAO

gpmIARC

LCARA

—

——

—

—

—

————

—

——

—

Acronyms and Abbreviations

American Conference of GovernmentIndustrial HygienistsAmerican Chemical Societycontrol areas sampled by EPA moni-toring studyCenters for Disease Control, PublicHealth Service, U.S. Department ofCommerceComprehensive Environmental Re-sponse, Compensation, and LiabilityAct of 1980, 42 U.S.C. 9601 et seq.,known colloquially as “Superfund”U.S. Department of Health andHuman Servicesemergency declaration areaEnvironmental Defense FundU.S. Environmental Protection AgencyFederal Emergency Management Actof 1978, CFR Title 44, Part 2, andassociated AgencyGeneral Accounting Office, U.S.Congressgallons per minuteInternational Agency for Research onCancer , U .N. World HealthOrganizationLove Canal Area RevitalizationAgency

LOD –LOQ –MDL –NAS –NBS –

NCI –NTPNYS/DEC -

NYS/DOH –NYS/DOL –OTA –

ppb —

ppm —ppt —RCRA –

TLV –TSCA –

µg/kg —

µg/1 —

limit(s) of detectionlimit(s) of quantitationmethod detection limit(s)National Academy of SciencesNational Bureau of Standards, U.S.Department of CommerceNational Cancer InstituteNational Toxicology ProgramNew York State Department of En-vironmental ConservationNew York State Department of HealthNew York State Department of LawOffice of Technology Assessment, U.S.Congressparts per billion, equal to microgramsper liter in water, micrograms perkilogram in solids or gasesparts per millionparts per trillionResource Conservation and RecoveryAct of 1976, 42 U.S.C. 6901 et seq.threshold limit valueToxic Substances Control Act of 1976,15 U.S.C. 2601 et seq.micrograms per kilogram, equal to ppbin solids or gasesmicrograms per liter, equal to ppb inwater

Terms

emergency declaration area (EDA): area sampled by environmental submedium: 1 of 22 components of theEPA monitoring study, outside 49-acre area sur- five environmental mediarounding the canal landfill, divided by EPA into 10 indicator substances: substance which was known tosubregions for sampling purposes. have been disposed into canal landfill

Love Canal: 49-acre area including canal landfillenvironmental medium: one of five environmental sampled by EPA monitoring study

components sampled by EPA, including water, soil, target substance: 1 of 50 substances monitored in EPAsediment, air, biota. monitoring study

vii

Chapter 1

Executive Summary

Chapter 1

Executive Summary

SUMMARY

In 1980, the Federal Government decided thata potential health threat existed in an area nearLove Canal, beyond the canal itself and two in-ner rings. This outer area was termed the emergen-cy declaration area (EDA). In cooperation withNew York State, the Federal Government pro-vided assistance for residents who wished to moveout of the area. There was no clear evidence atthe time of substantial and widespread contamina-tion of the EDA by toxic chemicals from LoveCanal.

From August to October 1980, the Environmen-tal Protection Agency (EPA) conducted a mon-itoring study to provide evidence for determin-ing whether the EDA was contaminated or couldbecome so. Later, the Department of Health andHuman Services (DHHS) became responsible fordeciding, on the basis of the EPA study and otherdata, whether the EDA was habitable. In July1982, after considering comments by the NationalBureau of Standards (NBS) on the proceduresEPA had used, and after further consultation withEPA, DHHS affirmed its earlier provisional deci-sion that the EDA was habitable. The decision wascontingent on effective safeguards against leakagefrom the canal, and cleaning up contaminationin the EDA.

OTA has reviewed and analyzed the EPA mon-itoring study, documents prepared by other Fed-eral agencies, plans for remedial action developedby EPA and New York State, and various inde-pendent critiques. OTA’s primary goal was to ex-amine the technical basis for the decision reachedby DHHS, in conjunction with EPA, that the EDAis habitable.

OTA’s principal finding is that: With availableinformation it is not possible to conclude eitherthat unsafe levels of toxic contamination exist orthat they do not exist in the EDA. The OTA anal-ysis does not support an interpretation of theDHHS decision that would lead to the immediateand complete rehabitation of the EDA. There re-

mains a need to demonstrate more unequivocal-ly that the EDA is safe immediately and over thelong term for human habitation. If that cannotbe done, it may be necessary to accept the originalpresumption that the area is not habitable.

Four arguments that support the principal find-ing are:

1.

2.

3.

4.

The current activities and long-term plansfor EDA cleanup and operation and main-tenance of the Love Canal remedial actionprogram pose difficulties and uncertainties.The design of the EPA monitoring study,particularly its sampling strategy, was inade-quate to detect the true level and pattern oftoxic chemical contamination that might ex-ist in the EDA.The EPA monitoring study contains impor-tant uncertainties over the levels of the tox-ic chemicals detected, and the possible levelsof those not detected. There are also uncer-tainties over possible synergistic humanhealth effects of multiple toxic chemicalspresent at low concentrations. These twoareas of uncertainty, as well as the lack ofdetailed documentation by DHHS of its anal-yses, place the decision on habitability byDHHS in doubt.OTA’s analysis of some data obtained in theEPA monitoring study provides limited, butnot conclusive, indication that there may becontamination in the EDA by toxic chemicalsfrom Love Canal. OTA examined those datafor chemicals known to have been disposedin Love Canal, as compared to the muchlarger universe of data analyzed by EPA.

Incremental rehabitation of the EDA is a possi-ble alternative to complete rehabitation, or to apresumption that the area is not habitable. OTAhas outlined several steps that could be taken tomove in this direction. By incremental rehabita-tion we mean a paced, cautious approach. Im-provements in scientific certainty to assure safe-

3

4

ty of the EDA are necessary for the success of thisapproach. Another benefit of improved certain-ty is to increase public confidence in policy deci-sions that are based on technically complex dataand analyses. Four key steps for moving towardincremental rehabitation are:

1. To address the technical problems and un-certainties in the current cleanup activitiesin the EDA and in the long-term plans foroperation and maintenance of the wastecontainment system at Love Canal.

2. To address the uncertainties related to in-stitutional stability and effectiveness overthe very long terms (i.e., hundreds of years)that reflect the long lifetimes of the chem-icals in Love Canal.

3. To consider performing additional monitor-ing for carefully defined areas of the EDA,perhaps for individual homesites. Thiswould make use of what has been learnedfrom the EPA monitoring study.

4. To develop a program for finding a perma-nent solution to deal with the large amountsof toxic wastes still in Love Canal (i.e.,waste destruction or detoxification insteadof the waste isolation approach now in ef-fect).

OTA's case study of the Love Canal EDAtouches on a number of issues of general impor-tance to the Federal Superfund program for clean-up of uncontrolled hazardous waste sites. EPA isnow aware of about 16,000 uncontrolled hazard-

BACKGROUND

In 1980, the Federal Government decided thata potential health threat existed beyond the canal(shaded area on figure) itself. The outer area wastermed the emergency declaration area (EDA). Seethe accompanying figure for a description of theEDA and Love Canal areas. In cooperation withNew York State, the Federal Government pro-vided assistance for residents who wished to moveout of the area. Not all residents in the EDA de-cided to relocate. The canal itself and Rings 1 and2 had been the subject of an earlier Federal stateof emergency which included total evacuation andintensive cleanup efforts. As for the EDA, therewas no clear evidence at the time of the emergency

ous waste sites nationwide. The use of monitor-ing studies to answer questions on relocation andhabitability will likely continue to be necessary.Therefore, it is important to learn as much as pos-sible from the Love Canal experience to make fu-ture efforts more effective and efficient. There areneeds to:

Examine the “How clean is clean?” question,and to develop standards for unacceptablelevels of contamination by toxic chemicals.Obtain much more information on the healtheffects of toxic chemicals, and better definethe Federal decisionmaking process concer-ning habitability of, and relocation of residentsfrom, uncontrolled hazardous waste sites.Develop technical guidelines for monitoringstudies, particularly for sampling and ana-lytical protocols, and for the way results arepresented and documented.Compel consideration of more permanent so-lutions for cleaning up uncontrolled wastesites, and to develop ongoing programs toevaluate technological opportunities foreventual permanent solutions to replacewaste containment “interim solutions. ” It isalso necessary to improve oversight by EPAof State implementation of chosen remedialaction programs.Explore answers to problems of long-term in-stitutional effectiveness, such as mechanismsto assure indefinite funding for operating andmaintaining waste containment systems.

declaration that substantial and widespread con-tamination by toxic chemicals existed in the area.The voluntary evacuation was considered a pre-cautionary measure.

Nevertheless, Government actions have raisedthe issue of habitability of the EDA. These actionsincluded: providing assistance for relocation, im-plementing monitoring studies to determine if con-tamination was present (and at what levels andpatterns), carrying out studies on possible healtheffects among residents of the EDA, and imple-menting the cleanup program to correct theknown problems in the canal itself and the adja-cent rings. In light of these activities, the hab-

5

Love Canal Revitalization Area —

s c a l e 1 = 2 0 0

—

\

II

J

\ ‘-. - . ‘, ●“% I

c , .\ . ” ., ~% ~ ‘

. . .> ..4 I ,’ ‘]-. . .

■ - UNOCCUPIED: HOUSES PURCHASED BY LCARA ‘.● -(423 AS OF 4/1/83) ‘. ,,

49- GRIFFON MANOR UNITS OCCUPIED. J.

17- SENIOR CITIZENS UNITS OCCUPIED.,\ \

NOTE: UNSHADED OUTLINES ARE PROPERTIESTHAT ARE OCCUPIED . .

21-893 0 - 83 - 2

6

itability of the EDA became, and still remains,an important issue.

Government agencies anticipated that scientificevidence of no contamination, or of acceptablylow levels of contamination, would be necessarybefore people could be allowed to move back intothe evacuated portions of the EDA. From Augustto October 1980, EPA conducted a monitoringstudy to provide evidence for determining wheth-er the EDA was contaminated by toxic chemicalsfrom Love Canal, or could become so. The re-lated, difficult task of deciding whether or not theEDA was habitable was undertaken by DHHS.DHHS was to make its decision primarily on thebasis of the EPA monitoring study, but was alsoto consider other data on contamination levelsand on health-related problems observed in res-idents, as well as professional judgments aboutpossible health effects that could result from ex-posure to chemicals present in the EDA.

EPA asked NBS to examine the analyticalchemistry procedures EPA used in its monitoringstudy. After completion of the initial NBS reportin May 1982, DHHS requested NBS to explain thesignificance of the negative data in the EPA study.Ninety percent of the samples taken by EPA hadrevealed either no detectable levels or traceamounts of contamination. DHHS wanted to besatisfied that the data were reliable enough toallow the conclusion that, in fact, only very lowlevels of chemicals were detected. The possibili-ty that EPA’s analytic techniques may have missedcontamination in the low parts-per-billion rangehad to be considered. DHHS was satisfied withEPA’s responses to the NBS comments on thismatter. In July 1982, DHHS affirmed its earlierprovisional decision that the EDA “is as habitableas the control areas with which it was compared.”However, DHHS made its decision on habitabilitycontingent on the understanding that the canal siteitself and Rings 1 and 2 would be “constantlysafeguarded against future leakage from the canaland that cleanup is required for existing con-tamination of local storm sewers and their drain-age tracts [in the EDA].”l

1“DHHS Evaluation of Results of Environmental Chemical TestingBy EPA in the Vicinity of Love Canal-Implications for HumanHealth–Further Considerations Concerning Habitability” (Wash-ington, D. C.: Department of Health and Human Services, July 13,

1982).

It must be stressed that EPA’s monitoring taskfor a large area possibly contaminated by hun-dreds of different chemicals was historicallyunique, technically complex, and very large inscope. EPA had no precedent for a similarly broadand complex monitoring study, targeted towarda decision on whether a site possibly contaminatedby toxic wastes was habitable or rehabitable.Time for the study was severely constrained be-cause of the Government’s strong desire to makea policy decision on habitability. The effort wasfurther complicated by other factors:

1.

2.

3.

Information on the existing or potential mi-gration of chemicals from the canal was lack-ing.

Boundaries for the EDA were arbitrary, un-related to technical considerations of possi-ble routes of transport of chemicals from thecanal.

The time constraints ruled out a pilot studyto define the requirements for ‘the largermonitoring program (e.g., to identify qual-itatively the levels of contamination to be ex-pected, thus influencing sampling design andthe choice of analytical procedures).

The findings of the study, moreover, made ahabitability decision difficult. The discovery ofhigh levels of contamination of even a few chem-icals can provide a reasonably certain basis fora decision of “nonhabitability,” but it is far moredifficult to contend with a situation involving lowlevels of contamination. The latter was the unex-pected result of the EPA monitoring study. Didthe low levels of contamination in the EDA foundby EPA accurately reflect reality? For the generalpublic as well as technical experts, there were con-cerns about the sampling, analysis, and interpreta-tion of EPA’s data. Thus, the goal of reaching apolicy decision on habitability quickly was notmet. Interpretation of the data collected by EPAwas difficult in itself. In addition, a number ofparties raised questions about the study which re-quired further analysis.

The problems confronting New York State inthe Love Canal cleanup were also unique, diffi-cult, and complex. Love Canal was one of the first

7

major uncontrolled hazardous waste sites in theNation where remediation was attempted. Therewas a paucity of previous experience to offerguidance on the technological problems of short-term cleanup and long-term maintenance. These

THE OTA ANALYSISScope of the OTA Analysis

At the request of the two U.S. Senators fromNew York, and based on OTA’s previous workon hazardous waste,2 OTA reviewed and ana-lyzed the EPA monitoring study, documents pre-pared by DHHS and NBS, the plans for remedialaction developed by EPA and New York State,and various independent critiques. OTA focusedon the technical aspects of sampling design andanalytical procedures used to obtain monitoringdata, on the statistical methods used to evaluatethe data, and on the immediate and long-termremediation plans for the entire Love Canal area.It was not OTA’s task to reach a finding concern-ing habitability of the EDA or to obtain new mon-itoring data. OTA’s primary goal was to examinethe technical basis for the decision already reachedby DHHS in conjunction with EPA. OTA wasalso asked to consider possible implications of thiscase study for the national Superfund programfor cleaning up uncontrolled hazardous wastesites.

2 Technologies and Management Strategies for Hazardous WasteControl (Washington, D. C.: U.S. Congress, Office of TechnologyAssessment, OTA-M-196, March, 1983).

problems must be solved. Moreover, EPA’s Su-perfund program has not yet reached the stageof providing complete, effectivepolicy guidance on remediation.

technical and

The principal findings:1.

2.

With available information it is not possi-ble to conclude either that unsafe levels oftoxic contamination exist or that they do notexist in the EDA.There are also serious concerns and uncer-tainties about progress in the remedial pro-gram to date and plans for the future.

The OTA analysis does not support an interpreta-tion of the DHHS decision that would lead to theimmediate and complete rehabitation of the EDA.There remains a need to demonstrate more une-quivocally that the EDA is safe for human habita-tion immediately and over the long term. If thatcannot be done, it maybe necessary to accept theoriginal presumption that the area is not hab-itable.

The three following sections discuss:

1.

2.

3.

Four arguments that support the above find-ings, with the more detailed supporting anal-yses provided in an appendix to this report.A number of Federal and State actions thatmight be undertaken for incremental rehabi-tation of the EDA over time.Implications and issues for the nationalSuperfund program of the Love Canal case.

THE FOUR ARGUMENTS FOR OTA’S PRINCIPAL FINDINGS

1. The current activities and long-term plans for The current activities and long-term plans forEDA cleanup and operation and maintenance cleanup of the entire canal area and for isolationof the Love Canal remedial action program of the wastes remaining in the canal must be ef-pose difficulties and uncertainties. fective (see the appendix for a description of these

activities and plans). Effective cleanup was a crit-ical contingency in the DHHS decision on hab-

Regardless of whether contamination of the itability, for two main reasons:EDA by toxic chemicals exists now, rehabitation 1. in several areas of the EDA there are con-of the EDA requires assurance about the future. firmed high concentrations of dioxin which

8

2.

pose a threat either if the dioxin stays whereit was found originally or if it migrates else-where; andthere remain in the canal itself very largeamounts of toxic wastes and contaminatedsoil, both of which pose threats unless theyare safely isolated from the environment,totally removed, or permanently destroyedor detoxified onsite.

The remediation approach at Love Canal iscontainment of the toxic wastes so that there isno migration of toxic chemicals into the surround-ing environment. This approach raises the ques-tion of what is meant by “long-term” in contain-ing these wastes. Both technically and institu-tionally, “long-term’’ for toxic wastes must be in-terpreted to mean hundreds or thousands of yearn.Why is it necessary to go beyond several decadesin assuring the effectiveness of the containmentsystem? Because many of the toxic chemicals pres-ent in the canal area are expected to remain stableand hazardous indefinitely. It is difficult to con-ceive of sanctioning rehabitation in the areas mostexposed to leakage of toxic chemicals from thecanal area without a high level of confidence thatthe remediation plan will remain effective beyondthe next few decades.

OTA sees three major reasons for concluding,at present, that further attention must be givento site cleanup and remediation before rehabita-tion of the EDA can proceed.

First, the areas in the EDA contaminated withhigh levels of dioxin have not yet been cleanedup. Moreover, until just a few months ago stormsewers leading from the canal area to the EDAand known to contain dioxin remained open. Itis possible that during the past few years—aftercompletion of the EPA monitoring study—dioxinmay have been transported within or beyond theEDA. A study to determine the full extent of con-tamination in and near the sewers is not yet com-pleted. When it is completed, it will greatly assistthe cleanup effort.

Second, there are technical problems with thecurrent activities and plans for the canal itself andthe immediate rings. These include:

• Leak Detection Systems. —The long-term in-tegrity of the remedial technology is not cer-

tain. Reliable methods are needed to allowearly detection of damage (leading to per-meability) to the two basic elements of thecontainment system. These elements, in-tended to minimize water entering the canal,are the cap over the canal area and the con-crete barrier wall to be built around it. Thereis no dispute about the need for repair andreplacement of the cap and leachate collec-tion system over time, Yet, how it will bedone is not clear. How structural damage orclogging of the drain system will be detected,and how repair and replacement can be car-ried out safely remain unanswered.Monitoring Programs. —Assurance of suffi-cient warning about any potential migrationand accumulation of chemicals from thecanal is essential. Plans are underway for de-veloping a long-term monitoring plan forground water in the area immediately adja-cent to Love Canal, but not in the EDA. Inthis same area (adjacent to Love Canal) it isalso necessary to design more extensive am-bient monitoring of environmental mediaother than ground water (e.g., air, soil, andbiota). Media other than ground water arepossible routes of exposure to toxic chem-icals. For example, depending on the proper-ties of chemicals disposed in the canal andthe soil through which the ground watermoves, some chemicals could be filtered outand could accumulate in soil or possibly inbiota. Humans might become exposed toeither. In addition, damage to the cap couldallow release of volatile compounds into theair.

third major area of uncertainty concerns thelong-term ability of government institutions toremember, fired, and carry out commitments forlong-term continued monitoring and maintenanceof the site. The full range of institutional issuessurrounding very long-term commitments formanaging uncontrolled hazardous waste sitesunder the Superfund program have nowhere yetbeen fully addressed. New York State is not alonein facing these questions. But Love Canal is some-thing of a historical first, and may be viewed asa model. Annual cost estimates for routine opera-tion and maintainance, as well as for replacement,of the leachate collection system at Love Canal

9

are about $0.4 million now, $4.2 million in theyear 2005, and $8.5 million in 2030. These costscould rise if the other leak detection and monitor-ing needs noted above are found to require at-tention. Even though the present State administra-tion is committed to providing these funds, thereis no guarantee that State officials 20 or 100 yearsfrom now will either remember or honor this com-mitment.

It is reasonable to raise the prospect that evenlarger funds might be needed at some time to takefurther corrective action at the Love Canal siteif the original containment system were to fail.Furthermore, there are few institutional mecha-nisms in place to assure continuity in transferringvital information on Love Canal from one genera-tion to the next. Nor does it appear that New YorkState has taken binding and permanent title to thecanal area in a manner that unequivocally rulesout future use of the site. * Finally, the idea thatthe present containment system is no more thanan acceptable “interim solution” requires more at-tention. No specific program has been undertakento find a more permanent remedy for removing,destroying, or detoxifying the wastes in the canal.

2. The design of the EPA monitoring study, par-ticularly its sampling strategy, was inadequateto detect the true level and pattern of toxicchemical contamination that might exist in theEDA.

The principal finding of the EPA monitoringstudy was that, except for a few locations withhigh levels of dioxin and some other chemicals,there were insignificant levels and patterns of con-tamination in the EDA attributable to wastes inLove Canal. The finding was based on analysisof the samples taken in the EDA. Our concernis that the design of the monitoring study was notadequate to detect all significant contaminationthat might be present.

The uncertainties which OTA sees as criticalinvolve such aspects of design as how many sam-

● This is not to imply that there is any serious consideration be-ing given to reuse of the canal. But some people may raise this pros-pect for the future. Reuse already took place once, when govern-mental bodies deemed the canal area safe for community develop-ment after it was no longer used for waste disposal.

pies were taken for specific chemicals, in howmany locations within the EDA, and in what en-vironmental media. Whether a monitoring studydetects contamination depends on how the searchis conducted. If the design of a monitoring studyis inadequate, then an erroneous false-negative in-terpretation may result.

The absence of a strong positive finding of con-tamination does not at all imply that a negativefinding (absence of contamination or absence ofhealth effects) follows logically or persuasively.In the case of a monitoring study, particularly onecarried out under serious time constraints andwithout the benefit of a pilot study, sampling in-adequacies can lead to a low level of confidencein the results. While there may never be absoluteconfidence that a study can find what it is look-ing for, the issue in the case of the EPA monitor-ing study is that the confidence level is low.

This lack of confidence in negative results (thefinding of an absence) presents substantial prob-lems to policymakers who desire a firm, scientificbasis for decisionmaking, but it is sometimes aninevitable outcome of scientific studies. Scientiststhemselves often find it difficult to give an answerof “I can’t determine, or I’m not sure” rather thana “yes or no” answer. Low confidence in the de-sign of a study to produce the desired informa-tion, it should be noted, is not the same as scien-tific uncertainty over the results of a study; uncer-tainty is discussed in a later section.

The following specific problems with the sam-pling procedures used by EPA led OTA to judgethe outcome of the study indeterminate with re-gard to the extent (or distribution) and level ofchemical contamination, and its site and regionalvariability y:

● The monitoring study sampled unevenlyacross environmental media and the 12 re-gions (10 in the EDA, the canal, and the con-trol). The numbers of sampling sites were notin proportion to sizes of the regions, whichvary by a factor of 10. One reason for thissituation was that EPA assumed that higherlevels of contamination existed closer to thecanal. Consequently, some regions fartheraway from the canal had very little sampling;

10

●

●

●

●

the distribution of sampling among regionsin the EDA was particularly inadequate. In-itial beliefs about possible routes of transportof toxic chemicals from the canal to andthrough the EDA may also have influencednumbers of sampling sites in environmentalmedia. To the extent that these assumptionsabout patterns remain unproven or unsup-ported by the results of the study, it can beconcluded that the sampling may not havedetected contamination present in the EDAwhich does not correspond to the patternsassumed initially by EPA.The numbers of sampling sites used were in-sufficient to determine accurately the levelof contamination within some regions.As for environmental media, the extent ofsampling was very broad and included air,surface and ground water, soil, sediment, andbiota. However, the effort across media wasuneven, and there was no examination ofyearly seasonal variations. Within the EDA,those media sampled most extensively weresoil, air, and sump water. Ground water wassampled less extensively and biota weresampled least often of any of the environ-mental media. Sampling in some media mayhave been inadequate to detect contamina-tion.Too few replicate samples were collected persite to evaluate site variability; thus, the dataon absolute concentrations of chemicals de-tected within any one region may not bemeaningful.The study lacked adequate control area data;thus, comparisons among regions are diffi-cult. However, as discussed more fully later,DHHS did not rely entirely on the controlarea data in its habitability decision.

The considerations outlined above apply to allthe chemicals sought in the EPA monitoringstudy. However, OTA has examined the samplingsituation for dioxin in greater detail because:

●

●

●

dioxin is generally viewed as a very toxicmaterial at very low concentrations,very high levels of dioxin were found in somelocations within the EDA,public sensitivity to dioxin contamination ishigh, and

• it is possible to make some comparisons be-tween the Love Canal dioxin sampling andthat done by EPA recently in Missouri.

Monitoring for dioxin was insufficient with re-spect to extent (distribution), level, and replica-tion. No conclusions can be drawn from the ab-sence of positive findings for dioxin in most ofthe EDA. There can be little assurance that thefindings accurately describe any contaminationthat could or could not exist there. Only 6 outof 21 environmental submedia in the EDA weresampled for dioxin; of the 10 regions in the EDA,only two were sampled for sump water contami-nation and three each for air and soil. In the 10EDA regions no more than five sites were sam-pled per region, except for storm sewer sediment.No attempt was made to take replicate samplesat all sites; this is particularly important becausedioxin binds strongly to organic particles. Somefurther indication that sampling for dioxin in theEDA was inadequate is that in three Missouri sitesthe number of samples ranged from about 4 to37 times more per acre than those used in theEDA.

3. The EPA monitoring study contains importantuncertainties over the levels of the toxic chem-icals detected, and the possible levels of thosenot detected. There are also uncertainties overpossible synergistic human health effects ofmultiple toxic chemicals present at low concen-trations. These two areas of uncertainty, aswell as the lack of documentation by DHHSof its analysis, place the decision on habitabilityby DHHS in doubt.

The results of the EPA monitoring study werethe major basis for the DHHS habitability deci-sion. Two lines of evidence have been offered tosupport the view that the EDA is not too con-taminated for habitation. DHHS has asserted that:

1. the EDA is no more contaminated than “con-trol areas” near the EDA, and

2. the absolute levels of contamination are solow as to present no health threat.

OTA has not emphasized in this discussion acomparison of findings in the EDA with those in

11

the control areas used in the monitoring study.As discussed in the appendix, OTA’s own anal-ysis, analysis in other studies, and, to a degree,even EPA’s own analysis disclose critical flaws inthe study’s use of control areas. As a result, mostof the comparisons made between the EDA andcontrol areas lack statistical confidence. In anycase, DHHS maintains that its decision on hab-itability did not solely depend on making com-parisons with the control areas.

Therefore, the following discussion focuses onthe data which did form the critical basis of theDHHS decision—-the levels of contamination de-tected in the EDA. EPA has presented data todemonstrate that the low levels and the lack ofpatterns of contamination it found are consistentwith levels found in industrialized areas nation-wide. However, it is not clear how DHHS usedthis information.

Like others who have examined the results ofthe EPA monitoring study, OTA questioned thereliability of the study measurements, whichfound low values for most chemicals detected inthe EDA; moreover 90 percent of all measure-ments found only trace amounts, or no detectableamounts, of contamination.

For the low values reported, the main issue isthe validity of the values and the uncertainty thatmight exist in such values. EPA reported resultsas parts per billion (ppb) and did not report resultsas ppb plus or minus some value. The fact that“plus or minus” values are lacking means that thestudy provides no information on a possiblespread in the detected levels. Such a spread couldresult from single or compounded errors in theentire chain of sampling, and analysis of the sam-ples.

Closely related to this issue is the level of con-tamination which is judged to be significant tohuman health. Human health effects are differentfor different chemicals, and they are also differentfor the same chemical in different environmentalmedia because of differences in exposure oppor-tunities. For example, suppose that 100 ppb is thevalue, for a specific combination of chemical andenvironmental medium, below which health ef-fects are not considered important or likely. Inthis case, a finding in a sample of 50 ppb plus orminus 10 ppb would be a firm basis for a deci-

sion that health is not likely to be affected. Butif the result is 50 ppb plus or minus 40 or 50 ppb,then such a decision becomes much less certain.And in fact, it is often difficult to achieve highlevels of certainty in associating a health effectwith a given contamination level.

Considering the uncertainties in data and esti-mates of health effects, as well as in detectionlevels, the combination of the two introduces sub-stantial uncertainty into a decision dependent onboth. All technical data have some uncertainties;nevertheless policy decisions can make use of suchdata. The issue is: How much uncertainty exists?The uncertainties for the EDA monitoring dataand with the health effects information used byDHHS are high, and they make policy decisionsbased on these technical inputs open to continu-ing debate.

EPA reported that its positive findings reliablyindicated contamination in the EDA no higherthan the low ppb range. NBS, which examinedthe EPA monitoring study to assess the adequacyof the analytical methodology, quality control andquality assurance programs, did not support theEPA contention. Based on a review of their anal-ysis, OTA believes the NBS assessment is validand has not duplicated it. NBS said:

The methodology selected and used by EPA isappropriate for measuring concentrations in thelow parts-per-billion range for air and watersamples. However, using appropriate meth-odology does not guarantee reliable results. Thus,the question that remains is the level of perform-ance of the laboratories conducting the analyses.At the low parts-per-billion level, the contractlaboratories displayed wide variability in per-formance. Well-documented statements of preci-sion and accuracy are critical since these providethe only valid basis for assessing the meaning ofthe numerical data to those who wish to drawtheir own conclusions from the report. Withoutsuch documented statements of precision and ac-curacy, the results of measurements are of limitedusefulness for making comparisons within andamong sites. 3 [Emphasis added. ]

3From a letter by Raymond G. Kammer, Deputy Director, Na-tional Bureau of Standards, Aug. 30, 1982, sent to Senators DanielP. Moynihan and Alfonse M. D’Amato and Congressman John J.LaFalce. This letter is the most recent statement from NBS and waswritten after EPA responded to the earlier comments of NBS, andalso followed congressional hearings on the subject.

12

NBS also noted that the "limitations in the state-of-the-art for measuring biota and soils and sed-iments resulted in EPA appropriately focusing itsstudy principally on air and water. ” What all thismeans is that there remains troubling uncertain-ty regarding the results of the monitoring studyindicating low levels of contamination in theEDA. Some of the positive results said to be inthe low ppb range maybe as high as several hun-dred ppb because of variability and uncertainty.Of even greater uncertainty are the 90 percent ofthe results termed “not detected” or “trace. ” Forthese results, NBS continued, in the letter quotedfrom above:

Unless measured values, including “none de-tected, ” are accompanied by estimates of uncer-tainty, they are incomplete and of limited useful-ness for further interpretation and for drawingconclusions. For these reasons, performance" . . . in the low parts-per-billion range” has notbeen demonstrated to our satisfaction in the docu-mentation.

Based on knowledge of the analytical meth-odology used by EPA and the documentation pro-vided by EPA, NBS has no reason to believe thatmeasurements labeled “none detected” or “trace”represent concentrations above one part-per-mil-lion. [Emphasis added. ]

DHHS has taken the results of the EPA mon-itoring study, including EPA’s reply to NBS con-cerns, to mean that the chemicals present in theEDA are present in amounts ranging from lowppb to perhaps several hundreds of ppb. Further-more, except for the locations in which high levelsof contamination, primarily dioxin, were found,DHHS has made the judgment that less than 1-part-per-million levels support a positive findingof habitability; i.e., that contamination is for themost part so low and so unexceptional as to ruleout possible health effects for those choosing toreside in the EDA.

OTA believes that the persistent concerns ofNBS remain valid. EPA has not yet provided allthe considerable details of its monitoring studythat would permit outside experts to reach EPA’sown level of confidence in its data. Thus, levelsof contamination may or may not be consistent-ly as low as DHHS concludes, and as EPA assures

them to be. Moreover, OTA has had difficultyin assessing the foundation for the DHHS deci-sion, as few details have been released (e.g., whathealth effects data for specific chemicals and en-vironmental media were considered, and howthese were linked to specific results of the EPAmonitoring study). DHHS faced problems withEPA’s monitoring study because details were ab-sent from the publicly available documentation.Ironically, this is now the case with materialsmade available by DHHS.

DHHS has not explained how it considered thepotential for synergistic health effects from themany toxic chemicals at Love Canal. Synergismmeans that low levels of contamination for severalchemicals may combine to pose health threats,even though the same chemicals present in-dividually at the same low levels of contamina-tion might not be considered threatening. Thiscomment should not be interpreted to mean thatmuch information on possible synergistic effectsexists. Unfortunately, there are few data onsynergistic effects for the many toxic chemicalswhich may contaminate the EDA. However, thislack of information contributes to uncertainty forpossible health effects. The importance of thisuncertainty to DHHS is not clear.

It is also possible that for a few chemicals (e.g.,hexachlorobenzene), contamination at levels ofonly hundreds of parts per billion may pose healththreats. Such levels, because of uncertainties, mayhave been present in samples recorded as havingno detectable or trace amounts. Of particular con-cern is uncertainty about the “no detect” resultsfor some two-thirds of samples tested for dioxin.First, the number of samples tested for dioxin wasvery small, only four soil samples in the entireEDA of over 200 acres were taken. Moreover, theproblems of analyzing for dioxin are well known.They may have been less known and, perhaps,even worse in 1980 than they are today.

Finally, the semantics and logic of the formula-tion of the habitability issue raise questions. Sodo the statements by DHHS on its habitability de-cision. DHHS defined the task for itself and itsconsultants in this way:

13

Based on available data, can it be concludedthat the area is not habitable?4

This formulation of the habitability decisiontask seems to demand one of two answers: eitherthe EDA is to be demonstrated as unsafe or it ishabitable. * In light of this formulation, DHHSsummarized its interpretation of the findings ofthe panel of 11 outside experts who advisedDHHS staff, as follows:

. . . a majority of the consultants concluded thatbased on the data available they could not con-clude that the area was not habitable. 5

However, as indicated earlier, the inability toconclude that the EDA is not habitable does notnecessarily imply that the EDA is habitable. Itmay be questioned whether the findings of theDHHS panel of consultants, as quoted above,support the statements DHHS ultimately made onhabitability.

The most recent statement on habitability, bya senior member of the DHHS team that madethe original habitability decision, includes noreference to control areas and is put in unusuallypositive terms:

Review of these [monitoring] data by the De-partment of Health and Human Services, withevaluation of technical methodologies by the Na-tional Bureau of Standards, led to federal recom-mendations in July 1982 that the general area sur-rounding the Love Canal was safe for human resi-dence outside the canal itself and the two ringsof homes surrounding it. It was also recom-mended that the storm sewers and their drainagetracts be cleaned and that special plans be madefor perpetual maintenance of the clay cap cover-ing the site.6

4 Statement of Edward N. Brandt, Assistant Secretary for Health,Department of Health and Human Services, hearing of Subcommit-tee on Commerce, Transportation, and Tourism, House Commit-tee on Energy and Commerce, Aug. 9, 1982.* This formulation of the problem might be interpreted different-

ly; i.e., that it demands a lesser burden of proof than a formulationwhich begins with the premise that the area is unsafe. Detecting con-tamination is, at least theoretically, easier than proving that an areais free of contamination. However, this is only the case if there isprecise design of the study and use of procedures which yield lowlevels of uncertainties for experimental results.

5Brandt, op. cit.6Clark W. Heath, “Assessment of Health Risks at Love Canal, ”

Fourth Annual Symposium on Environmental Epidemiology, May2-4, 1983, Pittsburgh, Pa. (Dr. Heath was the senior author of theDHHS habitability statement in July 1982. )

Relative to our earlier expressed concerns re-garding cleanup and maintenance of the site, thisstatement lacks the strong contingency elementof the original DHHS statement on habitability.Instead, it poses cleanup and maintenance needsas separate from the habitability conclusion. Itprovides less assurance that the site will in factbe cleaned up and the cleanup maintained; a con-cern which is emphasized in the OTA analysis.

Heath of DHHS notes that “one would hopethat this could be done while restoring the sur-rounding neighborhood to normal activity.” Thisrather general statement not only retreats fromthe implication that cleanup is a prerequisite tohabitability, it fails to address the potential prob-lems of cleaning up the locations contaminatedwith high dioxin levels without threatening thehealth of nearby residents.

The results of studies on health effects andchromosomal damage in Love Canal area resi-dents may appear relevant to the DHHS habita-bility decision. For the most part, these studieshave found little positive evidence of healthdamage. However, the following conclusion andcaveat by Heath should be noted:

. . . it can be said from current epidemiologicdata available at Love Canal that no striking in-creases in illness occurrence have thus far ap-peared in association with living near the canal.This does not mean that such occurrences mightnot yet appear or that some canal-related illnessmay not have occurred but at frequency levels notdetectable by the studies performed.

The present state of uncertainty about healtheffects for Love Canal area residents is not unique.It is common. That is why so much public policyin the environmental protection area is precau-tionary in nature. To wait for conclusive evidencefor adverse health effects in people would meanthat people could be unnecessarily exposed to tox-ic

4.

chemicals.

OTA’S analysis of some of the data obtainedin the EPA monitoring study provides limited,but not conclusive, indication that there maybe contamination in the EDA by toxic chemi-cals from Love Canal.

14

Despite the limits and uncertainties of EPA’smonitoring study previously discussed, OTA con-sidered the possibility that the positive detectionsfound might be analyzed so as to better discernwhether significant contamination existed in theEDA. Although some interesting, suggestive re-sults were obtained regarding contamination ofthe EDA, by no means are these results strongenough to support a conclusion of nonhabitabili-ty. The results also suggest how future monitor-ing studies may be designed to produce more cer-tain results.

OTA disaggregated the EPA monitoring datato allow an examination of a small universe ofdata, what we term “indicator compounds.” Theseare toxic chemicals which are known to have beendisposed of in Love Canal and which the monitor-ing study looked for in the control areas and theEDA. A statistical analysis was performed forOTA to determine whether any observed differ-ences in detections of contaminants in the EDAas compared to the control areas were statistical-ly significant. Four chemicals (1,2- and 1,3-di-chlorobenzene and 2- and 4-chlorotoluene) werefound to be present in the EDA at significantlygreater frequencies than in the control areas.However, the levels found for these chemicalswere quite low.

Why was it possible for OTA to find these sta-tistically significant differences between detection

of contaminants in the EDA versus the controlareas when EPA came to a general conclusion thatthe EDA was not significantly more contaminatedthan the control areas? The chief answer is thatOTA combined monitoring data for all environ-mental media tested, thereby enlarging the database per chemical. EPA analyzed data per chemi-cal for each submedia tested, which meant thatthere were very few data for the control areas.In this situation there were too few data from thecontrol areas to reveal statistically significant dif-ferences with the EDA. However, for severalchemicals EPA did find several significantly higherfrequencies of detection in the EDA samples ascompared to the control areas. Essentially, EPAdiscounted the few positive findings of significant-ly more frequent contamination in the EDA be-cause the number of negative findings was farlarger. This preponderance of negative findingswas based on the large number of chemicals whichEPA monitored in the study—about 150 chemi-cals, including 129 which are considered to be“priority” pollutants for general regulatory pur-poses but, by and large, have no history of dis-posal in Love Canal. In other word, with regardto its analysis and interpretation of data, EPAdirected a substantial portion of its efforts towardchemicals that are not necessarily unique and im-portant in the Love Canal situation.

POSSIBLE STEPS TOWARD REHABITATION

In making decisions on the habitability of theEDA, there are choices other than immediate,complete rehabitation and a continued presump-tion against people moving back into the area.A number of actions could be taken to movetoward incremental rehabitation of the EDA. Byincremental rehabitation we mean a paced, cau-tious approach resting on improvements in thescientific certainty for conclusions about the safetyof the EDA, and on increased public confidencein policy decisions based on technically complexdata and analyses. Habitability need not be seen“as an all or nothing” issue.

ing study unfortunately did not resolve, is thatsome portions may not be contaminated, whileothers may be. Over time, portions of the EDAmay be found, with a higher degree of confidence,to be free of contamination. Depending on theirlocation, they may then be judged habitable. Onecomplexity is that if some areas are found to becontaminated, then it will be necessary to cleanthem up and to assess the effect of the cleanupactions on uncontaminated and, perhaps, rehab-itated areas.

Four critical actions merit further consideration:

In fact, one of the difficulties in assessing the 1. The technical problems and uncertainties inhabitability of the EDA, which the EPA monitor- the current cleanup activities and long-term

15

2.

3.

plans for operation and maintenance of thecontainment system should be removed.There must be effective cleanup of areasknown to be contaminated with dioxin.The uncertainties of institutional stabilityand effectiveness over very long terms (i.e.,hundreds of years) should be addressed.There may be ways to assure that the fundsnecessary for indefinite monitoring, opera-tion and maintenance of the containmentsystem, and for possible corrective actions,will be available over the long term. It maybe possible to remove uncertainties relatedto the ownership, title, and future use ofLove Canal. There are probably ways to as-sure that critical information is retained andmade accessible over long periods.It should be possible to design well-focusedmonitoring studies to be conducted for care-fully defined areas of the EDA, perhaps forindividual homesites. To some extent, theEPA’s 1980 monitoring study can serve afunction similar to that of a pilot study. Thenew studies could monitor for fewer chemi-cals, sampling can be improved, and ana-lytical results can be presented in ways thatremove uncertainties over their reliability

4.

and accuracy. Costs are a concern, but mayprove to be a reasonable investment. It ap-pears that additional monitoring for in-dividual homesites (including several samplesfor dioxin and multiple samples for chemicalsknown to have been deposited in Love Ca-nal) might have a one-time cost of about$5,000 per site.Finding a permanent solution for the largeamounts of long-lasting toxic wastes still inLove Canal is highly desirable. The presentcontainment approach may best be viewedas an interim solution. Plans could be madeto:●

●

track technological developments for per-manently destroying or detoxifying thewastes, andassess the technical feasibility and eco-nomic cost effectiveness of applying suchdevelopments at Love Canal.

Finally, EPA should provide considerably great-er detail on the results of its monitoring study,which might resolve the uncertainties raised byNBS in its examination of currently availabledocumentation.

IMPLICATIONS FOR THE SUPERFUND PROGRAM

This case study provides insights into severalissues involved in the Federal Superfund programfor cleaning up uncontrolled hazardous wastesites.

1. The “ HOW Clean Is Clean?” Issue

Technical standards to determine unacceptablelevels of contamination by toxic chemicals are,on the whole, lacking. Some standards do existin various environmental programs, but there isno overall set of standards for the broad rangeof toxic chemicals associated with hazardouswastes. Having such standards, however, does notrule out the use of site-specific information (e.g.,concerning migration routes and potential ex-posure levels) to arrive at habitability decisions.

Without standards, Government agencies mustmake ad hoc decisions. Uniform protection na-tionwide is unlikely, and decisions may notalways be technically valid. This Love Canal casestudy illustrates three uses for standards for ac-ceptable (and unacceptable) levels of contami-nants:

●

●

With such standards, technical informationon contamination, even if only from pilot orpreliminary monitoring, could provide thebasis for decisions on relocation of residentsand the nonhabitability of areas.Standards could be very useful in designingdetailed monitoring studies to assess the fullextent and level of contamination. The prob-lem could be defined as assuring that stand-

16

ards are not exceeded, rather than detectinganything that might be present. Interpreta-tion of the results of monitoring studieswould also be improved. Standards wouldprobably make the use of control areas un-necessary.

● The choice of cleanup technologies could bemore effective, if based on standards that settargets and goals for the cleanup action.

2. Health Effects Data and Decisionson Habitability

Better understanding of the health effects of tox-ic chemicals is a critical need, particularly on thevarious exposure routes for toxic chemicals in haz-ardous waste sites. Using available data, it is pos-sible to establish some standards for acceptablelevels of contamination, but more data are neededto establish more reliable and complete standards.Health effects data can be collected from conven-tional research, but more effective implementa-tion of the congressional mandate in the Super-fund legislation is also needed to expand the epi-demiological data base for health effects in peo-ple already exposed to toxic chemicals. Moreover,policy decisions on evacuation and relocation ofresidents, and on (re)habitability, need a firmerunderpinning of administrative and analytic pro-cedures. The responsibility given to DHHS for theLove Canal EDA was not very clearly defined inthis respect. For example, no detailed analysis sup-porting the DHHS decision was provided. Therole of EPA relative to DHHS in substantiatingand reaching habitability decisions, and perhapsthe use of consultants and advisory panels byDHHS, needs further examination. In the case ofLove Canal, interactions between EPA and DHHSthat could have influenced the design and im-plementation of the monitoring study were notpossible because DHHS did not play a role untilafter the study was completed. If DHHS is to con-tinue to be responsible for decisions on relocationand habitability, then it may be appropriate forit to participate more actively in the design andimplementation of monitoring studies, and theanalysis of their results.

3. Technical Guidelines forMonitoring Studies

The entire experience in the EDA monitoringprogram underlines the need to develop appro-priate sampling and analytical protocols for anumber of toxic chemicals and environmentalmedia. It is also highly desirable to establish re-quirements for the presentation and documenta-tion of results of monitoring studies, includingestimates of error and uncertainty.

4. Selection and Implementation ofRemediation Programs

A very important implication of this case studyfor the Superfund program concerns the statutori-ly required analysis of alternative cleanup ap-proaches. At Love Canal, and apparently at manyother Superfund sites, there has been no detailedconsideration of permanent solutions. The cost-effectiveness study of which cleanup methods touse at Love Canal considered alternatives forisolating or containing the hazardous wastes inthe canal, but none that would destroy or detox-ify the wastes. Methods to accurately evaluate therelative cost effectiveness of containment versuspermanent solutions require more study.

The containment method adopted imposes op-erating and maintenance costs for all time(presumably for which New York State is respon-sible). It is also subject to technical uncertaintiesabout possible future failures and release of tox-ic substances into the environment.

This leads to an exceptionally important con-sequence of the Love Canal experience for theSuperfund program: the need to compel consid-eration of more permanent solutions, albeit withhigher capital costs (to be paid mostly by theSuperfund program), in analyses of alternativecleanup approaches. Some technical approachesfor onsite destruction and detoxification of toxicwastes and contaminated materials are availabletoday, and considerable effort is going into thedevelopment of more such techniques. Even if apermanent solution does not appear technicallyor economically feasible for a specific site at the

17

outset, there is a need for a formal, continuingprogram to evaluate relevant scientific researchand developing technological opportunities forsuch permanent solutions, site by site. The grow-ing trend to relocate residents away from criticaluncontrolled waste sites, followed by containmentof wastes and contaminated materials, means con-tinuing difficulties in establishing long-term safetyfor rehabitation.

Another apparent need is for improved over-sight by EPA of State implementation of selectedremediation programs. In the case of the EDA,the delay in cleaning up areas known to be highlycontaminated with dioxin—still not carried outseveral years after its discovery-is disturbing.However, New York State has faced a numberof critical tasks at Love Canal as well as at otheruncontrolled hazardous waste sites that have un-doubtedly strained its resources.

5. Long-Termand Issues

Institutional Capabilities

Love Canal may have yet another unique anduseful historical role. Some Government regula-tions recognize the long-term hazards of nuclearwastes and, reflecting concern for future genera-tions, impose requirements for assured isolationover thousands of years. Yet there has been verylittle attention to the long-term hazards of toxicwastes, and the extreme uncertainties of contain-ment approaches for controlling them. This ne-

glect is paradoxical and inconsistent. Many of themost toxic wastes will retain their hazardous char-acteristics indefinitely. Unlike nuclear wastes,most such toxic wastes (molecules) have no in-herent half-lives; i.e., they will remain stableunless they degrade through environmental in-teractions. Yet Government regulations for newland disposal facilities have well-defined re-quirements extending only 30 years. In addition,the Superfund program has given no considera-tion to the capability of institutions to carry outessential tasks of controlling toxic waste sites forhundreds or thousands of years.

In considering the long-term effectiveness of thecurrent remediation plan for the EDA, it becameapparent that there are currently no mechanismsto assure indefinite funding for monitoring, foroperation and maintenance, and possibly for cor-rective actions for the waste containment systemat Love Canal. Considering the recent failures ofmany public and private institutions to controltoxic wastes adequately, it is unreasonable to ex-pect the public to have confidence in a Superfundprogram that makes extensive use of long-termwaste containment but provides no assurances forthe long-term effectiveness of those “solutions.”

It is likely that waste containment at Superfundsites will continue to be used until there are morewidely available and low-cost technological alter-natives to destroy or detoxify wastes and contam-inated materials onsite. The issues of long-terminstitutional effectiveness cannot be escaped.

21-893 () - 83 - ~

Appendixes

Appendix A

Remediation

Summary

There are four areas of uncertainty that can affectprojections of the long-term integrity of the remedialtechnology. Before any decision on habitability canbe made, these uncertainties must be addressed andsolutions identified.

1. Remedial Action in the Emergency DeclarationArea (EDA).—The areas in the EDA contaminatedwith high levels of dioxin have not yet been cleanedup. Moreover, until just a few months ago stormsewers leading from the canal region to the EDA andknown to contain dioxin remained open. It is possi-ble that during the past few years—after completionof the EPA monitoring study—dioxin may have beentransported within or beyond the EDA. A study todetermine the full extent of contamination in and nearthe sewers is not completed.

2. Leak Detection Systems.—The long-term integri-ty of the remedial technology is not certain. Reliablemethods are needed to allow detection of damage(leading to permeability) to the two basic elements ofthe containment system. These elements, intended tominimize water entering the canal, are the cap overthe canal area and the concrete barrier wall to be builtaround it. There is no dispute about the need for repairand replacement of the cap and leachate collectionsystem over time. Yet how it will be done is not clear.How structural damage or clogging of the drain systemwill be detected, and how repair and replacement canbe carried out safely remain unanswered.

3. Monitoring Programs.—Assurance of sufficientwarning about any potential migration and accumula-tion of chemicals from the canal is essential. Plans areunderway for developing a long-term monitoring planfor ground water in the area immediately adjacent tothe canal but not in the EDA. It is also necessary todesign more extensive ambient monitoring of envir-onmental media other than ground water (e.g., air,soil, and biota). Media other than ground water arepossible routes of exposure to toxic chemicals. For ex-ample, depending on the properties of chemicals dis-posed in the canal and properties of the soil throughwhich the ground water moves, some chemicals couldbe filtered out and could accumulate in soil or possiblyin biota. Humans might become exposed to either. Inaddition, damage to the cap could allow release ofvolatile compounds into the air.

4. Institutional Mechanisms for Long-Term Protec-tion of the EDA Residents. —The fourth major areaof uncertainty concerns the long-term ability of gov-

ernment institutions to remember, fund, and carry outcommitments for long-term continued monitoring andmaintenance of the site. The full range of institutionalissues surrounding very long-term commitments formanaging uncontrolled hazardous waste sites underthe Superfund program have not been addressed. Cur-rent cost estimates for routine operation, maintenance,and replacement of the leachate collection system areabout $0.4 million now, $4.2 million in the year 2005,and $8.5 million in 2030. There is no guarantee thatState officials 20 or 100 years from now will eitherremember or honor this commitment. Furthermore,there are few institutional mechanisms in place toassure continuity in transferring vital information onLove Canal from one generation to the next. Nor doesit appear that New York State has taken unequivocal,binding, and permanent title of the canal area in amanner that prevents future use of the site.

Categories of Remedial Technology

Technical options for remedial action implementedunder the authority of the Comprehensive Environ-mental Response, Compensation, and Liability Act of1980 (CERCLA) can be categorized as either waste con-trol or environment control.l Table A-1 lists the typesof technologies in these two categories and illustratessome of the advantages and disadvantages of each.The implementation of any of them will depend onsite-specific conditions. In some situations, a combina-tion of waste and environment control strategies wouldbe required.

Waste control refers to the removal of the hazard-ous material from a site, followed by some treatmentthat reduces the potential harm of hazardous com-pounds and subsequent disposal of the waste or treat-ment residue in an appropriate facility. The treatmentcan involve destruction of toxic components of the ex-cavated material through chemical, physical or bio-logical processes, or immobilization of the hazardouscomponents.

At present, the application of destruction techniqueshas been limited to excavated materials or small areaspills treated by biodegradation and chemical proc-esses. Some thermal destruction technologies are avail-able. Thermal destruction of large volumes of contam-inated material, such as excavated soil, is a new ap-

1 Technologies and Management Strategies for Hazardous Waste Control,

“Chapter: Technologies for Hazardous Waste Management: UncontrolledSites” (Washington, D. C.: U.S. Congress, Office of Technology Assessment,OTA-M-l96, March 1983).

21

22—

Table A-l.—Advantages and Disadvantages of Control Technologies

Type Advantages Disadvantages

Waste control technologiesExcavation and removal followed • Good for containerized or bulk disposal • High initial costs

by treatment or disposal ● Potential higher risk during cleanup• Relocation of risk unless waste is

treated. Not cost effective for low-level

hazardous waste or uncontainerizedburied waste in large area

Excavation with onsite treatment ●

Neutral ization/stabilization

Expose waste to complete treatment ●

●

●

●

Useful in areas where waste excavated ●

prior to mixing ●

Low risk of exposure if injection methodis used ●

High initial costDifficult to assure monitoringeffectivenessSome risk of exposureNot cost effective for large amountof low-hazard wasteLimited applicationRequires long-term land useregulationsEventual off site migration if reactionis incomplete

Biodegradation ● Low costs ● Difficult to maintain optimumconditions to keep reaction going

Solution mining . Useful in homogeneous uncontainerized ● Can result in uncontrolled releasesolvent-soluble, buried solid hazardouswaste

Environmental control:

Isolation, containment, and ● Useful for large volumes of mixed . Effectiveness depends on physicalencapsulation hazardous and domestic waste and low- conditions

hazard waste ● Long-term O&M needed

Ground water diversion and ● Useful if soils are permeable or if there ● Requires wastewater treatmentrecovery are high perched water tables option

● Process is slow● O&M monitoring● Not effective for insoluble or

containerized material

Surface water diversion ● Easy to implement ● Can create flooding off site

● No transport of waste offsite

Ground and surface water ● Can be used onsite or offsite ● May generate hazardous sludges,treatment spent carbon

● Long-term monitoring

Gas collection or venting ● Low costs ● Site safety and fire hazards● Off site air pollution● Long-term monitoring and O&M

O&M—operating and maintenance.SOURCE: Office of Technology Assessment, op. cit., p. 210.

plication and data on its efficiency are limited. Thus,these technologies currently have not received wide-spread consideration as a remedial technology.

An alternative to destruction is the immobilizationof hazardous components. This is achieved by encap-sulating the excavated material in some impermeablematrix. When placed in soil or marine environments,migration of hazardous constituents is then prevented(or at a minimum, the rate of migration is decreased);

thus, the risk to public health and the environment isreduced.

These control technologies can be used effectivelywhen the waste has been deposited in containers andremoval from the site can be accomplished readily. Italso can be implemented at those sites where hazard-ous components have not become distributed through-out environmental media. For example, it is used atthose sites where bulk disposal has occurred and before

—

23

widespread migration of the material within the soilhas taken place. However, if the contaminated areais large, e.g., measured in several acres, waste con-trol techniques are difficult and costly to implement.

In the case of an accidental spill, removal and subse-quent treatment can be effective, if remedial action isnot delayed and boundaries of the spill can be iden-tified easily. Under appropriate site conditions, treat-ment techniques can be used without removal of thecontaminated material, e.g., in situ biological orchemical degradation of soil contaminated through anaccidental spill of hazardous chemicals.

Environment control options include those techni-ques that contain or isolate hazardous material, divertwater movement away from a site, or treat contam-inated water sources. A review of 23 landfill sites sug-gests that environment control is the more commonremedial strategy currently in use.2 The technologiesfor containment are not new; rather they are adaptedfrom structural or civil engineering procedures andconsist of the installation of caps, barrier walls, anddrainage systems. 3 4 At many sites, containment tech-nology is used in combination with water diversiontechniques. These latter include changing the flow ofsurface water to prevent flow into a contaminated siteor removal and treatment of ground water that hasbeen contaminated.

Because the strategy of environment control doesnot remove sources of contamination, it is necessaryto include safeguards that increase the likelihood oflong-term integrity of containment and reduce effectsof failure should it occur. Well-developed environmen-tal monitoring programs are essential safeguards.Monitoring should include all environmental media:water, air, soil, and biota. Moreover, some sort of leakdetection system is necessary to warn of possiblerelease of contaminants through the natural or syn-thetic barriers of the containment structure.

When comparing these two categories of remedialtechnology, advantages and limitations can be iden-tified. For example, environment control offers certainadvantages over waste control in that large areas ofcontamination (e.g., many acres) can be controlled.In addition, installation costs are generally less for en-vironment control technologies than for waste control.Environment control technologies eliminate the poten-

2E. Nagle, Environmental Law Institute, personal communication, April1983.

3C. Kufs, et al., “Alternatives to Ground Water Pumping for ControllingHazardous Waste Leachates,” Management of Uncontrolled Hazardous WasteSites, Hazardous Materials Control Research Institute, 1982, pp. 146-149.

4P. A. Spooner, R. S. Wetzel, and W. E. Grube, “Pollution Migration Cut-off Using Slurry Trench Construction, ” Management of Uncontrolled Haz-ardous Waste Sites, Hazardous Materials Control Research institute, 1982,pp. 191-197.

tial transfer of risk from one area to another; for somewaste control options this transfer of risk is a majorconsideration. Moreover the use of environment con-trol technologies does not create risks for transporta-tion accidents. Environment control, however, re-quires long-term (i.e., forever) operation andmaintenance. In contrast, waste control includestreatments that completely destroy the hazardous ma-terial, eliminating long-term hazards. Both environ-ment control and those waste control options that onlyimmobilize hazardous components must include long-term monitoring programs.

A major concern associated with either type ofremedial action is the limited experience with thesetechniques. Sites using either waste or environmentcontrol have not been in existence long enough to pro-vide sufficient data about the long-term integrity ofthe methods. For example, a review of sites where en-vironment control has been in place indicates that the“oldest site” has had a clay barrier wall (5-to 8-ft thick)only since 1976.5 Monitoring at this site has not yetindicated leakage through the wall. Remediation at theoldest sites incorporating barrier systems using syn-thetic materials (e.g., asphalt-bentonite, cement-bentonite) were completed only in 1979. 6 Thus, ourexperience regarding long-term integrity of contain-ment technology is limited.

Uncertainties exist for waste control options, also.Unless the extent of contamination can be character-ized in detail, i.e., the types and concentrations of allconstituents are known, complete destruction of haz-ardous elements cannot be validated. New constituentscould be formed as products of the biological, chemi-cal, or thermal processes taking place. These new con-stituents could be as, or more hazardous than, theoriginal compounds.

Much theoretical work has been done to predict theperformance of remedial technology. While the infor-mation gained through the use of theory and modelsis important, it must be emphasized that at present nofield experience exists. The persistence of many wasteconstituents is much longer than the effective lifetimeof the environment control technologies; the degreeof hazard for components in wastes may be increasedby waste control treatments. Thus, environment con-trol may simply postpone risks to public health andthe environment to future generations, and waste con-trol may create new hazards.

5Nagle, op. cit.6 Ibid.

24

Evaluation of Alternative Technologiesat Love Canal

In accordance with CERCLA requirements, EPA dida cost-effectiveness analysis of alternative technologiesfor remedial action at Love Canal. Their analysis con-sidered only environment control technologies. OTAidentified factors that are relevant to consideration ofwaste control options:

1.

2.

3.

4.

Given the large area of the landfill and adjacentland, waste control technologies likely would becostly, possibly greater than environment con-trol by orders of magnitude. For example, ex-cavating and treating 49 acres of contaminatedsoil to a depth of possibly 15 ft (equal to nearly2 million tons of contaminated soil) would be amajor and expensive task with current technolo-gies, particularly in water saturated zones.Workers as well as residents in the EDA wouldbe exposed to hazardous substances through theexcavation process and formation of potentiallyhazardous products by operation of a waste treat-ment system.Given the broad range of chemicals that wereoriginally dumped in the canal and the varietyof products that could result from natural andenhanced degradation as well as thermal combus-tion processes, the outcome of destruction effortsis uncertain with present technology. Demonstra-tion studies would be required to evaluate the effi-cacy of the waste control treatments. Thesestudies would delay completion of remedial ac-tion and possibly increase the risk to residents re-maining in the EDA.The problem of finding an ultimate disposal sitefor treatment residue would be difficult to resolvewithout knowing its hazardous quality. Disposalof such residues in a new site could result in mere-ly relocating health and environmental risks.

The environment control alternatives considered byEPA included four altematives:7

1. No additional action beyond operation andmaintenance of a leachate collection system.

2. Cut-off and plugging utility lines, in addition toalternative 1. All utility conduits that are possi-ble routes for lateral movement from the sitewould be plugged and all utility lines beyond thecontainment areas would be cleaned.

3. Alternatives 1 and 2, plus installation of a par-tial wall. A subsurface wall would be installed

‘CHJWfi”14 Immediate Remedial Action-Uncontrolled Hazardous Waste=lSites, zone L pdiminarydraftnwort tbr U.S. EPA Region IZ April

at points of natural migration routes from thesite—e.g., sand lens or drainage swales.

4. Alternatives 1 and 2, plus complete containmentof the contaminated area. Construction of a bar-rier wall that would completely enclose the site.

A summary of the lifecycle costs for these alternativesis presented in table A-2. Although initially the costsare greatest for option 4, over a period of 50 to 200years this alternative is expected to result in the lowesttotal cost to the State. As indicated in table A-3, eachalternative was also evaluated based on expectedperformance. Alternative 4 provides the greatestrelative protection for public health and theenvironment.

Table A-2.—Summary of Lifecycle Costs(present worth In 1981 dollars-1 x 10a)

1 year 50 years 100 years 200 years

Alternative 1:Capital . . . . . . . . .O&M . . . . . . . . . . .Replacement. . . .

Total. . . . . . . . .Alternative 2:

Capital . . . . . . . . .O&M . . . . . . . . . . .Replacement. . . .

Total. . . . . . . . .Alternative 3:

Capital . . . . . . . . .O&M . . . . . . . . . . .Replacement. . . .

Total. . . . . . . . .Alternative 4:

Capital . . . . . . . . .O&M . . . . . . . . . . .Replacement. . . .

Total. . . . . . . . .

0.25—

0.25

0.610.25

—

0.86

1.990.20

—

2.19

2.550.14

—

2.69

—12.73

1.0413.77

0.6112.73

1.0414.38

1.9910.090.79

12.87

2.557.021.49

11.06

—25.46

7.29

—50.9220.19

32.75

0.6125.467.29

33.36

1.9920.178.17

30.33

2.5514.04

1.49

71.11

0.6150.9220.1971.72

1.9940.3422.02

64.35

2.5528.0821.35

18.08 51.98NOTE: Alternative 1—No additional action beyond installa-

tion of Ieachate collection system.Alternative 2—Utility cut-off containment.Alternative 3—Partial slurry well containment.Alternative 4—Complete slurry wall containment.

SOURCE: CH2M-Hill, op. cit.

When Congress included the requirement ofconducting cost-effective analyses in the Superfundlegislation, the intent was that both waste and environ-ment control alternatives would be considered. Whileit is apparent that alternative 4 is preferred over alter-natives 1 through 3, OTA questions the omission ofsome consideration for of any waste control technol-ogy in the cost-effectiveness analysis. As indicatedabove, present waste control technology cannot han-dle efficiently the large volumes of contaminatedmaterial that exist at the Love Canal site. Therefore,choosing environmental control options makes sense

25

Table A-3.—Performance Criteria Evaluation

Ranka b c

Criterion Alt. 1 Alt. 2 Alt. 3 Alt. 4

Initial cost . . . . . . . . . . . . . . . . . .O&M cost . . . . . . . . . . . . . . . . . .Lifecycle cost . . . . . . . . . . . . . . .Long-term environmental impactShort-term environmental impactConstruction site health and

safety . . . . . . . . . . . . . . . . . . . .Community health and safety .Technical reliability . . . . . . . . . .System reliability . . . . . . . . . . . .Community acceptance . . . . . .Construction duration . . . . . . . .Achieve objectives . . . . . . . . . . .Meet project bid date . . . . . . . .

13341

14144141

2 33 24 23 22 3

2 33 22 23 23 22 33 22 2

41114

aRanking ranges from “1” = best to "4" = worst. Equal rankings denoted byequal low numbers.

bAlternative 1 . No additional action beyond Ieachate collection system.Alternative 2 - Utility cut-off containment.Alternative 3 - Partial slurry wall containment.Alternative 4 . Complete slurry wall containment.

cNo weighting factors have been applied to performance criteria

SOURCE: CH2M-Hill, op. cit.

as a short to medium term action, pending develop-ment of technology to deal permanently with thematerial. However, environment control cannot andshould not be considered a long-term or permanentsolution.

Many people have cited the great uncertainty of as-suring long-term protection using environment con-trol technologies. No effective alternative has been ad-vanced. New York State officials are convinced thatgreater efforts should be expended on research anddevelopment of detoxification and destruction techni-ques thus, eliminating the need for long-term com-mitments to protection of a large land area. Once thesetechnologies have been developed, they must be givenserious consideration in any cost-effectiveness analysesfor remedial technology. Although waste control op-tions might be extremely costly to implement, it ispossible that these would compare favorably with totalcosts over a 200-year time period for environment con-trol options. In addition, the complete elimination ofthe hazard due to waste control treatments mayoutweigh objections to the high cost for implementingthis type of remedial action and the short-term risksto workers and residents due to excavation of thematerial.

Control Action at the Love Canal Site

Because of the large area involved and environmen-tal distribution of wastes disposed in the canal, theremedial action chosen by EPA and the New York

State Department of Environmental Conservation(NYS/DEC) follows a strategy of environment con-trol. Two types of technologies are used: a leachatecollection system, for which construction began in1978; and a containment system, for which construc-tion work was planned for June 1983. These technol-ogies are commonly used for remedial action.8 910

The drainage system became operational in 1979 andis to continue indefinitely with planned repair andreplacement. The system consists of a clay cap cover-ing the immediate area of the original landfill; a Frenchdrain system rings the cap enclosing an area of approx-imately 23 acres.

Ground water migrates through the site into thedrainage system, is pumped into an onsite treatmentfacility, and put into clarification tanks where waterand sludge phases are separated. The average flowthrough the system is 8 gallons per minute (gpm); themaximum capacity is 200 gpm and peak flows of 48gpm have been recorded during the wet Season,ll Thewater phase is drawn through an activated carbonsystem and effluent discharged into the municipalsewage system.

Effluent standards have been established by the Cityof Niagara Falls specifically for discharge of effluentfrom this facility. For every day that the treatmentfacility is operational, analyses are performed to deter-mine whether these standards are being met. Analysesinclude tests for the presence of priority pollutants (seetable A-4) determination of effluent pH levels (the ef-fluent is neutral), and analyses of levels of total organiccarbons (tests the efficiency of the activated carbonsystem), and total chlorinated hydrocarbons. A reviewof available data on constituent levels within treatedleachate indicates that the highest value recorded forany of the priority pollutants was 46 parts per billion(ppb).12 Even this low concentration has been detectedonly occasionally; most results of the analyses indicateno detection. State and city officials consider valuesin the ppb range to be sufficiently low because the ef-fluent receives further treatment in the public waste-water treatment system. 13 Residues from the treatment

8Spooner, Wetzel and Grube, op. cit.*J. C. Evans and H. Fang “GeotechnicalAspects of tk De@ and Con-

struction of Waste Containment Systems, ” Management of UncontrolledHazardous Waste Sites, Hazardous Materials Control Research Institute, 1982.

IOHand~ok ~r Reme~”a] Action at Waste Disposal Sites (Washington,D. C.: U.S. EPA Municipal Environmental Research Laboratory, EPA-625/6-82-006, June 1982).

llCHzM-Hi]l, op. Cit., P. 3-1.IZW. J. McDoug~], R. A. FUSCO, and R. P. O’Brien, “containment and

Treatment of the Love Canal Landfill Leachate,’’Joumaf WPC’5 vol. 52, No.12, 1980, pp. 2,914-2,924.

I+J. Kolak, New York State Department of Environmental Con*rvation,

personal communication, March 1983.

26

Table A-4.—Priority Poiiutants

Volatile organic compoundsAcroleinAcrylonitrileBenzeneCarbontetrachlorideChlorobenzene1,1-Dichloroethane1,2-Dichloroethane1,1,1-Trichloroethane1,1,2-Trichioroethane1,1,2-2-TetrachloroethaneChloroethane2-Chloroethylvinyl etherChloroform1,1-Dichloroethylene1,2-Trans-dichioroethyiene1,2-Dichioropropane1,3-DichioropropeneEthylbenzeneMethylene chlorideMethyl chlorideMethyl bromideBromoformDichlorobromomethaneTrichlorofluoromethaneChlorodibromomethaneTetrachloroethyleneTolueneTrichloroethyleneVinyl chlorideBis (chloromethyl) ether

1,2-DiphenylhydrazineFluoranthene4-Chlorophenyl phenyl ether4-Bromophenyl phenyl etherBis (2-chloroisopropyl) etherBis (2-chioroethoxy) methaneHexachlorobutadieneHexachlorocyclopentadieneIsophoroneNaphthaleneNitrobenzeneN-nitrosodimethylamineN-nitrosodiphenylamineN-nitrosodi-n-propylamineButyl benzyl phthalateDi-n-butyl phthalateDi-n-octyl phthalateDiethyl phthalateDimethyl phthalateBenzo(a)anthraceneBenzo(a)pyrene3,4-BenzofluorantheneBenzo(k)fluoratheneChryseneAcenaphthyleneAnthraceneBenzo(ghi)peryleneFluorenePhenanthreneDibenzo(a,h)anthraceneIdeno(l,2,3-cd)pyrene

Base-neutral extractable organic compounds PyreneAcenacphthene Bis (2-ethylhexyl) phthalate

Pesticides and PCBsAldrinDieldrinChlordane4,4’-DDT4,4’-DDE4,4’-DDDa-Endosulfanb-EndosulfanEndosulfan sulfateEndrinEndrin aldehydeHeptachlorHeptachlor epoxidea-BHCb-BHCq-BHCw-BHCPCB-1242PCB-1254PCB-1221PCB-1232PCB-1248PCB-1260PCB-1016Toxaphene2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)MetalsAntimonyArsenicBerylliumCadmiumChromium

Benzidine1,2,4-Trichlorobenzene 2,4,6-TrichlorophenoiHexachlorobenzene Parachlorometa cresol MercuryHexachloroethane 2-Chlorophenol NickelBis (2-chloroethyl) ether 2-Nitrophenol Selenium2-Chloronaphthalene Pentachlorophenol Silver1,2-Dichlorobenzene 2,4-Dimethyphenol Thallium1,3-Dichlorobenzene 4-Nitrophenol Zinc1,4-Dichiorobenzene 2,4-Dinitrophenol Miscellaneous3,3’-Dichlorobenzidine 4,6-Dinitro-o-cresol Asbestos2,4-Dinitrotoluene 2,4-Dichlorophenol2,6-Dinitrotoluene

Total cyanidesPhenol

SOURCE: K. A. Brantner, R. B. Pojasek, and E. L. Stover, “Priority Pollutants Sample Collection and Handling,” Pollution Engineering, March 1981, p. 35.

process are presently being stored onsite for future cent data on water table elevations indicate thattreatment. The NYS/DEC plans to develop a pilot ground water in the region between the drain and theproject to investigate the potential for plasma arc in- landfill is being drawn into the collection system;cineration as a treatment process for the sludge.14 likewise, data for the area immediately beyond the

Hydrogeological assessments suggest that the leach- drain and adjacent to the EDA also indicate flowate collection system is operating successfully.15 Re- toward the collection system.16

————-Action Program Love Canal Project 1, Leachate Collection System, NiagaraFalls, New York Aug. 10,1982, prepared for Wald, Harkrader & Ross, Wash-

————.——— ington, D.C. According to Woodward-Clyde, their conclusions on groundI* For a d~riptic)n of this technology see Technologies ad Man%ement

Strategies hr Hazardous Waste Control, op. cit., pp. 172-173.water flow and the efficacy of the leachate collection system were essentiallythe same as CHzM-Hill.

15~= were ~~om~ by CHIM-Hill, contractors for U.S. EPA; their l~Data provided by J. L. Slack, NYS/DEC, May 19, 19S.3, duri% a meeti%report was not available to OTA. A separate evaluation of the system was with OTA, NYS/DEC, U.S. EPA, New York State Department of Healthmade by Woodward-Clyde Consultants, Evaluation of Proposed Remedial (NYS/DOH), and New York State Department of Law (NYS/DOL).

27

The containment component of the remedial actionwill involve the installation of a barrier wall aroundthe canal, encompassing an area of approximately 49acres. 17 The wall will be constructed of concrete (awidth of 24 inches) and will extend to a depth of about15 ft to be anchored into clay found at that depth. Thisclay is very impermeable with hydraulic conductivity(i.e., the rate at which water will move through thestrata) estimated to be in the range of 0.1 to 0.01 inchesper year.18 A synthetic membrane cap will be installedto cover the entire 49 acres, including the existing claycap. This membrane will extend beyond the barrierwall. Thus, it is expected that surface runoff will notpenetrate the enclosed area. Twelve inches of sterileearthfill will be placed on top of the membrane cover;6 inches of top soil will be the final cover. This topsoil will be grass seeded. All existing trees, shrubs, andother plants will have been removed from the areaprior to installation of the synthetic membrane cap.Only plants with a shallow root system can be allowedto be grown within the 49-acre area. Long rootedplants would eventually penetrate the cap.

The exact placement of the barrier wall will be deter-mined using two sets of data: results obtained fromthe 1980 Environmental Protection Agency’s (EPA)monitoring study and 1983 data on the extent of dioxincontamination in soils immediately surrounding thecanal. * EPA has concluded that major contaminationfrom Love Canal compounds does not extend beyondthe land immediately adjacent to the canal.**

According to Federal and State officials the wall willserve three purposes:19

1. The wall reduces the volume of water dawn intothe leachate collection system. Based on resultsfrom the 1980 EPA monitoring study, officialsassume that the water outside the 49-acre perim-eter is relatively clean—i.e., contaminants arepresent at concentrations of only parts per billionor less. By including this clean water in the col-lection system, the ongoing operation andmaintenance costs will be quite large (see tableA-2, alternative 1). These costs must be coveredby State funds into the indefinite future. Reduc-ing the volume of water that flows through the

17 Specifications for the barrier wall at Love Canal are provided inNYS/DEC, Love Canal fioject 1 Site Containment System, Niagara Falls,New York, vol. 1, August 1982.

1~L. R. Silka and J. W. Mercer, “Evaluation of Remedial Actions forGroundwater Contamination at Love Cam], New York,” Management ofLJncontrolkd Muar&us Waste Sites, Hazardous Waste Control Research In-stitute, 1982, pp. 159-164.

● The new data on dioxin contamination were not available to OTA. Theyare being collected by a contractor for NYS/DEC.

● *See app. C for OTA’s analysis of the EPA conclusions.I*R< ~wliW, U.S. EpA Region II, personal communication, March 1983;

and N. Nosenchuck, NYS/DEC, personal communication, May 1983.

2.

3.

drainage system should result in a decrease inoperation and maintenance costs for the State.Also once the 49 acres are contained by the bar-rier wall and surface covers, it is expected thatvery little precipitation will infiltrate into the con-taminated area. The rate of flow through thedrainage and leachate collection system is ex-pected to decrease below the current average rateof 8 gpm.The wall provides further control against migra-tion of contaminants from the canal into the res-idential areas. Should problems develop with theleachate collection system at some time in the fu-ture, the barrier wall will serve as backup pro-tection for EDA residents. Such protection, how-ever, is dependent on there being no undetecteddamage to the wall or cap over time. While leach-ate-collection system problems are being re-solved, the wall would postpone migration ofcompounds.The wall prevents migration of chemicals into thedeep aquifer below the landfill. Results of a re-cent modeling effort indicate a third advantageto having a barrier wall.20 After the wall is in-stalled, a reversal of waterflow is expected be-tween the shallow and deep aquifers, i.e., insteadof movement from shallow aquifer to deep aqui-fer, the flow will be from deep aquifer to shallowaquifer. While some reversal may be occurringdue to operation of the collection system, the ex-tent of the reversal should be greater with thewall. If the model conclusions are correct, thewall will provide the only real means for reduc-ing deep aquifer contamination.

NYS/DEC recognizes the need for continued moni-toring once the remedial action has been completed.Although not yet completed, a ground water monitor-ing strategy is being planned. Because NYS/DEC con-siders that all mobile compounds will be present in theground water, no soil or air monitoring is planned.State officials consider that any chemicals bound tothe soil will not be mobile. The synthetic cap is ex-pected to prevent volatilization, therefore air monitor-ing would not be necessary .21

The EPA monitoring study identified chemicals insediment from both storm sewers and storm sewerdischarge points in surface waters. Cleanup of thestorm sewers within the canal area has been completedand utility pipes were plugged in early 1983. Chain-link fences have been installed to discourage access to

‘OGeotrans, Inc., fioss+%wtional Simulations To Examine Proposed Wallat Love Canal, New York oral presentation given to OTA, May 12, 1983,

~ls~tments made by NYS/DEC during a meeting on May 19, 1981.

28

contaminated areas in Bergholtz and Blackfoot Creeks.Disposal of the contaminated sewage sediment awaitspermits from U.S. EPA. There are 375 drums of thismaterial being stored at the treatment facility. No ac-tion has yet been taken for contaminated storm sewerswithin the EDA.

NYS/DEC initiated a monitoring study to determinethe extent of contamination within the storm sewersystems located in the EDA. Chemicals of concern inthis study include priority pollutants and dioxin. Atotal of 1,000 samples have been analyzed. The studyresults are not yet available.

Discussions with the Love Canal Area Revitaliza-tion Agency (LCARA) indicate that decisions on thefuture use of the properties within the EDA have notbeen made.22 The Agency plans to delay any such deci-sions until the OTA review is released. An environ-mental impact assessment is required by State lawbefore any reuse of the EDA is allowed. LCARA hasbegun the assessment process .23 Some sense of urgen-cy is felt by the Agency to resolve the issue ofhabitability so that revitalization plans can bedeveloped. It should be noted that 100 residenceswithin the EDA are currently occupied. A majority ofthese (66 units in Griffon-Manor and Senior Citizenhousing) are situated adjacent to the canal area.

Uncertainties Associated Withthe Remedial Action

There are four areas of uncertainty that can affectprojections of the long-term integrity of the remedialtechnology:

1. Remedial action in the EDA.2. Leak detection systems for the barrier wall and

leachate collection system.3. Long-term monitoring programs.4. Institutional mechanisms for long-term protection

of EDA residents.While this brief OTA review cannot provide any sug-gestions for reducing the impact that these uncertain-ties may have, it is imperative that any decision ofhabitability consider them and their consequences forcontinued protection of the residences in the EDA.

Remedial Action in the EDA

Both EPA and NYS/DEC officials have based theiranalysis of the need (or lack thereof) for remedial ac-tion in the EDA, beyond that required for the storm

sewer system, solely on the results of the 1980 monitor-ing study. The major concern was whether contamina-tion observed in this area resulted from migration ofchemicals from the Love Canal landfill. Because theEPA monitoring study indicated that the only portionscontaminated by Love Canal wastes (within the EDA)were storm sewers and surface water sediments, nolarge scale remedial action is planned.

According to NY officials, the actual extent of con-tamination in the storm sewers has not been fullydeterrnined.24 A monitoring study is in progress andonce the data are available, a decision will be madeabout an appropriate method of cleanup. With the in-stallation of the barrier wall and cover, future con-tamination of the EDA from Love Canal chemicals isnot anticipated.

Unfortunately, the 1980 monitoring data were notsufficient to determine if hot spots of contaminationexist. Although data are recorded by subsection of theEDA, all values were averaged for the area as a whole.If hot spots do exist and remain untreated, the areawill continue to pose a threat to the health of theresidents.

Leak Detection Systems

Any analysis of the effectiveness of the remedial ac-tion must include some consideration of the capabili-ty to detect failure at some time after the system iscomplete. A major limitation of environmental con-trol systems, however, is that there are few methodsto test their continued integrity. Any cracks thatdevelop in the wall could serve as possible routes formigration of chemicals. If the leachate collectionsystem is working properly, such cracks should notpose a threat for outward migration of contaminants.If the system does not operate properly, however,pooling of ground water could occur near subsurfacestructures. These could consist of rock formationswithin the area as well as the basement structures con-taining rubble from destroyed houses on land im-mediately adjacent to the canal. Subsurface barrierscould impede downward movement of ground waterand facilitate lateral movement through breaks in thewall.

Officials at NYS/DEC estimate that a well-madeconcrete wall should last for at least 50 years. Evenif it lasts twice the expected lifetime, cracks can be ex-pected. The only means to detect these cracks wouldbe a decrease in the water table elevations. The syn-thetic membrane cap has an estimated lifetime of 20

ZZR. Morris, Umtjve Director, LCARA, Niagara Falls, N. Y., personalcommunication, March 1983.

~JStatments made by LCARA officials at a meeting on May 19, 1983.24J. Slack, NYS/DEC, personal communication, March 1983, restated at

a meeting on May 19, 1983.

29

years. Evaluation of water-table elevation data andchanges in volume of leachate collected in the drainagesystem are the currently available methods of deter-mining the existence of damage.

Monitoring Programs

A final area of uncertainty concerns long-term mon-itoring strategies. The monitoring effort that is plannedmay not provide sufficient warning about migrationand accumulation of chemicals outside the barrierwall. The State plan requires only a ground watermonitoring program. 25 It is presumed that all mobil-ized chemicals would eventually migrate into theshallow aquifer system within the barrier wall.

While ground water monitoring is a necessarysafeguard for containment technology, it is possibleto have contamination of soil and air before substan-tial levels of contaminants are detected in ground watersamples. For example, if cracks develop in the cap,volatile compounds would be released to the air ratherthan be transported through water. This situation ex-isted when damage to the original cap occurred, andnoxious odors were apparent around the canal area.26

Also, those chemicals that have a strong affinity fororganic material can be filtered out of contaminatedwater as it passes through soils high in organic com-ponents; this property is typical of clays found in thevicinity of the canal. Thus, any migration of contam-inated water outside of the barrier wall could lead toa build-up of such chemicals in the soil and perhapsbe taken up by vegetation. However, at present noplans exist to do any surveillance monitoring of air,soil, or biota.

Such accumulation and uptake of these types ofchemicals, often compounds that are very persistentin the environment, would not be detected throughground water monitoring. It is likely that the absenceof chemicals in the ground water samples would beinterpreted as no contamination of the area surround-ing the canal when, in fact, contamination in soil andbiota could be present. It may be prudent forNYS/DEC to develop a monitoring strategy that ob-serves biotic changes in areas adjacent to and outsidethe barrier wall as well as analyzing soil and groundwater samples.

Institutional Mechanisms for Long-TermProtection of EDA Residents

The first area of uncertainty surrounding theplanned remedial action concerns the need for long------- — ----- .—

25J. Slack, NYS/DEC, personal communication, March 1983.26Statements made by State officials during a meeting with NYS/DEC,

NYS/DOH, NYS/DOL, and LCARA on May 19, 1983.

term appropriations by the State of New York and fu-ture restrictions on the use of the canal property. Costsfor operation, maintenance, and replacement of thewall, covers, and leachate collection system are high.For example, current expenditures for operation andmaintenance of the treatment facility is approximate-ly $0.4 million.27 Included within the lifecycle costspresented in table A-2 are requirements for replace-ment of the following:

• synthetic cover every 20 years,. major equipment at the treatment facility every

20 years,● treatment plant building every 50 years,● leachate collection system every 50 years.Institutional and legal mechanisms are needed to

provide some assurance of a long-term commitmentto meet these costs. Although the current State ad-ministration may be completely committed to provid-ing sufficient funds for maintenance of the remedialaction, there are no guarantees that 10, 20, or 50 yearsfrom now the same commitment will hold. Becausethe remedial action chosen was environment controlrather than waste control, the source of contamina-tion will not be eliminated.

It should be emphasized that the current problemin Love Canal arose because the original use of thecanal was ignored or forgotten and improper use ofthe land initiated. The original deed given by HookerChemical Co. to the Niagara Falls Board of Educationincluded statements about the hazardous nature of thecontents of the canal .28 The Board chose to ignore thesewarnings and proceeded with construction of sewersystems that cut through the canal wall and a schoolthat damaged the cap.

Without strong institutional mechanisms that willguarantee continued protection for the EDA, theseoriginal problems could reoccur 50 years from now,when the current actors in this unfortunate drama haveleft the scene. At present the State has a temporaryeasement for an undetermined time, which providessome protection against improper use of the land.However, the canal property currently has three dif-ferent owners: the southern region is owned by aprivate citizen; the central section belongs to the Boardof Education; the northern portion is owned by theCity of Niagara Falls. If at any time in the future theState of New York relinquishes its temporary ease-ment, these owners will be free to utilize their proper-ty as they see fit. There are presently no strong legalor institutional mechanisms that will prevent resale andreuse of the land by the current owners.

-—— z7N, ‘Koiack, NYs/DEc, personal communication, April 1983.

“E. Zuesse, “Love Canal, the Truth Seeps Out,” Reason, February 1981,pp. 16-33.

Appendix B

Design of the EPA Monitoring Study

Summary

One aspect of the OTA review of the 1980 Environ-mental Protection Agency’s (EPA) monitoring studyfocused on the sampling design. The following specificproblems with the sampling procedures used by EPAled OTA to judge the outcome of the study indeter-minant with regard to the extent (or distribution) andlevel of chemical contamination, and its site andregional variability:

●

●

●

●

●

30

The monitoring study used uneven numbers ofsampling sites across media and 12 regions (10in the emergency declaration area (EDA), thecanal, and the control area). The numbers of sam-pling sites were not in proportion to sizes of theregions, which vary by a factor of 10. One reasonfor this situation was that EPA assumed thathigher levels of contamination existed closer tothe canal. Consequently, some regions fartheraway from the canal had very little sampling; thedistribution of sampling among regions in theEDA was particularly inadequate. Initial beliefsabout the possible routes of transport of toxicchemicals from the canal to and through the EDAmay also have influenced numbers of samplingsites in environmental media. To the extent thatthese assumptions about patterns remain un-proven or unsupported by the results of thestudy, it can be concluded that the sampling maynot have detected contamination present in theEDA which does not correspond to the patternsassumed initially by EPA.The numbers of sampling sites used were insuffi-cient to determine accurately the levels of con-tamination within some regions.As for environmental media, the extent of sam-pling was very broad and included air, surfaceand ground water, soil, sediment, and biota.However, the effort across media was uneven,and there was no examination of yearly seasonalvariations. Within the EDA, those media sampledmost extensively were soil, air, and sump water.Ground water was sampled less extensively andbiota were sampled least often of any of the en-vironmental media.Too few replicate samples were collected per siteto evaluate site variability; thus, the data on ab-solute concentrations of chemicals detected with-in any one region may not be meaningful.The study lacked adequate control data; thus

comparisons among regions are difficult. How-ever, as discussed more fully later, DHHS did notrely entirely on the control area data in itshabitability decision.

Scope of the EPA Monitoring Study

In 1980, the EPA designed and implemented an ex-tensive monitoring study of the EDA.1 The goals ofthe study were:

1. to determine the extent and level of contamina-tion in the area defined by President Carter inhis emergency declaration order (fig. B-1),

2. to assess the short- and long-term implicationsof ground water contamination in the generalvicinity of Love Canal, and

3. to assess the relative environmental quality ofthe EDA.

The design of the study was developed based on twoassumptions. First EPA expected that the levels of con-tamination would be very high. Also, EPA assumedthat the greatest contamination would exist nearest thecanal. EPA sampled five environmental media (air,soil, sediment, water, and biota). Sampling sites wereselected in 12 regions: 10 subregions of the EDA, aregion directly adjacent to the Love Canal, and a con-trol region that included selected sites throughout theNiagara Falls area. Distribution of sites within a regionwas generally random. The number of sampling sitesper region decreased with increased distance from thecanal. Additional sites were sampled at the request ofEDA residents and at places of possible migrationroutes from the canal landfill. A total of 150 chemicalswere chosen for analysis, including chemicals that wereknown to have been deposited within the landfill.

The EPA evaluated the data in two ways: one com-pared absolute concentrations of chemicals detectedwithin the EDA with available environmental stand-ards and with concentrations detected at controI sites.2

Of secondary importance to EPA was a comparisonof the frequency of detection of chemicals in the EDAto the frequency detected at control sites. EPA con-cluded that the only places within the EDA with signifi-cant contamination from Love Canal chemicals werethe sediments of storm sewer systems and their sur-face water outfalls.

‘Environmental Monitoring at Love Canal (Washington, D. C.: U.S. Envi-ronmental Protection Agency, EPA-600/4-82-030, May 1982), vol. I, II, III.

‘R. Dewling, U.S. EPA Region II, personal communication during a meetingwith OTA on May 12, 1983.

31

Figure B-1.—The 10 Subregions of the EDA Includedin the EDA Monitoring Study

10

/‘ -- Indicates approximate location of the canal landfill within the Love Canal

region.

SOURCE: Woodward-Clyde Consultants, Evaluation of Proposed Remedial Ac-tion program Love Canal, Project 1, Leachate Containment System,Niagara Falls, New York prepared for Wald, Harkrader & Ross, Wash-ington, D. C., August 1982.

Evaluation of the Sampling Effort

There are certain principles for environmentalsampling that must guide any monitoring program.3

OTA used these principles as criteria for an evalua-tion of the EPA monitoring effort. As indicated in tableB-1, the OTA analysis suggests that the number of sitesand replicate samples taken at each site were insuffi-cient to determine extent and level of contaminationfor all of the EDA.

1. Are spatial and temporal factors considered?

The EPA study attempted to investigate spatial pat-terns that could be evidence of chemical migrationfrom the landfill. Sites were chosen to represent threeregions: an area adjacent to the canal, the EDA, andcontrol area. No samples were taken directly from the

3R. H. Green, Sampling Design and Statistical Methods for Environmen-tal Biologists (New York: John Wiley & Sons, 1979).

landfill site for fear of interfering with the integrity ofthe cap and sidewalls of the canal. Because of the timeconstraints imposed on EPA, there was no attempt todetermine annual variability in the extent of con-tamination.

2. Were the choice and number of control sitesappropriate to distinguish among levels of con-tamination for the control area, EDA, andCanal regions?

There is a general consensus within the scientificcommunity that all environmental studies require base-line data to which the test area can be compared. Suchbaseline data can be control sites that are similar tothe test sites except for the variable of concern (in thisstudy the presence and concentrations of Love Canalchemicals); baseline data can also be established stand-ards for the chemicals of concern. If statistical analysesare to be conducted comparing control and test site,uncertainties in interpretation can be reduced if num-bers of control sites are equal to, or closely approx-imate, numbers of test sites.

Because of the way EPA designed the sampling ef-fort and because of the fact that the controls werelocated farthest from the canal, the number of con-trol sites was very small. For example, 11 control siteswere chosen for ground water samples. The numberof control sites for other environmental media also wassmall; a maximum number of nine sites was sampledas controls for soil analyses. For surface water andsediment, as well as drinking water, only five controlsites were identified. Sump water and storm sewers(both water and sediment) were sampled only at onesite in the control area.

Because the control sites for ground water were ad-jacent to and formed a ring around the EDA, there isconcern that some of these sites were not suitable ascontrols. For example, it is not clear that these con-trol sites were free of contamination from chemicalssimilar to those disposed in the canal landfill. Becausethe landfill had been in operation since the late 1940’s,there is the possibility that chemicals could have mi-grated within the ground water to sites designated ascontrols. For example, analysis of water table eleva-tions prior to installation of the leachate collectionsystem indicate that flow of the overburden groundwater system was away from the canal, toward thelocation of control sites.4 There is the added problemthat at least two of the control sites were located ad-

4Woodward-Clyde Consultants, Evaluation of Proposed Remedial ActionProgram LQve Canal Rvject 1, Leachate Conta inrnent System, Niagara Falk,I+&w York prqared for Wald, Harkrader & Rosa, Washington, D. C., August1982.

32

Table B-l.–Criteria Used To Evaluate EPA Sampling Effort

Criteria Design of the study

Spatial/temporal factors: ●

●

Choice/number of controls: ●

●

●

Equal number of sites: ●

●

Replicates: ●

●

●

Verification of methods: •

Only spacial factors included in the studyNo seasonal variations consideredControls for ground water adjacent to EDA; possibility ofchemical migration to control sites from CanalTwo control sites adjacent to another LandfillNumber of control samples too few for adequate analysisUnequal sample sites among regionsNot allocated in proportion to size of the regionsinconsistency in number of replicates taken per siteNot used to estimate variability within and among sitesReplicates treated as separate site samplesNo verification of sampling, handling, and analyticalmethods done prior to initiation of the study

EDA - emergency declaration area.

SOURCE: Office of Technology Assessment.

jacent to another known landfill, southwest of theLove Canal area.

3. Were equal numbers of sites allocated amongregions and environmental media?

Equal (or nearly equal) numbers of sampling sitesamong regions facilitate interpretation and reduceuncertainties in results of statistical comparisonsamong regions. If very different numbers of sites areused, indeterminance can exist for those sites sampl-ed least often. If equal numbers of sampling sites perregion are not possible, a standard practice is that thenumbers of samples be allocated in proportion to thesize of each region.

Because of the assumptions guiding the design of theEPA study, the number of sites sampled across regionswere not equal nor were they in proportion to the dif-ferences in sizes among regions, as indicated in tableB-2. For example, the entire EDA is approximately 3

1/3

times the size of the Love Canal region (see table B-3),yet the number of sampling sites do not reflect this dif-ference. It should be noted that the sampling effort didinclude sites chosen at the request of EDA residentsand because of possible migration routes from thelandfill.

Even for subregions within the EDA, the numberof sites were not allocated on a proportional basis (seefigs. B-2 through B-9.) For some media, site locationsare naturally limited. For example, sump samples wereobtained only in houses located in wet areas (fig. B-5);surface water and sediment samples were possible onlyfrom creeks (fig. B-7). However, for environmentalmedia such as air, soil, and ground water (both shal-low and deep) allocation could have been proportionalto the area of each subregion. If this had been done,the data would reflect a more accurate picture of en-

vironmental contamination over all of the EDA.It should be emphasized that by designing a sam-

pling effort with preconceived assumptions about theoutcome of the study, the use of the data can be lim-ited. In this instance, the EPA data were used to makejudgments about the overall habitability of the EDA.Because numbers of sites sampled decreased with in-creasing distance from the canal, uncertainties existabout contamination in some areas. For example,EDA-I had samples collected from only one or twosites (depending on the medium of concern) for the en-tire region. This sampling effort is hardly sufficient todetermine the overall level of contamination for thissubregion.

4. Were replicate samples taken for each site?

It is very important that any monitoring effort in-clude replicate samples (i.e., identical samples fromone site). The appropriate number of replicates willvary depending on the anticipated impact that vari-ability within individual sites may have on the con-clusions drawn from the data. If an assessment relieson absolute concentrations, replicate samples can beused to estimate the variance in concentration at a par-ticular site. Without replicates, confidence in the ab-solute concentrations cannot be indeterminate.

Replicate samples from one site are used to estimatethe amount of variance inherent at that site. Samplescollected at several different sites enable art estimationof the variability inherent in a region. These estimatesof variability may not be equal. Differences betweenthem will depend on the evenness of distribution ofa chemical within the environment, the properties ofthe medium being examined, and the presence of thosefactors that enhance or inhibit degradation of thechemical.

33

Table B-2.—Number of Sites and Samples for Target Substances

Regionsa

Environmental media 1 2 3 4 5 6 7 8 9 10 EDA LC CGround water

Shallow well:Sites . . . . . . . . . . . . . . . . . 1Samples b . . . . . . . . . . . . . 1

Deep well:Sites . . . . . . . . . . . . . . . . . 1Samples . . . . . . . . . . . . . . 0

Sump water:Sites . . . . . . . . . . . . . . . . . 2Samples . . . . . . . . . . . . . .1 -2

Drinking water:Sites . . . . . . . . . . . . . . . . . 1Samples. . . . . . . . . . . . . . 1

Surface water:Sites. . . . . . . . . . . . . . . . . 0Samples . . . . . . . . . . . . . . 0

Storm sewer water:Sites . . . . . . . . . . . . . . . . . 1Samples. . . . . . . . . . . . . . 1

Soil:Sites . . . . . . . . . . . . . . . . . 7Samples . . . . . . . . . . . . . .2 -5Volatiles. . . . . . . . . . . . . . 9

Sediment:Storm sewer:

Sites . . . . . . . . . . . . . . . . . 1Samples. . . . . . . . . . . . . . 1

Surface water:Sites. . . . . . . . . . . . . . . . . 0Samples . . . . . . . . . . . . . . 0

Air:Living:

Sites . . . . . . . . . . . . . . . . . 1Samples Tc . . . . . . . . . . . 9Samples P . . . . . . . . . . . . 3

Basement:Sites . . . . . . . . . . . . . . . . . 0Samples T..... . . . . . . . 0Samples P..... . . . . . . . 0

OutdoorSites . . . . . . . . . . . . . . . . . 0Samples T . . . . . . . . . . . . 0Samples P..... . . . . . . . 0

1311-13

9 58-9 5

32-3

32-3

3 2 42-3 1-2 1-4

4936-47

189-21

119-11

117-1o

9 54-5 2-4

52-4

21

3 2 61 1 2-4

176-13

1511-16

28-18

0 80 9-22

13-8

32-8

2 4 41-6 3-14 7-9

3392-104

136-16

11-5

31-3

1 71 3-7

31-3

42-4

3 3 22-3 2-3 2

3410-31

3l t ,3

5lt,4-5

o0

0 50 lt,2-4

o0

00

0 0 00 0 0

52-4

00

5lt,3-5

33t

3 6l,3t 1-4

11

1lt

1 1 41 1t 1,4t

11

9 10 106-9 3-10 8-1018 19 18

2313-23

45

1514-15

28

129-1224

91-918

11371-109

213

95-917

31-3

3 61-2 ,3t 1-3

00

1 1 41 1 1-2

225-15

11

41-4

11

00

0 50 1t,4

o0

00

0 0 00 0 0

53-4

00

51-5

66026

5 852 76-7920 48

560-6125-26

66627

6 6 549-50 62 46-47

35 43 22

54538-542

29230431

28

610-11

12

0 70 120 11

51211

6109

6118

000

6119

0 70 100 9

51212

69

7-8

61010

000

tSampled only for K-stable potassium, cesium, radium, Americium.aColumns 1 through 10 represent subregions in the emergen Cy declaration ares (EDA); LC represents the region adjacent to the Canal, and C represents the control region.bRepresents range of total analytical samples verified and entered into the EPA data base for each target substance; e.g., in EDA-2 some target substances

were analyzed using 11 shallow well samples and other substances using 13 samples.CT represents the total number of samples analyzed with the Tenax method; P represents the total number analyzed with the PFOAM method.

SOURCE: U.S. Environmental Protection Agency, Environmental Monitoring at Love Canal, Volume III.

As illustrated in table B-2, multiple samples per site that for some compounds as few as 8 samples werewere collected only for sump water samples, for vola- analyzed and recorded in the data base; for othertile compound analyses in soil, and for air analyses.5 chemicals, analyses were performed with 18 samplesFor example in EDA-2, two sump water sites were collected at the two sites. For all soil sites, two samplesidentified from which 8 to 18 samples were collected per site were collected for analysis of volatile com-for chemical analyses. The range (8 to 18) indicates pounds. Two methods were used to analyze air sam-

ples. For the Tenax method nearly 10 replicates persite were collected; four replicates per site were col-

‘Environmental Monitoring at Love Canal, op. cit., vol. 11. lected for analysis using the PFOAM method. It is dif-

34

Figure B-2.-Distribution of Shallow Well Figure B-3.–Distribution of Deep WellSampling SitesSampling Sites

c

c

c

cc

—

c

.1.

ccc

KEY: Numbers denote subregions; X indicates approximate location of eachKEY: Numbers denote subregions; X indicates approximate location of each

well; C represents location of control wells; ( ) indicates maximumWell; C represents location of control wells; ( ) indicates maximum

number of samples collected per region.number of samples collected per region.

SOURCE: Environmental Monitoring at Love Canal (Washington, D. C.: U.S. Envi- SOURCE: Environmental Monitoring at Love Canal (Washington, D. C.: U.S. Envi-ronmental Protection Agency, EPA-600/4-82-030, May 1982), vol. Ill, p. 14. ronmental Protection Agency, EPA-600/4-82-030, May 1982, vol. Ill, p. 19.

Table B-3.—Approximate Size of EDA Study Regions replicates, EPA did not follow the normal practice ofusing them to estimate variability within the site. Suchestimates are particularly important when low levelsof contamination are encountered and when an assess-ment of habitability is based on absolute concentra-tions of chemicals, as was the case in this study. Thedecision to treat replicates as actual samples was un-fortunate as it increases uncertainties about the sig-nificance of differences in observed concentrationsamong control, EDA, and Love Canal regions.

Square yards Football fielda

Region (in thousands) equivalence

3 . . . . . . . . . . . . . . . . . . . . .4 . . . . . . . . . . . . . . . . . . . . .5 . . . . . . . . . . . . . . . . . . . . .

5. Were the sampling and analytical techniquesverified?

Love Canal region . . . . . . 60

All sampling and analytical methods have certainbiases associated with them—e.g., differences in resultscan occur if slightly different procedures are followed,if different personnel perform the analyses, and if dif-ferent collection and analytical equipment are used.If results of environmental studies are to be properlyinterpreted, it is necessary to identify these biases.

aCalculated by assuming a football field is approximataly equal to 5,000 yd2.

SOURCE: Calculated from maps provided by the Love Canal Area RevitalizationAgency.

ficult to determine whether these multiple sampleswere actual replicates.

If, in fact, these multiple samples were collected as

35

Figure B-4.—Distribution of Soil Sampling Sites

KEY: Numbers denote subregions, X Indicates approximate location of eachsite; ( ) Indicates maximum number of samples (nonvolatile) collectedper region

SOURCE: Environmental Monitoring at Love Canal (Washington, D. C.: U.S. Envi-ronmental Protection Agency, EPA-600/4-82-030, May 1982), vol. Ill, p. 23.

Because of time constraints, EPA could not conductany preliminary analyses that would have identifiedbiases inherent in the techniques chosen for thismonitoring study.

In addition, EPA did not verify that the methodsused for sample collection and analysis were suitablefor the conditions at Love Canal. For example, themethods of obtaining samples could have been a ma-jor contributor of the large number of below detec-tion results that were obtained. Soil samples were ob-tained by using a soil corer, which obtained a core 1 3/8

inches in diameter and 6 feet in depth. Seven coreswere taken at a site; two cores, representing twosamples per site, were analyzed for volatile chemicals.The remaining five cores were homogenized andtreated as one sample per site. Such a method couldhave serious consequences for detecting soil contami-nation. If the compounds are present at low concen-trations and/or within only a small region of the core,the practice of compositing five cores to produce onesample will dilute any concentration level and makedetection extremely difficult. Because of this dilution

Figure B-5.—Distribution of Sump WaterSampling Sites

KEY: Numbers denote subregions; X indicates approximate location of sites;( ) indicates maximum number of samples collected per region.

SOURCE: Environmental Monitoring at Love Canal (Washington, D. C.: U.S. Envi-ronmental Protection Agency, EPA-600/4-82-030, May 1982), vol. Ill, p. 33.

factor associated with this particular sampling tech-nique, even if hot spots exist at a site the measuredvalues would be lower than actual environmental con-centrations. The use of such sampling techniques callsinto question the validity of using absolute contamina-tion values as the basis for a habitability decision.

Similar problems existed for ground water samples.For example certain ground water samples were invali-dated by EPA, because it was suspected that inade-quate purging had occurred.’ Hydrant water was usedas a drilling fluid during construction of the wells.Prior to collecting samples the wells had to be purged,removing all hydrant water. EPA officials thought thathydrant water had been collected rather than aquiferwater. Appropriate location of sampling wells is crit-ical to obtaining representative samples with which tojudge the extent of ground water contamination.Plumes of chemicals, which may have densities greaterthan water, can travel in directions different from the

6Environmental Monitoring at Love Canal, op. cit., vol. I, p. 238.

Figure B-6.— Distribution of Storm SewerSampling Sites

m

1

m

10(2)

(

KEY: Numbers denote subregions; X indicates approximate Iocation of eachstorm sewer; ( ) indicates maximum number of samples (water or sedi-ment) collected per region.

SOURCE: Environmental Monitoring at Love Canal (Washington, D. C.: U.S. Envi-ronmental Protection Agency, EPA-600/4-82-030, May 1982), vol. Ill, p. 41.

ground water flow and can migrate into undetectedfissures.’ Thus, when this occurs analysis of samplestaken from those wells placed to match ground waterflow patterns would not likely lead to detection of thecontaminated plume.8

Analytical methods likewise were not verified forLove Canal environmental conditions and study mech-anisms prior to initiation of the monitoring effort. IfEPA had attempted such verifications, problems asso-ciated with sample extraction (e.g., for dioxin),analyses of air samples using the PFOAM method, anduneven analytical capabilities among laboratoriescould have been resolved before enormous effort and

7The geology beneath the canal landfill has not been studied extensively.Fractures have been noted in parts of the Niagara Falls region. E. Koszalka,U.S. Geological Survey, Long Island, N.Y., personal communication, March1983.

‘Ground water monitoring problems are discussed further in Tiwhnologiesand kagement Strategies fix Hhzar&us Waste Controf, “Chapter T: Thecurrent Federal-State Hazardous Waste Program” (Washington, D. C.: U.S.Congress, Office of Technology Assessment, OTA-M-l%, March 1983).

Figure B-7.—Surface Water Sampling Sites Along theBlack Creek and the Bergholtz Creek

.

I

m

-

10

/

KEY: Numbers denote subregions; ( ) indicates maximum number of samplescollected for region 4.

SOURCE: Environmental Monitoring at Love Canal (Washington, D. C.: U.S. Envi-ronmental Protection Agency, EPA0600/4-82-030, May 1982, vol. III, p. 51.

resources had been expended. Absence of the verifi-cation was probably a direct result of the fact that EPAwas under great pressure to do the study quickly. Inretrospect, providing time prior to initiation of themonitoring study to verify methods likely would haveresulted in more definitive answers.

Proper sample handling is a critical element in en-vironmental monitoring programs. All samples takenin the field tend to lose a variable portion of thesubstances to be monitored during sampling, handling,and storage. A standard technique for determining thepercent of loss is to add a specified amount of knownsubstance (to “spike”) to certain field samples. The con-centration of the spiked substance is measured in thelaboratory; any differences between the amount addedin the field and the amount measured in the laboratoryrepresents the percent lost during sample handling.Analytical results of unspiked samples thus can be ad-justed to reflect these losses. Loss of concentrationsduring sample handling can result from chemicalsbonding to the sample medium or containers, to vola-

37

Figure B-8.-Distribution of Air Sampling Sites

—

—

—

b .

KEY: Numbers denote subregions; X indicates approximate location of eachsite; ( ) indicates maximum number of samples collected per region.

SOURCE: Environmental Monitoring at Love Canal Washington, D. C.: U.S. Envi-ronmental Protection Agency, EPA-600/4-82-030, May 1962), vol. Ill, p. 61.

tilization, or to other chemical-physical processes thatmay occur during handling and storage.

This technique was not employed in the EPA studyfor water, soil, or sediment samples. Blind spikedsamples were included in the field sample analyses forair. Consequently, reliable estimates of loss for mostsubstances cannot be made. (However, EPA did esti-mate percent recovery for extraction of dioxin fromfield samples.) Particularly for volatile chemicals andfor those samples collected during the warmest peri-ods of the monitoring program (which spanned Augustto October), loss of substances could have occurred.Analysis of air samples, however, did not reveal pro-nounced seasonal variations. It is uncertain whetherother media would reveal a similar lack of variation.

Field spiking was omitted from the EPA protocolsto eliminate the possibility of accidental contamina-tion of all field samples with target substances.9 In ad-dition EPA was presented with certain difficulties re-garding implementation of spiking for their fieldsamples. Spiking with all 150 target substances would

Figure B-9.—Distribution of Drinking WaterSampling Sites

xx

KEY: Numbers denote subregions; X Indicates approximate location of eachsite; ( ) indicates maximum number of samples collected per region.

SOURCE: Environmental Monitoring at Love Canal (Washington, D. C.: U.S. Envi-ronmental Protection Agency, EPA-600/4-82-030, May 1962), vol. Ill, p. 74.

be extremely difficult; choosing a few compounds toserve as representatives of the total set has severaluncertainties associated with it. For example, it wouldbe difficult to verify that the representative compoundsbehaved similarly to the 150 chemicals within the vari-ety of environmental media investigated in this study.Identification of these representative compoundswould have required additional time, and EPA wasunder pressure to complete its large study within 6months.

Conclusions About theSampling Strategy

Perhaps the most serious failings of this study were:

9J. Deegan, statement and supplementary testimony before the subcom-mittee on Commerce, Transportation, and Tourism of the Committee onEnergy and Commerce, U.S. House of Representatives, 97th Cong., serialNo. 97-197, Aug. 9, 1982.

38

the inadequate numbers of sites sampled in dif-ferent regions,varying intensity of sampling of different envi-ronmental media, andthe lack of replicate samples with which to esti-

ty, confidence in reported concentration values mustbe limited. For example, “trace” is considered to repre-sent levels less than 100 parts per billion (ppb) in theEPA study; the possibility exists that site variabilitycould range by an order of magnitude. Analysis of rep-

1 .

2.

3.licate sample; from a site would verify whether tracemate site variability in concentrations of chem-

icals.The numbers of collected samples per region and thenumber of measurements available per target sub-stance are insufficient to serve as a representative pic-ture of the potential contamination either in the EDAor within subregions of the EDA. For example, EDA-1covers an area of approximately 80,000 square yards(approximately equal to 16 football fields, as shownin table B-3). Only one or two sites, depending on theenvironmental media, were sampled to represent po-tential contamination. Similarly, nine or fewer siteswere sampled for all media to represent EDA-6, anarea approximately equal to 50,000 square yards anddirectly adjacent to the Love Canal region.

The lack of sufficient sites and inability to estimatesample variations within sites presents serious conse-quences for an assessment of habitability based on ab-solute concentrations. Without estimates of variabili-

represents 100 ± 10 ppb or 100 ±200 ppb. Becausehuman health effects can result from chronic exposureto concentrations in the ppb and parts per million(ppm) ranges, it is necessary to know with some levelof confidence that the values reported for an area rep-resent actual environmental concentrations.

Dioxin is probably the most toxic of the contami-nants (in low concentrations) known to be present inLove Canal, and an analysis of the data for this chem-ical illustrates the OTA concern about the samplingeffort. Monitoring for dioxin was insufficient with re-spect to extent, level and replication, as shown in tableB-4 and figure B-10. Only 6 out of 21 submedia weresampled in the EDA; 3 were sampled in control areas;and 7 submedia were analyzed in the Love Canal. Ofthe 10 regions in the EDA, only 2 were sampled forsump water contamination and 3 each for air and soil.No samples were collected in deep well, shallow well,

Table B-4.—Sampling Effort for Dioxin

EDA (10 subregions) Controls Love Canal regionTotal Total Total Total Total Total Total

regions sites samples sites samples sites samples

Water:Deep well . . . . . . . . . . . . . . . . .Shallow well . . . . . . . . . . . . . . .Sump . . . . . . . . . . . . . . . . . . . . .Surface . . . . . . . . . . . . . . . . . . .Drinking . . . . . . . . . . . . . . . . . . .Storm sewer . . . . . . . . . . . . . . .Sanitary sewer . . . . . . . . . . . . .

Soil . . . . . . . . . . . . . . . . . . . . . . . . .Sediment:

Sump . . . . . . . . . . . . . . . . . . . . .Storm sewer . . . . . . . . . . . . . . .Sanitary sewer . . . . . . . . . . . . .Surface water . . . . . . . . . . . . . .

Air:Living area. . . . . . . . . . . . . . . . .Basement . . . . . . . . . . . . . . . . .Outdoor . . . . . . . . . . . . . . . . . . .

Biota:Oatmeal . . . . . . . . . . . . . . . . . . .Potatoes . . . . . . . . . . . . . . . . . .Crayfish . . . . . . . . . . . . . . . . . . .Dog . . . . . . . . . . . . . . . . . . . . . . .Maple . . . . . . . . . . . . . . . . . . . . .Mice . . . . . . . . . . . . . . . . . . . . . .Worms . . . . . . . . . . . . . . . . . . . .

00430003

00030004

33500403

22200104

0400

0400

0901

01803

018

03

0000

0000

200

300

100

100

100

300

100

0000000

0000000

0000000

0000000

0000000

0000000

0000000— — . — —

Totals . . . . . . . . . . . . . . . . . . . 33 36 3 3 23 16SOURCE: U.S. Environmental Protection Agency, op. cit., Vol. Il.

39

Figure B-10.– Distribution of Dioxin Sampiing Sites

0●o0

A i r

.

Key:

● Stream sediment

Ground water

Storm sewer sedimentSOURCE: Environmental Monitoring at Love Canal (Washington, D. C.: U.S. Envi-

ronmental Protection Agency, EPA-600/4-82-030, May 1982), vol. Ill, p. 77.

drinking water, storm sewer water, or sanitary sewerwater. Surface water and sediment were sampled inone subregion. It should be noted, however, that thevolubility of dioxin in water is extremely low (0.2 ppb).Storm sewer sediment was collected in nine of the EDAregions. Also, this substance has not been detectedpreviously in air samples, except when present on dust,near incinerators, or in smoke from forest fires. Thus,EPA may have reduced the extent of sampling becauseof assumed distribution primarily in soil and sediment.

For a determination of the possible level of dioxincontamination in the EDA, the amount of samplingwas very limited for dioxin. For most of the regions,including the 10 subregions of the EDA, no more thanfive sites were sampled per region. Storm sewer sedi-ment sampled in the EDA is the one exception. Giventhe large area covered by both the Love Canal andEDA (see table B-3), this level of sampling effort seemsinsufficient to estimate the potential contamination inthe Love Canal area.

Within each of the major sampling regions (EDA,control area, and Love Canal), few replicate samples

were taken. Variability in concentrations within asingle site could be possible, but without replicates thisvariability cannot be estimated. Lack of replicates re-duces the certainty associated with comparison ofdioxin concentrations among the EDA, Love Canal,and control area. Distinctions between regions requireestimates of variation within each individual regionas a basis for comparison. Consequently, in the ex-treme case of no replication, it is impossible to deter-mine whether the regions, regardless of absolute con-centrations, differ because of normal site variabilityor because of actual regional variations.

OTA compared this extent and level of sampling fordioxin at Love Canal with recent EPA protocols fordioxin sampling in Missouri.10 The Eastern MissouriDioxin program has collected samples from some 30areas, including Denny Farm (1979), Times Beach(1982), and Quail Run (1983). Two important dif-ferences emerged.

1. In Eastern Missouri, preliminary surveys wereconducted to identify areas of highestcontamination,

2. Within the boundaries of these highly contami-nated areas, the level of sampling was between4 and 37 times as great as in the Love Canal EDA(table B-s).

Table B-5.—Comparison of Dioxin Sampling EffortBetween Eastern Missouri and the EDA

Area Total Samples(acres) samples per acre

Denny Farm (1979). . . . . . . . . . 4.5 30 6.7Love Canal EDA (1980) . . . . . . 200 0.18Times Beach (1982) . . . . . . . . . 640 5(X%OO 0.78-0.93Quail Run (1983) . . . . . . . . . . . 24 113 4.7SOURCE: Environmental Promotion Agency, Region Vll, Kansas City, Mo., May

Some important differences exist between the East-ern Missouri program and the Love Canal effort. First,the analytical capability to analyze for dioxin has ad-vanced tremendously since 1980. Second, the detec-tion limits are different for the two programs: 20 partsper trillion for Love Canal and 1 ppb for EasternMissouri. This leads to significant gains in turnovertime and laboratory capacities. Third, at Love Canal,EPA was faced with analyzing for a broad diversityof substances known to have been disposed in the land-fill, while in the Eastern Missouri program, dioxin wasthe only target substance. Fourth, in Eastern Missouri,EPA is not operating under a presidentially declaredFederal Emergy Management Agency state of emergen-cy as was the situation at Love Canal. Therefore, timewas not a limiting factor.

IOGale Wnsht, William Keffer, Will Bun, William Fairleaa, and JohnWhitland, U.S. EPA Region VII, Kansas City, Me., personal communica-tion, May 27, 1982.

Appendix C

Results of the EPA Study Relatedto the Habitability Decisions

Summary

Until there is agreement about the possible level ofchemicals in samples that contained no detectable con-centrations, it is pointless to dwell on the quantitativeaspects of health risk posed by chemicals from LoveCanal. If the concentrations are in the parts-per-billion(ppb) range, the risk has to be judged to be very lowand probably acceptable. If the concentrations forsome chemicals are 1,000 times higher, in the parts-per-million (ppm) range, the risks are probably not ac-ceptable. According to the Environmental ProtectionAgency (EPA), the concentrations are in the ppb range;according to the National Bureau of Standards (NBS),they could near 1 ppm.

OTA does not agree that the ppm estimate is realisticfor all chemicals, and it tends to accept EPA’s esti-mates, but OTA does agree with NBS that furtherdocumentation from EPA is necessary to settle thematter. A resolution between EPA and NBS might bereached by an examination of a subset of EPA’s rec-ords. Also, if additional monitoring is carried outbefore or during rehabitation of the emergency decla-ration area (EDA), EPA should consult with NBS toensure that quality control measures are adequate.

Basis of the Habitability Decision

The major input to the habitability decision wasdata generated by the EPA monitoring study.1 TheDepartment of Health and Human Services (DHHS)used absolute concentrations of chemicals found inEDA and assessed the relationship of these concentra-tions to potential health problems. DHHS also re-viewed data about health problems observed in EDAand Love Canal residents and used professional judg-ments about possible human health effects resultingfrom exposure to chemicals deposited in the canal land-fill.

The OTA review concentrated on the EPA monitor-ing data and possible health effects associated with

Love Canal chemicals. OTA inspected but did notevaluate the validity of reported health problems ofresidents nor question the professional judgments ofthe DHHS officials. The results of the OTA analysisindicate three areas where uncertainties in the datacould have a major impact on the DHHS decision.These areas include:

1. the range of variability associated with values re-ported for chemicals detected, nondetected, andtrace;

2. uncertainties in potential health effects associatedwith Love Canal chemicals; and

3. problems associated with comparing data for theEDA with data in control areas.

Problems With Statistical Comparisonsof EPA Results

A major statistical problem is related to the smallnumbers of controls used in the EPA analysis.2 Thepower to detect differences in contamination betweenthe EDA and control areas and between Love Canaland control areas has been questioned.3 These criti-cisms that the canal cannot be distinguished from thecontrol are accepted as valid by EPA. This createsuncertainty for a conclusion that the EDA is as hab-itable as the control areas to which it was compared.Silbergeld has attacked the EPA monitoring study oninsufficiency of statistical power:

The small number of control area sampling sites se-riously reduced the ability to detect differences inchemical contamination between the Declaration Areaand the control area.The absence of power to distinguish between the

canal and the control areas seriously compromises anyconclusions to be drawn from comparing Love Canalto the EDA and EDA to the control areas because, inmost cases, statistically there are no differences be-tween Love Canal and the control areas. The absence

*D. Rail, National Institute of Environmental Health Science, ResearchTriangle Park, N. C., and B. Paigen, Children’s Hospital, Oakland, Calif.,personal communications, May 1983. See Fmvironmental A40nitoring at LoveCanal: Interagency Review, comments by DHHS, NBS, and EPA (Wash-ington, D. C.: U.S. Environmental Protection Agency, Office of Research andDevelopment, May 1982).

‘E. Silbergeld, Environmental Defense Fund (EDF), testimony before theSubcommittee on Commerce, Transportation, and Tourism, Committee onEnergy and Commerce, U.S. House of Representatives, 97th Cong., serialNo. 97-197, August 1982, pp. 68-103.

3R. J. Cook, testimony before the Joint Public Hearing on Future Uses ofthe Love Canal Hazardous Waste Site and Adjacent Property, State of NewYork, Assembly Standing Committee on Environmental Conservation, As-sembly Subcommittee on Toxic and Hazardous Substances, Feb. 17, 1983.

40

41

of statistical power to distinguish between the LoveCanal and control areas results from the small numberof control area samples, and nothing can be done atthis time to make up for that deficiency. It is impor-tant to remember that differences almost certainly ex-ist in chemicals actually present in the Love Canal andcontrol areas. But the differences cannot be shown be-cause of too few control area samples.

Independent analysis of the EPA data shows thatthe greatest number of samples analyzed from any onemedium in the control area was 33, compared to 539in the EDA.4 For every chemical tested there werefewer than 10 samples in a majority of the media.Table C-1 describes the number of samples needed tohave a good chance of detecting differences betweentwo areas. Formally, this table gives required samplesize for a one-sided, alpha 0.10, Z-test on two propor-tions to achieve a power of 0.90. In less formal lan-guage, if the real frequency of positive detections ofchemicals in the control sample is equal to 5 percent(0.05) and the positive detection rate in the EDA isequal to 20 percent, a minimum of 61 samples fromeach region must be analyzed to have a 90-percentchance of detecting this (fourfold) difference. Becausethe maximum number of samples analyzed in the con-trol region was 33 from any one medium and most ofthe time it was only 10 samples, even a fourfold dif-ference in chemical detection rates would not be rec-ognized. Because differences in detection rates betweenthe EPA and control area are much smaller than thesevalues, statistical significance between the EDA andcontrols could not be expected.

OTA concludes that any decision based on differ-ences in detection frequencies between the Love Canal,EDA, and control area must be discounted because ofthe weak statistical basis of the EPA study. EPA ap-parently agrees with this assessment and asserted thatmaking such comparisons is not the normal way tojudge whether an area is contaminated.5 Rather, EPAwould rely on measured absences of chemicals to showthe area is not contaminated. Some type of baselinedata, however, are needed to make such judgments.These baseline measurements could be either controlarea analyses or established environmental standards.Unfortunately, few of the chemicals disposed in thecanal landfill have established environmental stand-ards.

Range of Variability for Reported Values

Except for some compounds detected in sumps andstorm sewer systems, concentrations of chemicalsreported for the Love Canal region were generallyquite low, as illustrated in tables C-2 and C-3. Themaximum values reported for organic chemicals de-tected in the EDA, Love Canal, and control regionsrange from 0.05 to 263 ppm. In the Love Canal, veryhigh concentrations were reported for sump sediment(16,500 ppm); however, these samples were taken fromsumps of homes that had been built directly adjacentto the canal landfill. Table C-3 provides reported max-imum values for those media where dioxin was de-tected. These values were 672 ppb found in stormsewer sediment within the EDA and 37 ppb detectedin surface water sediment, also in the EDA. For allother environmental media, results of dioxin analyseswere below EPA’s reported detection limits (20 ppt).In the Love Canal, very high values of dioxin werereported for sump and storm sewer sediment. No val-ues were reported for the control areas.

Table C-1 .—Number of Samples Required To DetectActual Differences Between the EDA

and Control Areas

Number of samples/per medium toDetection rates produce a W-percent chance of

Control EDA detecting a statistical difference0.03a 0.06b 6250.03 0.09 2030.03 0.12 1100.05 0.10 3620.05 0.15 1150.05 0.20 61

a The frequencies of detection in control area samples was between 3 and 5 percent

bAssumed detection rate in the EDA.

SOURCE: Cupples, op. cit.

Although these measures appear low, their absolutevalues for organic compounds can be questioned. TheEPA monitoring study used 19 different analyticallaboratories, each with varying capabilities.’ NBS wasasked by EPA to review the quality control protocolsused in the study.7 While NBS accepted the protocolsas adequate, the Bureau could not verify the certain-ty associated with performance of the different lab-oratories. As stated in a letter to Senator A.M.D’Amato: 8

4L. A. Cupples, Boston University School of Public Health, report sub-mitted to OTA, Industry, Technology, and Employment Program, May 1983.

‘Statements made by EPA officials during a meeting with OTA on May12, 1983.

6Environmental Monitoring at Love Canal (Washington, D. C.: Environ-mental Protection Agency, vol. I, pp. 36-37.

7Environmental Monitoring at Love Canal: interagency Review, op. cit.‘R. G. Kammer, letter to Senator A. M. D’Amato, August 1982.

42

Table C-2.-Maximum Vaiues (in ppm) Reported for Organic Compounds

Shallow well. . . . . . . . . . . . . . . . . . . . . . . . . . EDAControlLove Canal

Deep well . . . . . . . . . . . . . . . . . . . . . . . . . . . . EDAControlLove Canal

Soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EDAControlLove Canal

Sump water . . . . . . . . . . . . . . . . . . . . . . . . . . EDAControlLove Canal

Sump sediment . . . . . . . . . . . . . . . . . . . . . . . EDAControlLove Canal

Storm sewer water . . . . . . . . . . . . . . . . . . . . EDAControlLove Canal

Storm sewer sediment . . . . . . . . . . . . . . . . . EDAControlLove Canal

Surface water sediment . . . . . . . . . . . . . . . . EDAControlLove Canal

0.048 di-n-octylphthalate0.150 di-n-octylphthalate3.300 3-chlorotoluene0.230 phenol0.105 xylene0.050 acrolein3.120 chrysene0.420 benzene10.485 1,2,3,4-tetrachlorobenzene0.586 1,4dichlorobenzene0.002 Aroclor 12548.500 2,4-dichlorophenol

——

16,523 2,4dichlorotoluene0.062 1,2,4,5-tetrachlorobenzene0.0001 gammma-BHC0.120 hexachlorobutadiene123.000 di-(2-ethylhexyl)phthalate0.012 1,2-dichloroethane263.000 Aroclor 125420.000 delta-BHC23.645 di-(2-ethylhexyl)phthalate

—SOURCE: U.S. Environmental Protection Agency, op. cit., vol. Ill

Tabie C-3.–Maximum Vaiues Reported for Dioxin, ppb

Love ControlMedia EDA Canal area

Sump water. . . . . . . . . . . . . . . . . . . . — .6 –Sump sediment . . . . . . . . . . . . . . . . — 9570 –Storm sewer sediment . . . . . . . . . . 672 329 –Surface water sediment . . . . . . . . . 37.4 — —SOURCE: U.S. Environmental Protection Agency, op. cit., vol. Ill.

As we reported in our May 10, 1982 review, unlessmeasured values, including non-detected, are accom-panied by estimates of uncertainty, they are incompleteand of limited usefulness for further interpretation andfor drawing conclusions.

EPA’s response to NBS was to provide a “worst case”range for selected chemicals based on performance ofthe worst laboratory.9 While this provides some ideaabout the variability for these particular chemicals, itdoes not allow estimation of confidence limits for thetotal set of 150 chemicals used in the study.

As mentioned in appendix B, a major failing of theEPA effort was the improper use of replicates. For mostof the environmental media, no replicate samples weretaken at individual sites. In those few instances wherereplication was obtained, EPA treated them as separatesamples. Thus, there is no way to determine if the ab-solute values reported for any one chemical varies bytwofold, tenfold, or 100-fold.

‘Environmental Monitoring at Love Canal: Interagency Review, op. cit.

This concern is not trivial. All samples collectedfrom environmental media will vary to some extent.Because of the inherent variability of ecosystems andthe variations in interactions between chemicals andelements of the environment, a minimal level of uncer-tainty can never be overcome. An additional level ofvariability results during the analytical phase of amonitoring study. Such analytical variations arisewhen different people perform the same procedure inthe same laboratory. Even greater variability is in-troduced when different laboratories with differentcapabilities, experience, and equipment* are used inthe same study.

In addition to uncertainties associated with the ab-solute values reported for chemicals detected withinthe EDA, there is uncertainty associated with the detec-tion limits of the various laboratories. Within theEDA, 90 percent of the analytical measurements werebelow the laboratory detection limits. The obviousquestion is raised. Are the low values real or could theyresult from limitations of the various laboratories? Ifdetection limits were insensitive (i.e., too high), con-centrations significant for potential health problemsmay be overlooked. If detection limits were too var-iable, then extent of contamination may not be ac-curately identified. The adequacy of EPA’s reporteddetection limits for analytical methods is difficult to

● Although the same brand name of equipment can be used by differentlaboratory=, performance differences between similar equipment can beexpected.

— ———-—-

43

evaluate, in part because of the conceptual complexi-ty of detection limits.

Method detection limits (MDL) were used by EPA.The MDL is the lowest concentration of a substancethat can be detected with a 99 percent confidence thatthe reported concentration is greater than zero.10 EPAregarded detections reported as “trace” to be above theMDL for a particular analytical laboratory .11 There-fore, the upper limit for the frequency of reporting falsepositive (or false negative) readings should be 1 per-cent, i.e., one might not be able to determine with 99percent confidence that an undetected substance is, in-deed, absent in a sample at concentrations below theMDL. OTA finds that this is subject to uncertainty.

EPA reported detection limits for only a sample ofits target substances, on the grounds that this selectgroup was representative of each structural class pres-ent among the target substances. However, it is dif-ficult to determine whether MDLs accurately reflectroutine practices of particular analytical laboratories.As EPA acknowledged, except in the case of air sam-ples, analytical laboratories knew which samples wereperformance evaluation samples.l2 Performance eval-uation samples could, therefore, have been analyzedmore carefully than on the field samples.

Moreover, a different MDL was reported for par-ticular substances representing more than one ana-lytical method, and up to six analytical laboratories.In some cases, MDLs were estimated for several lab-oratories based on the performance of only one. Thesefactors contribute to uncertainty about detectionlimits.

Reported MDLs for metals were estimates only, andwere reported as aggregated value, which obscuresvariability among laboratories. Also, analysis forpesticides have only a single reported MDL, againobscuring laboratory variability. In contrast, a MDLfor many organic substances was reported as a meas-ured limit for specific laboratories.

Although most reported detection limits are quitelow relative to health standards (where available), theissue of uncertainty of detection limits is relevant tothe issue of the validity of EPA results. For example,beta-BHC is reported to have an overall (low) MDLof 0.006 ppb for analytical Method 608 in reagentwater. The same substance is reported to have MDLsthat range from 4.2 to 9.5 ppb using analytical Method625. The actual value depends on which laboratoryperformed the analysis. Thus, if different methods or

different laboratories were used to analyze for beta-BHC, then intended comparisons of frequencies ofbeta-BHC detections among components of the envi-ronment might represent comparisons of methods andof laboratory performance. Although, in conversation,EPA has asserted that this should not happen, nomechanism for reliably preventing it was presented.It should be emphasized that neither of these valuesmay be relevant to detection limits of actual samplesas the Love Canal samples which would contain com-peting contaminants that possibly lower analyticalpower.

Such variability in MDLs is not atypical. Detectionlimits for other compounds also varied widely. For ex-ample, reported MDLs for 1,2,3,4-tetrachlorobenzeneranged from 0.5 to 17 ppb, varying by a factor of 34for different laboratories employing the same ana-lytical method. Detection limits for several closelyrelated substances, the alpha, beta, delta, and gam-ma isomers of BHC, varied from 0.004 to 0.009 ppb,a factor of 2.25. Likewise, detection limits for DDT,DDD, and DDE varied by a factor of 3 (0.004 to 0.12ppb). MDLs for endosulfan 1, endosulfan 2, and en-dosulfan sulfate varied from 0.004 to 0.066 ppb (a fac-tor of 16.5), and for heptachlor and heptachlor epox-ide MDLs varied by a factor of 27.7 (from 0.003 to0.083 ppb).

MDLs were reported for only a subset (about one-third) of the total 150 chemicals; EPA considered thatthe subset of compounds spanned the range of com-pound classes used in the study .13 Consequently, EPAasserts that it should be possible to determine approx-imate detection limits for all substances:14

Similarly, it is reasonable to assume that the methoddetection limits of most of the organic analytes . . . fallinto the same range of 0.5 to 79 micrograms per liter.

Nevertheless, none are provided for substances knownto have been disposed into Love Canal.

This variability in detection limits introduces uncer-tainties in interpreting the meaning of the many sam-ples reported to be below the limits of detection. Thisuncertainty in turn casts doubt on any conclusionsabout the levels of contamination of the EDA or thecontrol area.

All reported MDLs are in or below the range of 0.5to 79 ppb. The observed variability of MDLs acrossmethods as well as for similar compounds and the lackof MDLs for most of the target chemicals calls intoquestion the ability to detect hazardous concentrations

IOJ. A. Glaser, et al., ‘Trace Analyses of Wastewaters,” Environmental Sci-ence and Technology, vol. 15, 1981, pp. 1426-1435.

ll~vuonmenta] Monitoring at Love Canal, op. cit.l~statmmts made by EPA officials at a meeting with OTA, May 12, 198s.

‘3 Deegan, op. cit., “it is true that MDL’s were determined for a subset ofthe target compounds, and the subset included model compounds for the com-plete set of target compounds . . a valid methodology . . . accepted wide-ly in scientific research,” p. 147.

ldfiv~onmental h40nitor@ at Love Canal, op. cit., VO1. I, P. 228.

44

of the target substances. Were the detection limits foreach compound sufficiently low and were actual lab-oratory performances sufficiently high to allow a con-clusion that those chemicals not detected would bepresent in such low levels as not to pose a threat tohuman health? Is the range of variability for MDLssufficiently low to be certain that estimates of variancefor absolute concentrations are not within hazardousconcentrations for all target substances? Until MDLvalues are reported with estimates of variances foreach, uncertainties about the meaning of none detectedand trace, remain.

What If EPA’s Numbers Are Wrong?

Most of the samples in which EPA detected meas-urable amounts of chemicals revealed concentrationsin the ppb range. If those numbers are accurate, theassumption can be made that the samples in whichonly traces of chemicals were detected or in which noteven traces were detected contain even lower concen-trations of chemicals. Looking at the data reportedlater in tables C-5 and C-6, it can be seen that if thetrace measurements are in the ppb range, the levelsof chemicals in the EDA are indeed so low as to posean acceptable health risk (except for hexachloroben-zene and dioxin).

NBS was asked to comment on the amount of chem-icals that might have gone undetected in EPA’s mon-itoring program at Love Canal. NBS was not con-vinced that the absence of detectable levels of chem-icals in the EPA analysis was consistent with concen-trations as low as parts per billion.15 Instead, it is con-fident that the concentration of a chemical reportedto be below detection is no more than 1 part per mil-lion.

OTA asked officials of DHHS who participated inmaking the habitability decision if they would persistin their conclusion that the declaration area was hab-itable if many chemicals were present in the near 1 ppmrange. The response was that they would stick by theirearlier decision with a demur about certain chemicals.

OTA would not be so sanguine about the safety ofthe EDA if the concentration of all or most of the 150chemicals approached 1 ppm. If the NBS estimate thatthe “no detectable limit” might be as high as 1 ppmis applied to monitoring of drinking water, then everylimit shown on table C-5 would be exceeded in theEDA. If the conservative NBS estimate is not appliedto drinking water because it is to be expected thatdrinking water would be cleaner than other waters andsoils, finding concentrations in the 1-ppm range in

1 5Krammer, op. cit.

other media would still show that contamination ofthe EDA was widespread. In that case, the chance ofhuman exposure would have to be reckoned as sub-stantial.

Some toxic chemicals exhibit “synergism,” i.e., thetoxic effect of simultaneous or sequential exposure totwo (or more) chemicals greatly exceeds the toxic ef-fects predicted from adding together the effects of theindividual chemicals. Without consideration of syn-ergism and with consideration of only additive effectsof chemicals, OTA would not consider the EDA hab-itable if many of the 150 chemicals were present at con-centrations near 1 ppm. For instance, if 10 carcinogensare present in concentrations such that each one posesa 1 in 100,000 chance of a person developing cancer,then the 10 together may pose a 1 in 10,000 risk, whichmay well be so high as to be unacceptable. For the veryreason that so little is known about carcinogenic poten-tials and other toxic potentials, OTA would come tothe conclusion that the uncertainties about health ef-fects from many chemicals being present at near 1 ppmeach would preclude considering the declaration areato be habitable.

However, it is impossible to interpret the NBS opin-ion as supporting the idea that all chemicals for whichMDLs were reported might be present in concentra-tions near 1 ppm. First, the ability of laboratories todetect chemicals varies from substance to substance.The basis of the NBS conclusion, that no concentra-tions higher than 1 ppm would have gone undetectedmust be based on consideration of the properties ofthe chemicals most difficult to detect and measure.Therefore, the MDLs for chemicals that are more easilydetectable must be lower, perhaps in the low ppb rangeclaimed by EPA. The second reason is that there is littlereason to believe that all 150 chemicals monitored byEPA were actually present in significant amounts inthe Love Canal dump. Therefore, to assume that allthe 150 chemicals could be present at concentrationsof up to 1 ppm poses an immediate question about theorigin of all these chemicals. OTA’s concentration ona subset of chemicals known to be present in the land-fill eliminates the problems associated with assigninga possible concentration to chemicals that are not pres-ent.

It would be a tedious task for EPA to supply esti-mates of variance to support the contention that allMDLs were in the low ppb range. However, it mightbe a manageable job for EPA to examine the recordsfor the 16 or so chemicals known to be in the landfillin significant amounts. Because the argument aboutpossible health effects hangs on knowing the absoluteconcentrations of chemicals in the EDA, further anal-ysis of the EPA data seems worthwhile.

45

Uncertainties in Potential Health Effects

Adverse health effects from exposures to toxic sub-stances are conveniently divided into two broadgroups: acute and chronic. Acute effects are observedsoon after exposure, typically to relatively large con-centrations of toxic materials. For example, when aruptured train car spilled nitric acid and the wind car-ried fumes into residential areas of Denver, Colo., peo-ple experienced choking and difficulty in breathing.Less spectacular and more common effects include eyeirritations from air pollutants during periods of poorair quality.

Acute toxic effects are marked by the body’s re-sponding to an insult from ingestion, inhalation, ordermal exposure to a substance. When the insult is re-moved, the affected cells and organ systems of thebody may recover, may die, or maybe replaced. Im-portantly, new body cells, those formed after the in-sult has passed, are not affected.

Chronic toxic effects encompass three dread events:mutations, cancer, and birth defects. These may resultfrom exposure to extremely low concentrations andmay not be observed until years after the exposure oc-curred. (Neurological disorders, which may be causedby low doses of some materials, e.g., lead, are alsochronic health effects, but they are not discussed inthis paper. ) Mutations and cancer differ from acutetoxic effects in that cells are altered genetically and thedamage caused is perpetuated in progeny cells formedfrom the one originally harmed. In contrast, birthdefects that result from in utero exposure of the fetusto chemicals do not necessarily involve genetic altera-tions. Some may result from biochemical changes incritical developmental processes. Because most con-cerns about health risks to EDA residents has centeredon chronic effects, the OTA review focused on muta-tions, cancer, and birth defects.

Mutagenesis and Cancer

Mutagenesis —the causation of mutations—is thebest understood chronic effect, from the standpoint ofmechanism. An environmental contaminant interactswith the DNA of a germ (reproductive) cell and altersthe genetic information within it. If that germ cell, anegg or a sperm, is involved in the formation of an or-ganism, every cell in the new organism will bear thealteration, the mutation. If this organism has progeny,half of those progeny, on average, will bear the muta-tion. Thus, mutations are chronic in the sense that onceintroduced into a population they may be propagatedin every succeeding generation. Some mutations arebeneficial, but most of those that are detected in

humans are associated with deleterious effects.16

Cancer also involves an interaction between a con-taminant and DNA, but the mutational event occursin a somatic, or body, cell rather than in a germ cell.Thus, the mutation is not passed on to the next genera-tion. Instead, a mutation that results in cancer causesrapid proliferation of cells. The rapidly growing cells,all of which may derive from a single mutated cell,in turn, produce a tumor.17

Many mutational events, whether they occur in so-matic or germ cells, may have no effect because theycause changes in DNA without biological conse-quences. Others may produce small but undetectedchanges, either beneficial or detrimental. Although itis likely that only a few DNA changes produce a de-tectable mutation or tumor, our awareness of muta-tional events in humans has been heightened by in-creasing knowledge of their sometimes devastatingeffects.

Methods for Identifying Health Effects

During recent years, much effort has been expendedin identifying carcinogens, agents that cause cancer. 18

The methods used for identifying cause-effect relation-ships between manmade or natural substances and tox-ic effects in humans can be illustrated by a discussionof the methods used to identify carcinogens. Effectsfrom carcinogens (and toxic substances, in general) canbe identified through results of epidemiology—thestudy of diseases and their causes in human popula-tions—and various laboratory tests.

Epidemiology is the only method that establishesassociations between a substance and human toxici-ty. However, it is limited as a technique for identify-ing chronic effects that appear years or decades afterexposure, because people are difficult to study, movefrom place to place, change their work environment,and change their living habits. Also, it is hard to locatethose people who may have been exposed to a par-ticular carcinogen several years previously. Estimatingpast exposures to suspect agents is very difficult.

Testing suspected chemicals in laboratory animals,generally rats and mice, is the backbone of current tox-ic substances identification. A chemical is administeredto animals either in their food, water, air, or (less fre-quently) by force feeding, skin painting, or injection.

‘bFor examples see The Rok of Genetic Testing in the Prevention of@-cupational Disease (Washington, D. C.: U.S. Congress, Office of TechnologyAssessment, OTA-BA-195, April 1983).

IT-ment of TK~olog;m for Btermining Cancer fiks from t~ fi-

vironment (Washington, D. C.: U.S. Congress, Office of Technology Assess-ment, OTA-H-138, June 1981.

18Ibid.

46

The animals are observed over a specified time periodto identify acute or chronic effects.

The reliability of animal tests, bioassays, dependson their design and execution. Guidelines for cancerbioassays were published by the National Cancer In-stitute in 1976. Bioassays now cost between $100,000and $1 million and require up tos years to complete.Clearly such expensive tools can be used only to testhighly suspect chemicals, and much effort is devotedto selecting chemicals for testing.

Molecular structure analysis and examination ofbasic chemical and physical properties are used tomake preliminary decisions about the likelihood of achemical being toxic and whether or not to test it. Forinstance, greater suspicion is attached to chemicals thatshare common features with identified toxic sub-stances. “Paper chemistry” and “paper toxicology” areused most extensively to estimate properties of newchemicals. 19 Unfortunately, not all chemicals withina single structural class behave similarly; thus, limitsare placed on the use of these approaches.

New developments in laboratory testing has resultedin the greater use of short-term (a few days to months)tests. Test costs range from a few hundred dollars toa few thousand. Such tests depend on measuring bio-logical interactions between chemical and DNA. Thebest known test, the “Ames test,” measures mutationsin bacteria. Other short-term tests use nonmammalianlaboratory animals, as well as cultured human andanimal cells. Some tests measure mutagenicity; othersmeasure either the capacity of a chemical to alter DNAmetabolism or to transform a normal cell into one withabnormal growth characteristics. Problems in inter-preting mutagenicity tests arise from the ease of do-ing them; the possibility of false-positive tests increaseswith the number of tests that are done, and since neg-ative tests frequently are not reported, there is somedanger of overrelying on positive test results. Furthercomplicating interpretation, a substance may test outas a mutagen in one assay system and not in another.

A critical problem in estimating human health ef-fects is the need to extrapolate results from studies in-volving large concentrations of chemicals to expectedresults from exposure to low levels actually seen in theenvironment. The idea of dose response (that the per-centage of people suffering adverse effects will decreaseat lower exposures) is well accepted. However, the ex-act relationship between dose (exposure levels) and re-sponses (numbers of affected people) is disputed. Inparticular, some knowledgeable observers argue thatthere are doses of chemicals so low that they will cause

19OTA, The Information Content of Premanufacture Notice, OTA-BP-H-17, Government Printing Office, Washington, D. C.: 1983.

no disease. In other words, a threshold has to be ex-ceeded before any adverse effects will be seen. In gen-eral, thresholds are better accepted for acute effectsthan for chronic effects. In particular, if the interac-tion between a single molecule of a chemical and DNAis sufficient to produce a mutation, no threshold valueis likely for mutagenic and carcinogenic effects.

The problems of extrapolation are more complexwhen data from laboratory studies are the only avail-able information. In those cases, a method must bechosen to translate the meaning of a toxic effect in theanimal to an expected toxic effect in humans. Almosteveryone accepts that animal results are important topredicting human effects; toxicology is based on thatpremise. However, there can be endless argumentsabout the applicability of a particular animal test.

A 1979 IARC report summarized the agency’s anal-ysis of 354 chemicals and chemical processes that ithad reviewed in its program, which began in 1971.20

IARC found sufficient epidemiologic information toevaluate carcinogenicity in humans for fewer than 100chemicals. For 18 of those, IARC considered that theevidence was sufficient to support a conclusion thatthe chemical causes cancer in humans. For an addi-tional 18, the evidence was sufficient to support a con-clusion that the agent was a probable human car-cinogen. In the cases of the remaining 318 chemicalsand chemical processes, the data from human studieswere insufficient to support a conclusion that thesubstance or process is a human carcinogen or a prob-able human carcinogen.

IARC also reviews the worldwide literature aboutthe testing of chemicals for carcinogenicity in animals.About animal tests, it says:

. . . in the absence of adequate data in humans it isreasonable, for practical purposes, to regard chemicalsfor which there is sufficient evidence of carcinogenici-ty (i.e., a causal association) in animals as if theypresented a carcinogenic risk for humans.IARC has reviewed the literature about the testing

of 354 chemicals in animals. For 142 of those, IARCconsidered that the animal evidence was “sufficient,”and that those substances should be considered to posea carcinogenic risk for humans. In 1980, scientists atthe National Cancer Institute (NCI) reviewed all of thecancer tests carried out in animals there.21 In the caseswhere both IARC and NCI evaluated data about thesame chemicals, the results of the two organizations’analysis were generally consistent .22

20Chemicals and Industrial Processes Associated With Cancer in Humans,Monographs, Supplement 1, (Lyon, France: IARC, 1979).

~lR. A. Griesemer and C. Cuets, Jr., ‘Toward a Classification Scheme forDegrees of Experimental Evidence for Carcinogenicity of Chemicals for Ani-reals, ” in Mokctdarand Cellular Aspects of C%ucinogen f%wming Tests, I-I.Bartsh and L. Tomatis (eds. ), Lyon France, 1980.

‘lOTA, 1981, Op. cit .

4 7

Comparing the relatively small number of chemicalsfor which human evidence is available to the largernumber of chemicals which has been tested in animalsillustrates the importance of animal tests. The numberof tested chemicals is much smaller than the numberof all chemicals, and there are major efforts underwayto use “short-term tests, ” most of which measuremutagenicity to provide information about carcino-genicity.

The IARC review provides an example of scientistsand policymakers wrestling with the problems of ex-trapolating results of epidemiology and animal testsabout toxicity to estimates of human effects. Althoughfindings of IARC are not binding on governments,they are generally accepted as authoritative and pro-vide an example of a successful ongoing effort toevaluate scientific evidence. A number of approachesto evaluating evidence about carcinogenicity and othertoxicities for regulatory purposes, all of which involvea centralized panel of experts to consider the toxicityof substances, have been advanced by Governmentagencies,

23 Members of Congress, and by trade asso-ciations. 24 A recent National Academy of Sciencescommittee document argued against a central commit-tee for making decisions about carcinogenicity for allGovernment agencies, but urged that a central com-mittee be formed and charged with developingguidelines for making those decisions.25

To a major extent, the interest in expert review stemsfrom a desire to grapple with uncertainty. Uncertain-ties in estimating the hazards posed by chemicals resultfrom difficulties with test design and execution, thescantiness of data, and methods for extrapolation. Theactivities of an expert panel (e.g., IARC) to review dataand conclusions reduce the uncertainties in a few casesinvolving carcinogenicity. In the absence of such ex-pert review, the reader or scientist interested in tox-icity must develop a critical eye, inspect and evaluatethe evidence presented by others, and discuss opinionswith other interested parties.

Health Effects Associated WithLove Canal Chemicals

Toxic effects associated with 18 chemicals knownto have been deposited in the canal landfill are listedin table C-4. The minimum lethal doses for these

chemicals are much greater than environmental con-centrations reported or expected within the EDA (seetables C-5 and C-6). For example, maximum en-vironmental concentrations of chlorobenzene were onthe order of 3 to 5 micrograms per cubic meter of air(table C-6), an amount less than 1/1,000,000 the levelneeded to kill the most sensitive laboratory animal(table C-4).

Data for birth defects or reproductive effects (calledhere “teratogenic effects”) are not available for mostof these compounds. Only pentachlorobenzene andhexachlorobenzene were reported to have been testedfor teratogenicity.

26 Both were found to be positive inat least one test, but the quality of the data was notevaluated.

Mutagenic test results are reported on 12 of the sub-stances, and of these, 8 were positive in at least onetest. In addition, additional mutagenicity tests areplanned for lindane and hexachlorobenzene.27 Thechemical 1,4-dichlorobenzene provides an example ofa chemical that was mutagenic in one test and not inanother. It has been shown to cause mutations in bac-teria, but not in a test involving the use of Chinesehamster ovary (CHO) cells. Nevertheless, whatevercaveats are attached to finding a positive response intesting for a mutational effect, the finding serves towarn of a possible hazard.

The fact that IARC found adequate animal data toevaluate the carcinogenicity of five of the chemicalslisted in table C-4 indicates that there had been con-cern about the carcinogenicity of those chemicals. Inaddition, five chemicals, including three of the IARC-reviewed chemicals, are currently under test at the Na-tional Toxicology Program (NTP). In other words, 7of the 18 Love Canal chemicals have been tested orare being tested for carcinogenicity, This level of ef-fort does not mean that many or most of the chemicalsare carcinogens, but it does mean that scientists haveexpressed sufficient concern about them that tests arenecessary to provide more information.

For three of the five chemicals that IARC reviewed,data were sufficient to conclude that the chemicals arecarcinogenic in laboratory animals. In the other twocases, IARC reached the conclusion that there was“limited” evidence rather than “sufficient” evidence tosupport a conclusion that the chemicals were carcino-gens.

‘~STp, “Identification, Characterization, arid Control of Potential HumanCarcinogens: A Framework for Federal Decision-Making,” J. Natioml Gncerhstitute, 64:169-176, 1980.

Mmc I+oP~I hr a scie~ce panel (Searsda]e, N. Y.: AIHS, ~9W).~~Natio~ R=mch Counci], Risk Assessment in tk k&ra] &vemment:

Managing the Process (Washington, D. C.: National Academy Press, 1983).

lb~e Department of Health and Human Services, Report of the %bcom-mittee on the Potential Health Effkts of Toxic Chemicals Dumps of theDHEW Committee to Coordinate EnvkonmentaI and Related Programs,undated.

27National Toxicology Program, National Toxicology Program: Fiscal Year1983 Annual Plan, Research Triangle Park, N. C., 1982.

Table C-4.–Summary of Test Results Available on Health Effects of Chemicals Disposedin Love Canal and Monitored by EPA

Regulation or standardd

Substance Minimum lethal dosea Mutagenicity b Carcinogenicity C ACGIH OSHA EPA

Lindane (gamma-hexachlorohexane) . . Ingestion, animal,

Chlorobenzene . . . . . .

1,2-Dichlorobenzene . .

1,3-Dichlorobenzene . .1,4-Dichlorobenzene . .

1,2,3-Trichlorobenzene1,2,4-Trichlorobenzene

1,3,5-Trichlorobenzene

180 mg/kg

. . . . . . . . . . . . . . Inhalation, animal,15 g/m3

. . . . . . . . . . . . . . Inhalation, animal,821 ppm

. . . . . . . . . . . . . . —

. . . . . . . . . . . . . . Man, 221 mg/kg

—. . . . . . . . . . . . . .. . . . . . . . . . . . . . Ingestion, animal,

758 mg/kg. . . . . . . . . . . . . . Implant, animal

LD50 20 mg/kg1,2,3,4-Tetrachlorobenzene . . . . . . . . . . —1,2,4,5-Tetrachlorobenzene . . . . . . . . . . Ingestion, animal,

Pentachlorobenzene f

Hexachlorobenzenef.

2-Chloronaphthalene

Alpha-Chlorotoluene.

2-Chlorotoluene. . . . .

3-Chlorotoluene. . . . .4-Chlorotoluene. . . . .2.4-Dichlorotoluene. .

1,035 mg/kg. . . . . . . . . . . . . . . Ingestion, animal,

2,000 mg/kg. . . . . . . . . . . . . . . Man, 220 mg/kg

. . . . . . . . . . . . . . . Ingestion, animal888 mg/kg

. . . . . . . . . . . . . . . Ingestion, animal,LD50 1,200 mg/kg

. . . . . . . . . . . . . . . Inhalation, animal,175,000 ppm

. . . . . . . . . . . . . . . —

. . . . . . . . . . . . . . . —

. . . . . . . . . . . . . . . —

Cytogenic changesto be tested (NTP)negative (CCERP)

—

Positive (CCERP)

Positive (CCERP)Point mutagen,negative, CHOtest (NTP)

Negative (CCERP)Negative (CCERP)

Negative (CCERP)

Positive (CCERP)

Positive, CHO(NTP)

Positive (CCERP)to be tested (NTP)

—

Point mutagenDNA replicationpositive (CCERP)

Negative (CCERP)

Animal + (IARC)Animal - (NCI)

Under test (NTP)

Animal ? (IARC)Animal - (NTP)

Animal ? (IARC)Under test (NTP)

—

Under test (NTP)

Animal + (IARC)

Animal + (IARC)Under test (NTP)

Yes

Yes

Yes

—Yes

—Yes

—

——

—

—

—

Yes

Yes

———

Yes

Yes

Yes

—Yes

——

—

——

—

—

—

Yes

Yes

———

National Drinking Water StandardW.Q.C.e

W.Q.C.

W.Q.C.

W.Q.C.W.Q.C.

W.Q.C. insufficient dataW.Q.C. insufficient data

W.Q.C. insufficient data

W.Q.C.W.Q.C.

W.Q.C.

W.Q.C.

W.Q.C. insufficient data

—

———

aAll data in this column are from 1980 Registry of Toxic Effects of Chemicals SUbStances (NlOSH, 1982).bReferences for data in this column are from NIOSH (1982) unless otherwise indicated: NTP is National Toxicology Program: Fiscal Year 1983 Annual Plan (NTP, 1982); and CCERP is Report of the Subcommitteeon the Potentia/ Health Effects of Toxic Chemica/ Dumps of the DHEW Committee to Coordinate Environmental and Related Programs (Department of Health and Human Services, undated).

The appearance of a test name means that the chemical was found to cause the named effect: cytogenic changes: microscopically visible chromosomal changes; point mutagen: chemical altered a specificgene in the test organism, a standard test; CHO test: a test of the capacity to alter growth patterns of Chinese hamster ovary cells, a standard test; DNA replication: a test of the capacity to alter DNA replication,a standard test; positive: CCERP reported that at least one test has shown the chemical is a mutagen; the quality of the data was not reviewed; and negative: CCERP reported that the chemical had been testedand none of the results showed the chemical to be a mutagen; the quality of the data was not reviewed.CReferences: IARC is IARC Monographs Supplement 1 (IARC, 1979); NCl is R. A. Griesemer, and C. Cueto, Jr. in Molecular and Cellular Aspects of Carcinogen Screening Tests (IARC, 1980); and NTP is same as under (b). + means the agency judged the substance to be an animal carcinogen; ? means the evidence about carclnogenicity was limited; and – means the evidence was negative.

dAcronyms: ACGIH—American Council of Government lndustrial Hygienist, Osha-Occupational Safety and Health Administration, EPA—Envimnmental Protection Agency, IARC—lnternational Agency for Research

on Cancer, NCI—National Cancer Institute, and NTP—National Toxicology Program.eW.Q.C.: Water Quality Criteria Document (45 F. R., 11/28/80). W.Q.C. means EPA recommended a standard; W.Q.C. insufficient data means that there were insufficient data to base a standard.fAt least one test result indicates that this substance has teratogenic properties.

4 9

Table C-5.—Comparison of Regulated Exposure Limits to Detected Maximum Concentrationsin the EDA: Water

Ratio:Regulated limit Maximum concentration found in EDAa maximum detected level/

Substance (exposure through ingestion) [concentration (medium)] standardOne substance regulated under National Drinking Water Standard:Lindane . . . . . . . . . . . . . . . . . . 4.0 ppbb 5.3 ppb (sanitary sewer) 1.3

3.4 ppb (storm sewer) 0.85Substances for which water quality criteria have been published:Lindane . . . . . . . . . . . . . . . . . .

Chlorobenzene. . . . . . . . . . . .1,2-Dichlorobenzene . . . . . . .1,3-Dichlorobenzene . . . . . . .1,4-Dichlorobenzene . . . . . . .1,2,4,5-Tetrachlorobenzene. .

Hexachlorobenzene . . . . . . .

No safe limit (carcinogen)1 0-5 risk,water and fish, 0.186 ppb 5.3 ppb (sanitary sewer)

3.4 ppb (storm sewer)fish only, 0.625 ppb 5.3 ppb (sanitary sewer)

3.4 ppb (storm sewer)488 ppb Trace c (shallow well)400 ppb 15 ppb (deep well)400 ppb 80 ppb (sump)400 ppb 586 ppb (sump)Water and fish, 38 ppb 62 ppb (storm sewer)fish only, 48 ppb 62 ppb (storm sewer)No safe limit (carcinogen)1 0-5 risk,water and fish, 7.2 pptb Trace (sump water, sanitary sewer)fish only, 7.4 ppt Trace (sump water, sanitary sewer)

28.518.38.55.4

0.040.21.460.610.77

aM u c h h i g h e r c o n c e n t r a t i o n s , s o m e t i m e s i n e x c e s s o f 1 0 , 0 0 0 p p b , w e r e f o u n d i n s u r f a c e w a t e r s e d i m e n t n e a r s e w e r o u t f a l l s a n d i n s e w e r s e d i m e n t s . T h o s e t w o c o n -

taminated media are to be cleaned up in the remediation process.bppb: parts per billion, 1 µg of chemical/liter water; ppt: parts per trillion, 0.001 µg of chemical/liter water.cTrace: detectable, but not measurable concentrations.

SOURCE: From several published sources.

Table C-6.—Comparison of Regulated Exposure Limits to Detected Maximum Concentrationsin the EDA: Air

Maximum concentration Ratio:Regulated exposure limita found in EDA maximum detected level/

Substance (exposure through inhalation) [concentration (medium)] standardLindane . . . . . . . . . . . . . . . . . . . . . . . . . 500,000 µg/m3 0.098 µg/m3 (living area) <0.001

(OSHA, ACGIH)b

Chlorobenzene. . . . . . . . . . . . . . . . . . . 350,000 µg/m3 3.5 µg/m3 (basement)(OSHA, ACGIH)

<0.001

1,2-Dichlorobenzene . . . . . . . . . . . . . . 300,000 µg/m368 µg/m3 (living area) <0.001

(OSHA, ACGIH)1,4-Dichlorobenzene . .............450,000 µ/m 3 25 µg/m3 (living area)

(OSHA, ACGIH)<0.001

1,2,4-Trichlorobenzene . . . . . . . . . . . . 40,000 µg/m3 (ACGIH) Tracec (living area) <0.0012-Chlorotoluene . . . . . . . . . . . . . . . . . . 250,000 µg/m3 (ACGIH) 8 µg/m3 (living area) <0.001aLimits are generally expressed in units of milligrams/cubic meter. To facilitate comparison of limits and maximum concentrations, limits are converted to units of

micrograms/m 3 here.bOSHA—Occupational Safety and Health Administration; ACGIH—American Council of Government Industrial Hygienists.cTrace: detectable, but not measurable concentrations.

SOURCE: From several published sources.

The National Cancer Institute’s (NCI) evaluation of it is an animal carcinogen. IARC concluded that lin-the carcinogenicity of lindane differed from that of dane was a carcinogen.IARC.28 Data from NCI did not support the idea that For 1,2-dichlorobenzene, IARC found that there was

only limited information about carcinogenicity. A sub-sequent test by NTP reveals that the substance does

2 1G r i esemer, op. cit. not cause cancer in either rats or mice. Therefore, ad-

50

ditional evidence has reduced the level of concernabout possible carcinogenic effects for that chemical.A related chemical, l,4-dichlorobenzene is now undertest at NTP as is alpha-chlorotoluene. Data about thecarcinogenicity of chlorobenzene and pentachloroben-zene under test at NTP, have not been reviewed byIARC.

Table C-4 also shows that workplace and environ-mental exposures to many of these chemicals are reg-ulated. The Occupational Safety and Health Admin-istration (OSHA) regulates workplace exposure, andrestrictions on workplace exposures have been recom-mended by a group of industrial health experts, theAmerican Conference of Government Industrial Hy-gienists (ACGIH). EPA has published water qualitycriteria for 13 of the 18 chemicals. In addition, a Na-tional Drinking Water Standard regulates exposure tolindane.

The data summarized in table C-4 show that thechemicals deposited in the canal landfill include somerecognized as presenting hazards to human health. Animmediate objection to drawing any conclusions abouthealth effects from these chemicals at Love Canalderives from EPA’s observations that the concentra-tions within the EDA were very low.

Comparison of Regulated Exposure Levels andEnvironmental Concentrations

EPA has published National Drinking Water Stand-ards for 16 inorganic chemicals, 6 pesticides (includinglindane), 1 group of organic chemicals, and total dis-solved solids. As is shown on table C-5, the maximumconcentration of lindane found in one sample eachfrom sanitary sewers and storm sewers slightly ex-ceeded that limit. It should be emphasized that whilethis concentration is higher than the drinking waterstandards, the water in both systems is not likely tobe ingested by humans.

Water quality criteria documents have been pub-lished for 64 chemicals to serve as guidelines for ac-ceptable concentrations in drinking and fishing waters.There was some emphasis on protecting against car-cinogenic risks in the criteria documents, and, in keep-ing with the idea that there is no dose of a carcinogenbelow which there is no risk, the Agency declared thatthere was no safe limit for carcinogens. Instead, it cal-culated the amount of the substance, that if ingestedover the course of a lifetime, would cause an incremen-tal risk of cancer equal to 1 case of cancer in 100,000people. The magnitude of that risk can be judged bycomparison to the figure that about 20 percent ofAmericans (or 20,000 out of every 100,000) die fromcancer. As is shown in table C-6, the maximum de-

tected concentrations of lindane in sewers exceeded thestandards for water to be used for drinking or fishing.The measured levels of 1,2- and 1,3-dichlorobenzeneand 1,2,4,5-tetrachlorobenzenes were less than the lim-its established by the water criteria documents. Only1,2-dichlorobenzene, found at 0.04 of the recom-mended guideline, was detected in deep wells and like-ly would be associated with human ingestion. No nu-merical measurements were reported for chloroben-zene and hexachlorobenzene, and EPA claims that con-centrations of those chemicals would be in the low ppbrange. If the concentrations are that low or lower, thelevel of chlorobenzene would be below that recom-mended by EPA.

Hexachlorobenzene presents an analytical problem.The water quality criteria document associates a 10-5

cancer risk with a 7 ppt concentration of hex-achlorobenzene. However, that chemical cannot bemeasured at concentrations lower than a few ppb.Thus, hexachlorobenzene could be present in watersamples in the Love Canal study in concentrations upto 1 ppb, 130 times the level associated with a 10-5

cancer risk. But such concentrations are possible in allwater; methods are not available to measure this chem-ical at 7 ppt. The fact that hexachlorobenzene was de-tected only in waters that humans do not drink or fishmeans that opportunities for exposure are limited.

To summarize, measured concentrations of somechemicals found in sewer and sump waters in the EDAapproached or exceeded levels recommended fordrinking and fishing waters. However the contam-inated waters in the EDA are not likely to be con-sumed, and exposure through ingestion is unlikely.

OSHA or ACGIH or both have established limitson workplace exposure to six of the chemicals. In theworkplace, concern is about exposure by inhalationor through the skin. In the case of lindane, the ex-posure limit shown on table C-6, based on inhalation,is to be further lowered if there is any chance of thechemical reaching the worker’s skin. The workplaceexposure limits are based on consideration of acutetoxic effects and are designed to protect againstworkers’ becoming ill soon after exposure. They arenot designed to protect against any chronic healtheffect-cancer, birth defects, or mutations. Therefore,it is no surprise that the maximum levels of airbornecontamination found in the EDA is much less than theworkplace limits. Although there is little reason tobelieve that the workplace limits protect againstchronic health risks, the airborne concentrations atLove Canal are much lower, and are as low as levelsdetected in many areas of the country.

EPA can consider chronic health effects in settinglimits for air pollutants under the Clean Air Act. To

51

date, it has not published nor made public any con-sideration of possible regulations of the chemicals listedin table C-6. Therefore, there are no standards orguidelines to compare to the detected levels. EPA didcompare airborne concentrations of chemicals in theLove Canal area to concentrations in cities around thecountry, and there were no striking differences re-ported.

The Special Case of Dioxin

Dioxin (more precisely, 2,3,7,8-tetrachlorodibenzo-p-dioxin, or 2,3,7,8-TCDD, or TCDD) is one of themost toxic substances. It is known to cause cancer inlaboratory animals, and it has been associated withtumors in humans. The “buyout” of Times Beach,Me., was based on the premise that dioxin present atconcentrations greater than 1 ppb in the soil presenteda health threat.

There is no Federal standard that restricts envi-ronmental exposure to dioxin. EPA has carried out anassessment of the health risk posed by the presence ofdioxin in incinerators. The Agency concluded that0.0004 micrograms/m 3 of stack air, which would bediluted 100,000-fold by the time it reached groundlevel, where it might be inspired, would not “presenta public health hazard. ”29 The government of the Prov-idence of Ontario has set a permissible limit for diox-in in air equal to 0.00003 micrograms/m 3, which isabout 1/10 the level found by EPA in the incineratorstacks. (An average person breathes about 20 m3 ofair daily. Over a 70-year lifetime, a person breathingthe maximum limit permitted by Ontario would in-hale 219 micrograms of dioxin. )

One of the surprises from Times Beach is the obser-vation that dioxin is very stable in soil. The stabilityis probably related to the fact that dioxin binds veryfirmly to particles in the soil, and being bound pro-tects it from degradation. It is so difficult to extractdioxin from soil, that it may be that the chemical isnot removed from soil particles that are inhaled or in-gested. If that is the case, soil-bound dioxin would poselittle threat to human health. NTP is currently con-ducting studies about the bioavailability of soil-bounddioxin and expects to have results by the end of thesummer of 1983. It has already been reported that rootcrops, such as carrots, that were grown in dioxin-con-taining soil did not take up appreciable quantities ofdioxin .30 The low levels found with the carrots mighthave resulted from contaminated soil sticking to theoutside of the root.

29D. Barnes,U.S. EPA, personal communication, May 1983.JOR Kimbrou~t, Center for Disese Contro], Atlanta, Ga., personal com-

munication, May 1983.

EPA has been working on a water quality criteriadocument for dioxin for some time, Although it is notknown what level the final document will recommend,a draft suggested that dioxin levels should not exceed1 part in 1016 parts of water. This very low concen-tration presents analytical difficulties. Although bothEPA and the Canadian Government have perfectedmethods to measure low levels of dioxin in water, theminimum detection level are now 100-fold higher,about 1 part of dioxin in 1014 parts of water. In otherwords, water with no detectable levels of dioxin mightharbor 100 times the concentration that may be recom-mended as a guideline to protect health. (See discus-sion above, also of hexachlorobenzene. )

EPA claims that it was able to detect 1 to 20 ppt(0.001 to 0.020 ppb) dioxin in the samples taken atLove Canal. OTA has serious reservations about theintensity of sampling for dioxin, but laying those asidefor the moment, EPA reported dioxin in sumps in theLove Canal homes and in storm sewers. The Agencyreported finding no dioxin in air, which is not surpris-ing because it has not been reported in air except instack gases. And no dioxin was reported in waters.The Ring 1 homes are not being considered for rehab-itation, and the storm sewers are to be cleaned up aspart of the remediation effort. Therefore, the knownsources of exposure to dioxin are going to be elimi-nated.

OTA is concerned that few samples were analyzedfor dioxin in the EPA study. If the decision is madeto rehabitate the EDA, and, if, as part of that effort,monitoring is carried out, dioxin should be includedas a monitored substance more frequently than it wasin the EPA study reported in 1982. It may be that thereare good reasons for the sketchy sampling carried outearlier, but it would not reassure the public to samplethis important contaminant any less frequently thanother chemicals.

Studies of Health Effects in theLove Canal Population

Although the habitation decision was based on con-sideration of the extent of chemical contamination inthe EDA, there are other data that also bear on thequestion of health effects. Those data derive fromobservations made on the population of people wholived near the Canal. Love Canal, A Special Reportto the Governor and Legislature, in 1981 summarizesevidence collected by New York Department of Healthofficials until that time.31

Jl~sj~H, ~Ve Canal, a Special Report to the Governor and J%-islature, 1981. See also, Love C&naL Public Health Thne Bomb, NYS/DOH,1979.

52

Briefly, pregnant women who lived near the canalwere found to be at greater risk of suffering a miscar-riage or of delivering a ‘low birth-weight baby.” Morefocused research into the location of residences asso-ciated with these adverse effects revealed that thewomen at greatest risk lived either on 99th Street,directly adjacent to the canal, or in formerly wet areasjust east of the canal. The New York State analysisof these data led to the conclusion that the frequencyof spontaneous abortion (miscarriage) reached a peakin the 1960’s and early 1970’s. In addition, the percent-age of children born with birth defects was largeramong those delivered by women who lived on 99thStreet or in formerly wet areas as compared to womenwho lived beyond what is now the declaration area.The excess of birth defects was not, however, statis-tically significant. (Neither the study of reduced birthweight nor birth defects has been published in a peer-review scientific journal. )

Cancer rates of residents of the census tract in whichLove Canal is located have been compared to cancerrates in other census tracts in Niagara Falls. The femalepopulation around Love Canal was found to have ex-perienced twice the number of respiratory cancers asa control female population; the excess among menwas less, respiratory cancer among male Love Canalcensus tract residents was 1.7 times that observed inthe control population. 32 Scientists differ in what im-portance to attach to excesses of cancer that are lessthan 2, but most agree that a twofold excess, as wasseen for respiratory cancer in women, merits attention.Further analysis showed that the incidence of respira-tory cancer within the Love Canal census tract doesnot vary with distance from the canal, which weakensthe argument that substances in the canal were asso-ciated with the excess cancers. The incidence of somecancers, lymphoma, leukemia, and liver cancer amongmen living in the Love Canal census tract was onlyhalf that observed in the control areas. The incidenceof genital organ and urinary cancers among womenwas lower among Love Canal area residents thanamong women in other areas of Niagara Falls.

Any conclusions to be drawn from the study of can-cer incidence around Love Canal are weakened by thesmall number of cancers that occurred during theperiod (1966-77) over which the study was conducted.Furthermore, since cancers may develop years or dec-ades after exposure, the study done by New York Statemay have been too early to detect an effect, if thereis one. A better answer to whether or not living nearthe canal is associated with higher cancer rates may

J~Janerich, et a]., “cancer Incidence in the Love Canal Area, ” science 212,

1981, pp. 1404-1407.

become available as more people who lived in thecanal area are located and studied.33

In May of 1983, the DHHS released a study ofchromosomal abnormalities in a small population ofpeople who had lived near the canal.34 The results ofthat study were negative; that is, the frequency ofunusual chromosomes among the canal area residentswas no greater than the frequency found in a controlpopulation. There are some problems with this study.(There may never have been a study without someproblems. ) In particular, in the opinion of many scien-tists, a chromosomal abnormality caused by exposureto toxic chemicals may be short lived and may be re-paired over time. Therefore, since the exposures oc-curred years ago, there might be no discernible effectfrom them now. Despite that reservation (so far asOTA can determine, that reservation is based on tech-nical opinion and not upon prolonged observation ofhuman populations) and other technical reservations,the consensus is that the study does not show anychromosomal abnormalities as a result of living nearLove Canal. This negative study provides evidencethat an earlier study, which detected a high frequen-cy of chromosomal abnormalities among Love Canalarea residents might have been in error.

Also in May 1983, Beverly Paigen and coworkerspresented a paper at a meeting of the Society forPediatric Research.35 They compared health effectsobserved in the population of people who had livedin the EDA to effects observed in groups of people wholived in control areas. They reported that low birth-weight babies were more common among the EDApopulation and that the average weights of babies bornthere were lighter at each week of gestation. Further-more, children born and raised (for at least 75 percentof their lives to date) in the EDA were shorter thanchildren in the control areas, and the parents reportedthat these children had more episodes of six differentmedical complaints than control area children.

Paigen confirmed that low birth weights were con-fined to families who had lived in the formerly wetparts of the declaration area.36 The medical complaintsin children were not restricted to families living in thewet areas, but decreased with distance from the canal.She thinks that the play of children might bring theminto closer contact with contaminated areas, and dis-

33C. WI. Heath, Jr. Assessment of Health Risks at Love Canal, presentedat the Fourth Annual Symposium on Environmental Epidemiology, Pitt-sburgh, Pa., May 1983.

sac, W. Heath, Jr., et al., A Study of Cytogenetic piitte171S h pemflS h“v-

ing Near the Love Card Center for Disease Control, Atlanta, Ga., May 1983.JSB pa~en, et al., Growth ~d Health in Children Livins Near a fizar~

ous Waste Site, presented at Society for Pediatric Research, May 1983.S*B. paigen, Children’s Hospital, Oakland, Calif., personal communica-

tion, May 1983.

—

53

tance from the canal would decrease the frequency ofthe children reaching such areas.

To accumulate the data in the paper was difficult.The 220 births reported in the declaration area and the697 in the control areas occurred over a 15-year inter-val. Analyzing such data requires careful attention todetail, and reviewing the data and analysis requiresan opportunity to review details. At the present time,the results presented by Paigen, et al., are being pre-pared for publication, and final evaluation of theirwork must wait until the paper, with more detaileddescriptions of the study and data, is complete.

A major study of health effects is expected sometimethis fall. The NYS/DOH has located as many as possi-ble of all the people who lived in the canal area sincethe 1940’s.37 Those people were interviewed about theirhealth status, and the results of that study will pro-vide important information about the health of formerresidents of the Love Canal region, and current andformer residents of the EDA.

37N, Vianna, NYs/mI-1, personal communication, May 1983.

The completed studies are sufficient to show thatthe Love Canal population has not experienced anyadverse health effects at rates more than twice thoseexperienced by control area populations. Spontaneousabortions and low birth-weight babies were statistical-ly more frequent among the populations in the LoveCanal region, and birth defects, although not sta-tistically more frequent, were observed more often inthat population. The incidence of some cancers in thearea has been higher than in control areas; the in-cidence of other cancers has been lower.

As more data become available in the near future,it may become possible to draw firmer conclusionsabout the health impacts of living near Love Canal.Unsatisfying as it is, the consensus of opinion probablyis that the studies of the Love Canal area populationhave produced more leads to follow up than strongconclusions about the safety of the EDA. Evidence ismore convincing that serious health effects were as-sociated with living in the Love Canal homes.

Appendix D

Analysis of the EPA Data

Summary

OTA examined a subset (20 chemicals) of the EPAmonitoring data. Differences were noted in frequen-cies of detection for this subset between the emergen-cy declaration area (EDA) and control areas. Fourchemicals (1,2- and 1,3-dichlorobenzene and 2- and4-chlorotoluene) were found to be present in the EDAat greater frequencies than the control areas. The dif-ferences were significant, but the concentrations werelow.

Statistical Analysis ofIndicator Substances

OTA disaggregated the Environmental ProtectionAgency (EPA) monitoring data to allow an examina-tion of a subset of “indicator substances, ” toxicchemicals known to have been disposed in the canallandfill, Twenty chemicals were identified for the OTAanalysis; only 16 of them had sufficient data. Frequen-cies of detection for the 16 substances were comparedbetween the EDA and control areas.’ The method ofanalysis was the Mantel-Haertszel procedure. This pro-cedure assumes similar patterns across those mediathat contain one positive sample; if a chemical is foundin one medium in a region, it is assumed that the chem-ical may be found in the other media in the sameregion. The finding of no detectable levels in bothregions has no effect on the assumptions.

The results of Mantel-Haenszel analysis are shownin table D-5. The EDA was found to have a significant-ly higher frequency of positive detections than the con-trol area for four chemicals: 1,2- and l,3-dichloroben-zene, 2- and 4-chlorotoluene. Table D-1 illustrates theresults for one compound, 1,2-dichlorobenzene. Ahigher rate of positive samples was found in the EDAand the difference is statistically significant. * The oddsof finding a positive sample in the EDA is 7.5 timesgreater than that for the control areas. It should benoted that no comparison can be made with six of thesubmedia, as no control samples were analyzed. TheEPA data used in this analysis are presented in tablesD-2, D-3, and D-4. EPA found significant differencesfor 2-chlorotoluene and 1,2-dichlorobenzene in air be-tween EDA and control areas. It should be noted that

IL. A. Cupples, Boston University School of Public Health, report sub-mitted to OTA, Industry, Technology, and Employment Program, May 19S3.

“Significant refers to a 90-percent confidence level or greater.

5 4

Table D-l.—Odds Ratioa and Detection Ratesfor l,2-Dicholorobenzene

EDA V. Love Canalcontrol area v. control area

7.5 (.001) 4.2 (.05)

Medium EDA Control area

Shallow well. . . . . . . . . . . . . .

Deep well . . . . . . . . . . . . . . . .

Sump water . . . . . . . . . . . . . .

Surface water. . . . . . . . . . . . .

Drinking water . . . . . . . . . . . .

Storm sewer water . . . . . . . .

Sanitary sewer water . . . . . .

Soil . . . . . . . . . . . . . . . . . . . . .

Storm sewer sediment . . . . .

Surface water sediment . . . .

Living area air . . . . . . . . . . . .

Basement air . . . . . . . . . . . . .

Outdoor air. . . . . . . . . . . . . . .

Oatmeal . . . . . . . . . . . . . . . . .

Potatoes . . . . . . . . . . . . . . . . .

Crayfish . . . . . . . . . . . . . . . . .

Mice . . . . . . . . . . . . . . . . . . . .

Worms. . . . . . . . . . . . . . . . . . .

0.0b

47=

0.0428

0.01040.040.0

300.1191.0010.01

1050.07

150.3040.42

5390.34

880.10

83—0

—00.0

310.0

350.0

19

0.0110.0

140.040.0

40.01—

00.09—

00.040.10

30—

0—

0—

0—

o0.090.0

320.05

aThe odds ratio (P-value) indicates the odds of observing a sample withl,2-dicholorobenzene in one region compared to another.

bFrequency of samples containing 1,2-dichlorobenzene.cTotal number of samples analyzed.

SOURCE: Cupples, op. cit.

1,4-dichlorobenzene was found to be more frequent-ly detected in the control areas than in the EDA. Theabsence of significant differences for the other com-pounds may be due to inadequate sampling for con-trols and cannot be interpreted as strong evidence thatthe EDA and controls are similar in levels of contam-ination. It should be noted that these same four chem-icals were detected in leachate sludge obtained at the

55

c

Table D·2.-Detection (+) of ndicator Substance and Numbers of Samples Analyzed (n) for Environmental Media in the Control Areas

t:nVlronmental meala ana suomeala-

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Indicator substance +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n +/n ·/n +/n +/n

Gamma-BHe (Lindane) .......... 3/11 3/15 115 0/5 0/5 0/1 010 0/9 010 010 010 3/5 0/28 010 010 010 010 10/9 010 0/0 1/33 0/5 Chlorobenzene ................. 0/11 3/16 0/5 0/5 115 0/1 010 0/17 010 0/1 010 0/4 0/31 010 010 0/4 0/3 10/0 010 010 010 010 1,2-Dichlorobenzene ............. 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 0/4 3/30 010 010 010 010 10/9 010 010 0/32 0/5 1,3-Dichlorobenzene ............. 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 0/4 0/0 010 010 010 010 10/9 010 010 0/32 0/5 1 ,4-Dichlorobenzene ............. 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 214 116130 010 010 010 010 10/9 010 010 0/32 115 1,2,3-Trichlorobenzene ........... 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 0/4 0/28 010 010 010 010 10/9 010 010 0/32 0/5 1,2,4-Trichlorobenzene ........... 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 1/4 Ol28 010 010 010 010 10/9 010 010 1/32 115 1,3,5-Trichlorobenzene ........... 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 0/4 Ol28 010 010 010 010 10/9 010 010 0/32 0/5 1,2,3,4-Tetrachlorobenzene ....... 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 0/4 Ol28 010 010 010 010 10/9 010 010 10/32 015 1,2,3,5-Tetrachlorobenzene ....... 010 010 010 010 010 010 010 010 010 010 010 010 OlO 010 010 010 010 10/0 010 010 010 010 1,2,4,5-Tetrachlorobenzene ....... 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 0/4 OlO 010 010 010 010 10/9 010 010 0/32 0/5 Pentachlorobenzene ............ 010 010 010 010 010 010 010 010 010 010 010 010 0/28 010 010 010 010 10/0 010 010 010 010 Hexachlorobenzene ............. 0/11 0/14 0/4 0/5 0/4 010 010 0/9 010 010 010 0/4 0/28 010 010 010 010 10/9 010 010 0/32 0/5 2-Chloronaphthalein ............. 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 0/4 OlO 010 010 010 010 10/9 010 010 010 0/5 a-Chlorotoluene ................ 0/11 0/16 0/5 0/3 0/5 0/1 010 0/17 010 0/1 010 0/4 OlO 010 010 0/4 0/3 10/0 010 010 010 010 2-Chlorotoluene ................ 0/11 0/16 0/5 0/3 0/5 0/1 010 0/17 010 0/1 010 0/4 2130 010 010 0/4 0/3 10/0 010 010 010 010 3-Chlorotoluene ................ 1/11 2116 0/5 0/3 0/5 0/1 010 0/17 010 0/1 010 0/4 OlO 010 010 0/4 0/3 10/0 010 010 010 010 4-Chlorotoluene ................ 0/11 0/16 0/5 0/3 0/5 Oj1 010 0/17 010 0/1 010 0/4 2131 010 010 0/4 0/3 10/0 010 010 010 010 2,4-Dichlorotoluene ............. 0/11 0/14 0/4 0/5 0/4 0/1 010 0/9 010 010 010 114 OlO 010 010 010 010 10/9 010 010 6/32 0/5 a,a,2,6-Tetrachlorotoluene ........ 010 010 010 010 010 010 010 0/0 010 010 010 010 OlO 010 010 010 010 10/0 010 010 010 010

Totals ....................... 4/187 81249 1174 0177 1/74 0/17 010 0/193 010 0/5 010 7/69 25/348 010 010 0/20 0/15 0/108 010 ')10 81353 2160 Detection freauencies (%) ....... 2.1 3.2 1.4 0 1.4 0 0 0 10.1 7.2 0 0 0 5.1 3.3

aWater: Soli: Sediment: AIr: Biota: 1) shallow well, 8) soli, 9) sump, 13) living area, 16) oatmeal, 2) deep well, 10) storm sewer, 14) basement, 17) potatoes, 3) sump, 11) sanitary sewer, 15) outdoor; and 18) crayfish, 4) surface, 12) surface water; 19) dog, 5) drinking, 20) maple, 6) storm sewer, 21) mice, 7) sanitary sewer; 22) worms.

SOURCE: US/EPA, op. cit., vol. III.

Table D-3.—Detection (+) of Indicator Substance and Numbers of Samples Analyzed (n) for Environmental Media in the EDA

Environmental media end submediaa

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 18 17 18 19 20 21 22- -+ /n +/n + /n + /n +/n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n +/In + /n + /n + /n + /n + /n + /n

Gamma-BHC (Lindane) . . . . . . . . . . 12/47 3/29 19/105 1/3 3/31 4/7 1/1 7/109 0/0 6/13 0/0 4/4 2/292 0/84 0/79 0/0 0/0 0/31 0/0 0/0 0136 0/8Chlorobenzene . . . . . . . . . . . . . 1/43 2/30 2/104 0/4 7/31 0/9 0/1 3/212 0/0 3/15 0/1 3/4 8/540 1/88 2/83 0/13 0/12 0/0 0/0 0/0 0/0 0/01,2-Dichlorobenzene. . . . . . . . . . . . . 0/47 1/28 0/104 0/4 0/30 1/9 1/1 1/105 0/0 1/15 0/0 2/4 227/539 29/86 8/83 0/0 0/0 0/31 0/0 0/0 0/35 0/191,3-Dichlorobenzene. . . . . . . . . . . . . 0/47 1/28 2/104 0/4 0/30 1/9 1/1 0/105 0/0 1/15 0/0 2/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 1/35 0/191,4-Dichlorobenzene. . . . . . . . . . . . . 0/47 1/28 12/104 0/4 0/30 1/9 0/1 1/105 0/0 1/15 0/0 2/4 81/539 14/86 1/83 0/0 0/0 0/31 0/0 0/0 0/35 0/191,2,3-Trichlorobenzene. . . . . . . . . . . 0/47 0/28 0/104 0/4 0/15 2/9 1/1 0/105 0/0 0/7 0/0 1/4 0/292 0/84 1/79 0/0 0/0 0/31 0/0 0/0 0/35 0/191,2,4-Trichlorobenzene. . .........0/47 0/28 O/1O4 0/4 0/30 3/9 1/1 0/105 0/0 6/15 0/0 2/4 6/292 0/84 1/79 0/0 0/0 0/31 0/0 0/0 0/35 4/191,3,5-Trichlorobenzene. . . . . . . . . . . 0/47 0/28 0/104 0/4 0/15 0/9 1/1 0/105 0/0 0/8 0/0 0/4 0/292 0/84 1/79 0/0 0/0 0/31 0/0 0/0 0/35 0/19l,2,3,4-Tetrachlorobenzene . . . . . . . 0/47 0/28 O/1O4 0/4 0/15 2/9 1/1 0/105 0/0 3/6 0/0 2/4 15/292 3/84 0179 0/0 0/0 0/31 0/0 0/0 8/34 0/191,2,3,5-Tetrachlorobenzene . . . . . . . 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/01,2,4,5-Tetrachlorobenzene . . . . . . . 0/47 0/28 O/1O4 0/4 0/15 2/9 1/1 O/1O5 010 1/6 0/0 0/3 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/34 0/19Pentachlorobenzene . . . . . . . . . . . . 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 2/291 0/84 0/77 0/0 0/0 0/0 0/0 0/0 0/0 0/0Hexachlorobenzene . ............0/47 0/28 1/104 0/4 0/30 0/9 1/1 O/1O4 0/0 1/15 0/0 2/4 0/292 0/84 0/78 0/0 0/0 1/31 0/0 0/0 0/35 0/192-Chloronaphthalein . ............0/47 0/28 0/104 0/4 0/30 0/9 1/1 O/1O5 0/0 0/15 0/0 1/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/35 0/19a-chlorotoluene . . . . . . . . . . . . . . . . 0/43 0/31 0/90 0/0 0/31 0/9 0/1 0/213 0/0 0/6 1/1 0/4 0/0 0/0 0/0 0/13 0/12 0/0 0/0 0/0 0/0 0/02-Chlorotoluene . ...............0/43 1/30 0/90 0/0 0/31 1/8 0/0 0/213 0/0 0/6 0/0 2/4 141/541 14/86 12/83 0/13 0/12 0/0 0/0 0/0 0/0 0/03-Chlorotoluene . . . . . . . . . . . . . . . . 0/43 4/30 0/90 0/0 0/31 1/8 0/0 0/213 0/0 0/6 0/0 2/4 0/0 0/0 0/0 0/13 0/12 0/0 0/0 0/0 0/0 0/04-Chlorotoluene . . . . . . . . . . . . . . . . 0/43 1/30 0/90 0/0 0/31 1/8 0/0 0/213 0/0 0/6 0/0 2/4 69/541 12/86 7/83 0/13 0/12 0/0 0/0 0/0 0/0 0/02,4-Dichlorotoluene . . . . . . . . . . . . . 0/47 1/28 0/104 0/4 0/15 1/9 1/1 O/1O4 0/0 0/9 0/0 1/4 0/0 0/0 0/0 0/0 0/0 5/31 0/0 0/0 0/35 3/19a,a,2,6-Tetrachlorotoluene. . . . . . . . 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0

Totals . . . . . . . . . . . . . . . . . . . . . 13/779 15/448 38/1,673 1/51 10/441 20/148 11/14 12/2,328 0/0 23/167 1/2 28/67 451/4,737 73/1,018 33/965 0/65 0/80 6/372 0/0 0/0 9/419 7/217Detection frequencies (O/.) . . . . . . . 1.7 3.1 2.2 2.0 2.3 13.5 78.6 0.5 0 13.8 50.0 41.8 0.1 7.2 3.4 0 0 1.6 0 0 2.1 3.2aWater soil: Sediment:

1) shallow well, 8) SoiI, 9) sump,2) deep well, 10) storm sewer,3) sump, 11) sanitary sewer,4) surface, 12) surface water;5) drinking,6) storm sewer,7) sanitary sewer;

SOURCE: US/EPA, Op. cit., vol. Ill.

Air. Biota:13) living area, 16) oatmeal,14) basement, 17) potatoes,15) outdoor; and 18) crayfish,

19) dog,20) maple,21) mice,22) worms.

Table D-4.—Detection (+) of Substances Not Found in Control Areas and Numbers ofSamples Analyzed (n) for Environmental Media in the EDA

Environmental media and submedia a

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22Indicator substance + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n + /n

2-Chlorophenol . . . . . . . . . . . . . . . . . . . 0/47 1/27 0/104 0/4 0/30 0/9 0/1 0/104 0/0 0/15 0/1 1/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/35 0/194-Chlorophenol . . . . . . . . . . . . . . . . . . . 1/35 3/24 2/92 0/4 0/15 0/5 0/1 0/71 0/0 0/15 0/1 0/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/35 0/192-Nitrophenol . . . . . . . . . . . . . . . . . . . . . 0/47 0/28 0/104 0/4 0/30 0/9 0/1 0/104 0/0 0/15 0/1 0/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/35 0/192,3,6-Trichlorophenol . . . . . . . . . . . . . . 0/36 0/24 0/92 0/4 0/15 0/5 0/1 0/71 0/0 0/5 0/1 0/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/35 0/194-Chloro-3-methylphenol. . . . . . . . . . . . 0/47 0/28 0/104 0/4 0/30 0/9 0/1 0/104 0/0 0/15 0/1 0/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 1/35 1/191,3,5-Trichlorobenzene . . . . . . . . . . . . . 0/47 0/28 0/104 0/4 0/15 0/9 1/1 0/105 0/0 0/6 1/1 0/4 0/292 0/84 1/79 0/0 0/0 0/31 0/0 0/0 0/35 0/1 9Acenaphthylene. . .................2/47 2/28 1/104 0/4 0/30 0/9 0/1 2/105 0/0 0/15 0/1 0/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/35 0/19Dibenzo (a,h) anthracene . . . . . . . . . . . 0/47 0/28 0/104 0/4 0/15 0/9 0/1 0/105 0/0 0/15 0/1 0/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/35 0/192,4,6-Trichloroaniline. . . . . . . . . . . . . . . 0/47 0/28 0/104 0/4 0/15 0/9 0/1 0/105 0/0 0/6 0/1 0/4 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/34 0/191,2,4,5-Tetrachlorobenzene. . . . . . . . . . 0/47 0/28 0/104 0/4 0/15 2/9 1/1 0/105 0/0 1/6 1/1 0/3 0/0 0/0 0/0 0/0 0/0 0/31 0/0 0/0 0/34 0/19Acrylonitrile . . . . . . . . . . . . . . . . . . . . . . 0/43 0/31 0/104 0/4 0/31 0/9 0/1 0/213 0/0 0/6 0/1 0/4 0/0 0/0 0/0 0/13 0/12 0/0 0/0 0/0 0/0 0/01,1,2-Trichloroethane. . . . . . . . . . . . . . . 0/43 1/31 0/103 0/4 2/31 0/9 0/1 0/213 0/0 0/15 0/1 0/4 0/0 0/0 0/0 0/13 0/12 0/0 0/0 0/0 0/0 0/01,2-Dichloroethane . . . . . . . . . . . . . . . . 1/43 0/31 1/104 014 0/31 0/9 0/1 0/213 0/0 0/15 0/1 0/4 0/0 0/0 0/0 7/13 0/12 0/0 0/0 0/0 0/0 0/0

Totals . . . . . . . . . . . . . . . . . . . . . . . . . 4/576 7/364 4/1,327 0/52 2/303 2/109 2/13 2/1,618 0/0 1/149 2/13 1/51 0/292 0/84 1/79 7/39 0/36 0/310 0/0 0/0 1/348 1/190Detection frequencies (o/o)

(Control = 0°/0) . . . . . . . . . . . . . . . . . 0.7 1.9 0.3 0 0.7 1.8 15.4 0.1 0 0.7 15.4 2.0 0 0 1.3 18.0 0 0 — — 0.3 0.5aWater Soil: Sediment: Air. Biota:

1) shallow well, 8) soil, 9) sump, 13) living area, 16) oatmeal,2) deep well, 10) storm sewer, 14) basement, 17) potatoes,3) sump, 11) sanitary sewer, 15) outdoor; and 18) crayfish,4) surface, 12) surface water; 19) dog,5) drinking, 20) maple,8) storm sewer, 21) mice,7) sanitary sewer; 22) worms.

SOURCE: US/EPA, op. cit., vol. Ill.

58

Love Canal treatment facility in volumes approaching1 percent of the total sludge volume.2

A finding that 4 out of 16 chemicals known to havebeen disposed in the canal landfill calls into questionthe EPA conclusion that EDA is not contaminated byLove Canal chemicals except for sediments of stormsewers and surface water sediments at sewer dischargepoints. The discrepancy between the OTA and EPAfinding can be explained. EPA aggregated data for 150compounds including 129 priority pollutants, the ma-jority of which were not known to have been disposedin the canal landfill. OTA focused on only those chem-icals with a history of disposal.

The detection of four substances at higher frequen-cies in the EDA does not necessarily mean that thesesubstances originated in the canal landfill. Frequenciesof detection for 12 indicator substances were not sig-nificantly different and one had a greater frequencyof detection in the control areas. It is to be expectedthat, at the 90-percent confidence level applied in thesestatistical analyses, 10 percent of the statistical testsshowing significance might be in error. In this case,this means that 1.6 (or really 2) of the 16 analysesmight be a result of chance. However, two or possiblyfour of the significant differences could be real indica-tions of contamination in the EDA as compared to the -

control areas.

Table D-5.—Summary of Mantel-Haenszel Results

EDA V. Love CanalIndicator substance control area v. control areaLindane . . . . . . . . . . . . . . . . . .Chlorobenzene. . . . . . . . . . . .1,2-Dichlorobenzene . . . . . . .1,3-Dichlorobenzene . . . . . . .1,4-Dichlorobenzene . . . . . . .1,2,3-Trichlorobenzene . . . . .1,2,4 -Trichlorobenzene . . . . .1,3,5-Trichlorobenzene . . . . .1,2,3,4-Tetrachlorobenzene. .1,2,3,5-Tetrachlorobenzene. .1,2,4,5-Tetrachlorobenzene. .Pentachlorobenzene . . . . . . .Hexachlorobenzene . . . . . . .2-Chloronaphthalene. . . . . . .0-Chlorotoluene . . . . . . . . . . .2-Chlorotoluene . . . . . . . . . . .3-Chlorotoluene . . . . . . . . . . .4-Chlorotoluene . . . . . . . . . . .2,4-Dichlorotoluene . . . . . . . .a,a2,6-Tetrachlorotoluene . .

NoNoYesYes(a)NoNo—No—NoNoNoNo—

YesNoYesNo—

NoYesYesYes(a)NoYesNoNo—No—

Yes——

YesNoYesNo—. .

Yes: Significantly greater contamination in EDA (Love Canal).No: No significant difference between areas.—: Insufficient data.(a): Control area shows significantly greater contamination.

SOURCE: Cupples, op. cit.

2U. S. Environmental Protection Agency, Analysis of the Love Canal Treat-ment Plant Sludge Sample, memo from W. L. Budde, J. W. Eichelberger,P. Olynk to T. Hauser, Aug. 22, 1980.

o