Baltimore
Integrated Environmental Management Project:
Phase I Report
Regulatory Integration Division
Office of Policy Analysis (PM-220)
Office of Policy, Planning, and Evaluation
U.S. Environmental Protection Agency
May 1987
MD 00355
-------
230R87OO5
BALTIMORE
INTEGRATED ENVIRONMENTAL MANAGEMENT PROJECT
PHASE I REPORT
Prepared for the
Baltimore Management Committee
by
Andrew Manale
Hope Pillsbury
of the
Regulatory Integration Division
Office of Policy Analysis
Office of Policy, Planning and Evaluation
U.S. Environmental Protection Agency
Washington, D.C. 20460
May, 1987
-------
ACKNOWLEDGEMENTS
This report was prepared by staff of the Regulatory Integra-
tion Division of the Office of Policy Analysis of the US Environ-
mental Protection Agency. Credit must go to David Erickson who
started the project for EPA. Also, special credit and thanks
go to John Williams who followed Mr. Erickson as site director
until very recently. He provided considerable energy, creativity,
and technical knowledge to advance this integrated environmental
management study through its first phase. The authors would like
to thank Robert Currie, Daniel Beardsley, Sam Napolitano, and
John Chamberlin for their managerial support at EPA. For their
substantial technical assistance, the authors also thank Roberta
Grossman, David Lee, Cathy Crane, Janice Talsky, Alan Jones.
Special thanks to Dr. Erik Rifkin of Rifkin and Associates for his
help in establishing the fabric of the Baltimore IEMP and to Randy
Freed previously of Sobotka, Inc., Michael Alford and David
Sullivan of VERSAR, Inc., for their help in developing IEMP method-
ologies
While we benefitted from the generous assistance of other
people too numerous to name, we would especially like to thank
the following individuals and organizations for their overall-
direction and development of this study. In this EPA-funded
study, these groups manage the Agency's effort to complete the
project, a unique arrangement for EPA and local government. We
appreciate the strong commitment and effort of:
Members of the Baltimore IEMP Management Committee
Mr. J. James Dieter, Assistant Director, Bureau of Environmental
Service, Baltimore County.
Mr. Claude Vannoy, Assistant to the County Executive for Land
Use, Anne Arundel County.
Mr. Robert Perciasepe, Chief of Capital Improvement Program,
Department of Planning, City of Baltimore.
Dr. Max Eisenberg, Director, Science and Environmental Health,
Science and Health Advisory Group, Office of Environmental
Programs, State of Maryland, Department of Health and Mental
Hygiene
Ms. Virginia Kearney (emeritus, represented the City of Baltimore).
Mr. Daniel Boyd (emeritus, represented Anne Arundel County).
Members of the Baltimore IEMP Technical Advisory Committee
Jared L. Cohon, Vice Provost for Research, and Professor of
Geography and Environmental Engineering, Johns Hopkins
University (Chairman, Technical Advisory Committee)
Frank Hoot, Assistant Commissioner, Environmental Health, Baltimore
City Health Department (Chairman, Human Health Subcommittee).
-------
Don Andrew, Administrator, Engineering & Enforcement Programs,
Office of Environmental Programs'.
Charles Billings, PhD; Associate Professor, Environmental Health
Engineering, School of Hygiene and Public Health, Johns
Hopkins University.
Katherine Parrell, MD, MPH; Chief, Division of Environmental
Disease Control, Office of Environmental Programs.
David Filbert, Director, Division of Air Pollution Control,
Baltimore County Department of Health.
Sam Martin, Consultant, Vice Chairman of TAG, (represented
Regional Planning Council during Phase I).
Philip Clayton, Manager, Cooperative Clean Water Program, Regional
Planning Council.
Paul Keenan, Administrator, Pollution Control Section, Baltimore
City Department of Public Works, City of Baltimore.
Tom Ervin, Environmental Planner, Anne Arundel County Office of
Planning and Zoning (Chairman, Ecological Subcommittee).
Janice Outen, Supervisor of Water Quality, Baltimore County
Department of Health.
Bill Wolinski, Water Quality Coordinator, Department of Public.
Wo rk s.
Emery Cleaves, Principal Geologist, Maryland Geological Survey
(Chairman, Ground-Water Subcommittee)
N. Singh Dhillon, Director, Environmental Health, Anne Arundel
County Department of Health.
Colin Thacker, Director, Northern Environmental Services,
Baltimore County Health Department.
John W. Koontz, represented the Enforcement Program, Waste
Management Program Administration, State of Maryland.
Additional Participants in Phase I Baltimore IEMP
Through their Membership on Workgroups
Joseph Abey
Anne Arundel County Health Department
Anthony S. Bonaccorsi
Eastern Stainless Steel Company
Daryl Braithwaite
Clean Water Action Project
Elkins W. Dahle
Baltimore City Health Department
Allan B. Heaver
Baltimore Building Owners and Managers Association
Michael K. Hettleman
The Southern Galvanizing Company
-------
Bruce W. Jacobs
General Physics Corp.
Thomas King
Mueller Associates
Jack R. Lodge
Baltimore Gas and Electric
Genevieve M. Matanoski
Johns Hopkins School of Hygiene and Public Health
Arthur Nierberding
Mueller Associates
Darryl W. Palmer
FMC Corporation
Velma Rector
American Lung Association of Maryland
John Shumaker
Baltimore County Department of Health
Melissa Wieland
Baltimore Gas and Electric
Susan S. G. Wierman
Office of Environmental Programs, State of Maryland
Bruce Windsor
American Lung Association of Maryland
Joe Macknis
U.S EPA Chesapeake Bay Program
Rich Batuk
U.S. EPA Chesapeake Bay Program
Mary Jo Garreis
Office of Environmental Programs, Maryland
Mary Dolan
Baltimore City
David Carroll
Baltimore City Department of Planning
Stuart May
Dept. of Natural Resources, State of Maryland
-------
Harold M. Cassell
Dept. of Natural Resources, Maryland
Ajax Eastman
State Water Quality Advisory Council
Daryl Tuckey ,
Bethlehem Steel Corporation
Bill Burgess
Department of Natural Resources, Maryland
John Hobner
Baltimore County Health Department
Edwin C. Weber
Department of Natural Resources, Maryland
Neil Thompson
Office of Environmental Programs, Maryland
Tom Kusterer
Office of Environmental Programs, Maryland
Bernard Bigham
Office of Environmental Programs, Maryland
Mark Farfel
Johns Hopkins University
Tad Aburn
Air Management Administration, Maryland
Gregg Mellon
Anne Arundel County
Susan Guyaux
Office of Environmental Programs, Maryland
-------
CONTENTS
Page
EXECUTIVE SUMMARY ES-1
i. 'INTRODUCTION
0 Traditional Approach 1-1
0 Integrated Environmental Management 1-3
0 The Baltimore IEMP 1-6
II. BACKGROUND ON THE BALTIMORE STUDY AREA
Boundaries II-l
Climate II-l
Surface Waters I1-2
Water Supply II-3
Air Quality II-3
Housing . II-3
Population Growth and Development I1-4
Mortality Statistics II-8
Structure of State and Local Government I1-8
III. INSTITUTIONAL STRUCTURE OF THE BALTIMORE IEMP
AND ITS EVOLUTION
° The Institutional Structure 'III-l
0 Evolution of the Institutional Structure III-4
* Public Education and Involvement III-7
IV. OVERVIEW OF THE PRIORITY-SETTING PROCESS
0 General Structure of the Priority-Setting IV-2.
Process
" Priority-Setting in Practice IV-3
Preliminary Defining of the Scope of IV-4
Study
* Selection of Initial List of Phase I IV-5
Study Issues
* Methods for Setting Priorities IV-6
* Ranking of Issues against the Primary IV-ll
Criteria
-------
V. ANALYSIS OF HUMAN HEALTH RISKS
0 Risk Assessment and its Limitations: V-l
An Overview
0 Organics in the Ambient Air V-7
0 Metals in the Ambient Air and Lead in the V-14
General Environment
0 Indoor Air Pollution V-19
* Trihalomethanes in Drinking Water V-23
0 Risk Comparisons V-28
0 Summary of Results V-30
VI. ANALYSIS OF SOURCES WITH POTENTIAL ADVERSE
IMPACTS ON GROUND-WATER
0 Objectives VI-1
0 Overview of the Methodology VI-2
0 Background of Members of Ground-water VI-4
Subcommittee .
0 Data Collection VI-4
e Performing the Scoring of Ground-Water VI-8-
Resource Damage
0 Results of the Rankings VI-10
0 Limitations of the Analysis VI-12
0 Additional Analysis of Metals in VI-15
Ground-water
VII. ANALYSIS OF ECOLOGICAL IMPACT
0 Defining an Approach to the Harbor Issues VII-2
0 Harbor Overview VI1-3
0 Defining an Approach to Studying the Harbor VII-6
0 Monitoring Results VII-7
0 Determining Methodologies for Setting VII-10
Priorities within the Harbor
0 Indexing VII-11
0 Results of Indexing Exercise VII-13
VIII. COMPLETION OF THE PRIORITY-SETTING PROCESS
0 Reducing the Number of Issues for Study VIII-1
0 Reducing the Number of Issues from Ten VIII-2
to Six
0 Applying the Secondary Criteria VIII-3
Developing the Phase II Workplans VIII-5
0 Special Attention given to Metals in the VIII-6
Environment
0 Priority Setting Through Budget Allocation VIII-10
-------
Page
IX. PHASE II WORK PLANS
Air Toxics IX-1
Indoor Air IX-3
Underground Storage Tanks IX-4
Multimedia Metals . IX-5
Baltimore Harbor IX-7
X. CONCLUSIONS
0 Conclusions Regarding Institutional X-2
Arrangements
' Analytic Findings X-2
APPENDIX A GENERAL IEMP METHODOLOGY
APPENDIX B LIST OF ISSUE PAPERS; SAMPLE
ISSUE PAPER
APPENDIX C EPILOG
-------
EXECUTIVE SUMMARY
Baltimore
Integrated Environmental Management Project
Phase I Report
-------
This report describes the first phase of the two-phase
Baltimore Integrated Environmental Management Project (IEMP) con-
ducted by the Regulatory Integration Division of the Environmen-
tal Protection Agency (EPA). EPA initiated the project in
Baltimore as part of its pursuit of new approaches to environmen-
tal management and policy. The purpose of the IEMP is to identi-
fy and assess the significance of a selected set of environmental
issues that concern management, to set priorities for action
among these issues, and to assist local authorities in responding
to environmental problems they have identified.
The IEMP approach is based in part on risk assessment and on
risk management. It uses estimates of risk (that is, the prob-
ability of adverse effects) as the common measure for comparing
problems and setting priorities among issues affecting human
health, involving different pollutants, sources, and exposure
pathways. The risk assessments are used in risk management, a
process in which policymakers balance programs to reduce risks
against available resources to support those programs. In its
simplest form, it requires an examination of how large the risks
are, how much the risks can be reduced by various regulatory
controls, and the cost of controls. Also, the projects are
intended to involve all responsible local parties and agencies in
actually managing and coordinating the projects, ensuring that
issues of greatest local concern are adequately addressed.
Projects typically have two phases. In the first the
decision-making structure of the project is established, key
environmental issues are identified, and priorities for detailed
study are set among them. In the second the IEMP studies the
priority issues in greater depth and develops potential strate-
gies for their control or resolution.
THE BALTIMORE IEMP
Description of the Baltimore IEMP
The Baltimore IEMP is a cooperative effort among EPA and
the governments of the State of Maryland, the City of Baltimore,
Baltimore County* and Anne Arundel County. The Baltimore area
was chosen, in large part, because EPA and local officials wanted
to explore better ways to identify, assess, and manage the human
health risks of environmental pollutants in the area. It repre-
sents the second of five, full-scale geographic projects initi-
ated to date. Most important, the Baltimore IEMP is different
from other IEMP projects in that the authority for project direc-
tion and policies and resource allocation res.ides with state and
local officials"! Though EPA has set the overall study objectives
and the level of resources for the project, it has played only
a support and advisory role in overall project management. This
ES-1
-------
experiment in local participation in, project management of the
Baltimore project is intended to achieve greater local commitment
and to test a different approach to local environmental management.
Emphasis is on local control rather than federally-directed quan-
titative analysis of environmental problems.
The study area (described in Chapter II) covers Baltimore
City, which includes the Port of Baltimore, and Baltimore and
Anne Arundel Counties (see Figure ES-1). The greater Baltimore
area is representative of older, industrialized cities of the
East Coast in transition from smokestack to more diversified,
service economies. Its present environmental concerns derive
largely from the industrial and commercial activities that are
ongoing and of its past and from present-day cars and trucks.
The Institutional Structure of the IEMP
The institutional structure (described in Chapter III) of the
Baltimore IEMP evolved in response to EPA's decision to test the
hypothesis that delegating management authority will lead to
active local participation and commitment to project objectives.
The Management Committee (MC) is the vehicle for State and local
participation and provides project and policy direction. The
Technical Advisory Committee (TAG), composed of local and State
environmental and public health professionals, provides advice to
the MC. EPA provides administrative, technical, and analytical
support. In the concluding stage of Phase I, the general involve-
ment in the IEMP widened to nearly 60 people representing industry,
public interest groups, government, and academia. They serve on
workgroups established to develop and excute workplans for Phase
II issues. For the second phase of the project, the Management
Committee has constituted a Risk Assessment Review Panel, con-
sisting of scientists from Johns Hopkins University, to provide the
MC with scientific and technical advice on questions related to
risk assessment.
The Process for Setting Priorities in Phase I
The purpose of Phase I was to identify issues for further
study in Phase II. The Baltimore IEMP set priorities among the
issues on the basis of available information, supplemented by
data from a brief ambient air monitoring effort conducted by the
EPA. It was not a strictly scientific endeavor, but rather an
exercise in policy analysis using scientific information; expert
judgment; and reasonable assumptions, where gaps in data existed.
Two sets of criteria were used: primary decision criteria which
relied heavily upon scientific data and professional expertise
and secondary criteria which drew on pragmatic considerations
by local officials regarding the best use of study resources.
ES-2
-------
FIGURE ES-1
BALTIMORE I.E.M.P. STUDY AREA
PENNSYLVANIA
-------
The TAG had the responsibility for developing both the ini-
tial list of environmental issues and,the priority-setting proc-
edures that would be used in ranking issues on this list. The
MC had the final say in the adequacy of the initial list and in
the selection and relative funding of issues for Phase II study.
Defining the Scope of the Project
The scope of the project evolved from numerous EPA discus-
sions with State and local governments. Toxic pollutants were
chosen because of the general consensus that the greatest analy-
tic contribution could be made in this area. The same discus-
sions persuaded the Baltimore IEMP to include ecological effects
of both conventional and other than conventional water pollutants
and exclude issues that could not reasonably be handled with
anticipated project resources^ (e.g., conventional air pollu-
tants, occupational exposures, food chain exposures).
Selecting the Initial List of Environmental Issues
The Technical Advisory Committee drew up a list of thirty-
two topics (described in Chapter IV) for preliminary screening,
based on their previous experience with pollution problems and
on professional judgment. The issues dealt with risks to human
health and potential to cause damage to ecosystems and natural
resources, such as ground water. The TAG did not consider the
thirty-two issues a comprehensive list of the most pressing envi-
ronmental problems in Baltimore. This process was neither de-
signed nor ever intended to identify systematically every envi-
ronmental problem; no process could do so given data limitations.
Developing the Method for Setting Priorities in the Initial List
of Issues
The problem for the TAG lay in developing a procedure that
permitted comparing different environmental problems (e.g., sub-
stances that may cause cancers versus industrial effluents that
disrupt ecosystems), yet could make best use of available scienti-
fic information.
Rather than making this type of tradeoff immediately, the
TAG decided to rank the issues against three separate measures
of risk: risk to human health, potential to cause adverse ecolo-
gical impact, and potential adverse impacts to ground-water.
Each type of risk required a different method for determining
the ranking of .these issues with regard to the measure. The
problem of comparing different environmental problems (e.g., human
health vs. ecological impact) was deferred until after the eval-
uation of issues within the categories of human health, ecological
ES-3
-------
impact, and impact on ground-water resources. As it happened,
selection of issues for Phase II study did not require making
these tradeoffs.
In developing measurements of human health risk for air
pollutants, the human health subcommittee used quantitative risk
assessment for carcinogens. For non-cancer health risks, the
subcommittee used EPA Reference Doses (RfDs) that indicate what
levels of a pollutant may pose noncancer risk. For indoor air,
the subcommittee used best professional judgment regarding the
risks because the only existing data were for exposure levels in
other cities. The work of the subcommittee is further described
in Chapter V.
The ground-water subcommittee developed an index (described
in Chapter VI) to rank the relative importance of various sources
and classes of pollutants that may damage ground-water resources.
The subcommittee relied primarily on their professional judgment
to assign scores to different potential threats to ground-water
which in turn were used to establish their relative ranking on the
index. The index had two basic components capturing different
aspects of the possible effect of a potential source on ground
water: pollution and economic impact.
The ecological subcommittee used indexing as the priority-
setting tool for Phase I.The process (described in Chapter VII)
compares existing pollutant concentrations in the ambient water
to generally applicable reference values.
All three of the TAC.'s methods included appraisals of the
degree of uncertainty associated with the analysis of each prob-
lem. These appraisals were qualitative and based on the best
judgment of the committee members.
Ranking the Initial Issues against the Primary Criteria
For setting priorities among potential health'problems, the
human health subcommittee defined the primary criterion as aggre-
gate expected increases in the incidence of disease. On the ba-
sis of this criterion and the quality of the available data, the
subcommittee recommended five issues to the full TAG for study
in Phase II: trihalomethanes in drinking water; toxic volatile
organic compounds of low molecular weight organics; benzene;
metals in air; and indoor air pollution.
The ecological subcommittee recommended three issues to the
full TAG based on the criterion of potential ecological impact:
toxic metals in ambient water; previously contaminated sediments
as a source of water contamination; and bioaccumulation of toxics
in aquatic organisms.
ES-4
-------
Figure ES-2
Flow Chart of How the TAC and the Workgroups
Pared down the List of Issues from Ten to Five
10 Issues
Human Health Risk Subcommittee
Benzene
Tonic Air Pollutio
Trihalomethanes in
Drinking Water
Metals in Air
Indoor Air Pollu
Ecological Impact Subcommittee
Combine and
Redefine to
6 Issues
Air Pollution
Trihalomethanes in
Indoor Air Pollution
Metals-
Sediments '
Bioaccumulation
Toxics in
Aquatic Organises
Metals in the
Environment
Harbor
Ground-Water Resource ^Damage Subcommittee
Metals in Ground'
Water
Underground Storage
Tanks
Underground Storage
Tanks
Apply Secondary
Criteria
Tonic Air Pollution
Indoor Air Pollution
Metals in the
Environment
Harbor
Underground Storage
Tanks
Redo
Primary/Secondary
Criteria for Metals
Resulting in Final
List of Issues
Tonic Air Pollution
Indoor Air Pollution
Lead
Harbor
Underground Storage
Tanks
1\Continued in Phase II for comparison, but not for further study'
-------
After evaluating source categories against the criterion of
potential ground-water impact, the ground-water resource subcom-
mittee recommended two issues to the full TAG: pollution of
ground-water by metals and pollution from underground storage
tanks.
In summary, the first part of the priority-setting process
reduced the initial list of thirty-two issues to ten (see Figure
ES-2). Overlap among the ten study topics allowed the TAG to con-
dense these into six issues. Benzene and toxic air pollution
became one issue, toxic air pollution. The three issues relat-
ing to metals became multi-media metals. The two remaining eco-
logical problems became the focus of a study of the Harbor.. The
remaining 3 topics were indoor air pollution, underground storage
tanks, and trihalomethanes (THMs) in drinking water.
Ranking Issues against the Secondary Criteria
The TAG then completed the priority-setting process through
the application of secondary criteria. These included:
1) Likelihood of making a significant contribution to lo-
cal environmental management through a Phase II study;
2) Lack of duplication with existing analyses or control
programs;
3) Technical and political feasibility if implementing
controls for each issue studied; and
4) Feasibility of performing the analysis within time and
fiscal constraints.
Of the six topics the TAG determined that trihalomethanes in
drinking water did not meet the first criterion. The exposures,
risks, and controls for THMs are well understood for purposes of
risk management and, further, are under study by EPA for national
regulations. Thus THMs were not included as a Phase II study
topic (although it remains as a comparison point for the Phase II
risk'estimates). The other five topics remained for Phase II.
(See Figure ES-2)
ANALYTIC RESULTS
A summary of the analytic results developed in Phase I is
presented in this section. However, it is important to under-
stand the major assumptions on which the results are based. This
is particularly true in this case, as the analysis was used only
to select issues for further studynot to support risk manage-
ment decisions or control strategies nor to document a local
problem.
ES-5
-------
Limitations and Caveats of the Analysis
The reader should "keep in mind that the risk assessments do
not directly examine disease incidence in the local population.
Quantitative risk assessment uses models that are conceptually
simple. Ambient monitoring data or, where these do not exist,
estimated ambient levels based upon source emission estimates,
are used in conjunction with exposure factors to estimate human
exposure to each substance under study. These estimates of expo-
sure in turn are combined with toxicological estimates of potency
to yield quantitative estimates of individual risk. These are
expressed as the incremental probability of disease incidence
(not death) that would conservatively be expected to result from
that exposure. Combined with data on population densities, the
information on individual risk can be extrapolated to yield numer-
ical estimates of disease incidence in the population attributable
to exposures to each pollutant.
The models are deliberately designed to yield conservative
estimates both of individual risk and of aggregate disease inci-
dence to average ambient values for the pollutant in question
over long periods of constant exposure. Their primary usefulness
is thus in setting priorities and allocating resources rather
than in predicting absolute risk. For each of the pollutants
analyzed, the methodology is more likely to overstate risk than
to understate it. However, they do not take into account inter-
mittent peak levels of pollutants to which an individual may be
exposed during his or her daily activities and which, through
dilution in the ambient air, are not well represented by ambient
average daily concentrations. Where these are expected to occur
frequently, the models may be of limited use. The models also
cannot handle possible synergistic or antagonistic effects of
simultaneous exposure to more than one pollutant and rely on
simple additive assumptions, consistent with EPA guidelines.
Consequently, our analyses cannot provide definitive answers
regarding past or current risks. They allow only a rough estima-
tion of health effects or environmental conditions that may occur
in the future for the issues analyzed. Where situations that
contribute to these health effects or environmental conditions
change, these estimates of risk will no longer apply. Because
of the many uncertainties and potential omissions, we cannot say
whether our evaluation of risks to health and the environment are
under- or over-estimated. For those chemicals for which the IEMP
was able to make quantitative estimates of risks and for the ex-
posure scenarios presented, the risks are more likely to be over-
estimated than underestimated. To the extent that toxic chemi-
"cals about' which we currently know little have been left out,
risks may be underestimated.
ES-6
-------
Also, the scope of the study unavoidably excludes numerous
issues that may be of environmental importance. Only a very small
number of pollutants were examined for their health and environ-
mental effects. The IEMP, for example, did not estimate risks
from occupational exposures or from the ingestion of pesticide
contamination of foods. Furthermore, the estimates of pollutant
concentrations in the various media were severely limited by the
lack of actual monitoring data. Nor were risks from chance occur-
rences, such as accidental spills or releases of toxic chemicals,
studied. Also, we did not monitor ground-water resources for con-
tamination or conduct a comprehensive survey of the health of the
harbor. Finally, the examination of hospital records of chil-
dren's exposure to lead in dust suggests only that an immediate
health concern exists for a limited population. The health con-
cern cannot be generalized to the population as a whole.
Risk assessment in Phase I was conducted as part of an exer-
cise in policy analysis to help local decision-makers set priori-
ties. The goal of our risk assessments is to determine which
issues were suitable for a more detailed examination of risks and
control options in Phase II. The results which we present are
not statements about the incidence of disease in the Baltimore
area.
The results of Phase I of the IEMP must, therefore, not be
looked upon as the products of a comprehensive appraisal of envi-
ronmental risks in the Baltimore area. The value of the IEMP
methodology is that it allows an evaluation and comparison of the
risks from chemicals about which we know something. Management
of these risks, based on the best current information, can pro-
ceed, while research continues on the effects of chemicals about
which little is currently known*
Results of the Analyses
Before summarizing the results of our analysis of human
health effects, we want to provide some general guidance to help
understand them. We feel it is important to provide a point of
reference, a baseline, for our numerical estimates of the health
effects. Thus, from Maryland statistics for cancer, we know there
were 4285 cancer deaths in 1984 in the study area (see Chapter II,
Table II-4). In 1983, the American Cancer Society estimated there
were 1.92 cancer cases for every cancer death. Using these two
factors, we derived an estimated baseline incidence of 8227 cases
for the study area.
Human health issue; Organics in the Ambient Air
The limited air monitoring conducted by the IEMP and r.eview
of monitoring conducted by the Maryland Air Management Administra-
tion provided estimates of ambient levels for a small number of
noncriteria pollutants. The IEMP also conducted limited air dis-
perison modeling to estimate the contribution of emissions from
publicly-owned treatment works (POTWs) to ambient levels.
ES-7
-------
Of the organic compounds evaluated in the ambient air, the
highest measured concentrations were for benzene, xylene, toluene,
and ethyl benzene. Lead had the highest value for the metals.
Comparison of levels of pollutants at different monitoring sites
suggests that the concentrations of some organics and metals can
vary significantly throughout the Baltimore area.
The highest identifiable cancer risk from organics is attri-
butable to benzene. Benzene alone accounted for about two-thirds
of the upper-bound estimate (assuming the EPA standard of 70 years
of exposure) of roughly 3 excess cancer cases a year from the
organic air toxics examined. Table ES-1 summarizes these results.^-
Monitoring data suggested that there may also be increased
risk of noncancer health effects for benzene and chloroform.
Human health issue; Metals in the Ambient Air
For the metals examined, health risks were much lower than
the organics, with the possible exception of chromium. The
health risk from exposure to chromium depends on the species of
chromium in the air. The techniques employed in determining
chromium levels in the Baltimore area, however, do not allow for
determining the relative concentrations of hexavalent chromium, a
potent carcinogen, to trivalent chromium, a relatively weak car-
cinogen. Under the worst case, and unlikely, assumption that all
detected chromium is hexavalent, the upper-bound estimate (assuming
the EPA standard of 70 years of exposure) of cancer risk is roughly
four excess cancer cases a year. This roughly comparable to the
total risk from the organics described above. Table ES-2 summari-
zes these results.
The upper-bound estimate (assuming the EPA standard of 70
years of exposure) of cancer risks to individuals in areas with
the highest concentrations of the pollutants examined (both for
metals and for organics) did not exceed one chance in ten thousand
for any pollutant, except where one assumes that all detected
chromium was hexavalent. In that case, the upper-bound estimate
(assuming the EPA standard of 70 years of exposure) risk would be
four chances out of ten thousand.
1 Potentencyestimates used were from EPA's Cancer Assessment
Group (CAG). They have received extensive peer review. The
exception is 1,2-Dichlbropropane. For this compound, toxicolo-
gical staff of the Regulatory Integration developed the potency
score using CAG methodology. Review of this score has been
more limited.
ES-8
-------
Table ES-1
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES Of ANNUAL EXCESS CANCER INCIDENCE:
ORGANICS IN THE AMBIENT AIR
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
Upper-Bound Annual Cases1,2
Pollutant 1985 Revised
(weight of evidence)3 Analysis4 19865
Benzene (A) 1.6 1.8
Trichloroethylene (B2) 0.1 0.02
Perchloroethylene (82) 0.2 0.1
1,2-Dichloroethane (82) 0.04 0.1
Chloroform (82) 0.2 0.4
Carbon Tetrachloride (82) 0.3 0.3
1,2-Oichloropropane (C) 0.1 0.1
Total 2.5 2.8
:THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON
CONSERVATIVE ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND
ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS IN
SEVERAL AREAS OF THE ANALYSIS, SUCH AS EXPOSURE CALCULA-
TIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE CALCU-
LATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF
ACTUAL CANCER RISKS IN BALTIMORE. ACTUAL RISKS MAY BE SIG-
NIFICANTLY LOWER; IN FACT, THEY COULD BE ZERO. THE PROPER
FUNCTION OF THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT
AND EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS
EXAMINED.
2RID'S ESTIMATE OF THE ACTUAL NUMBER OF CANCER CASES IN THE
STUDY AREA IN 1984 IS 8,000 CASES. (SEE II-8 AND 9.) THIS
NUMBER SHOULD SERVE ONLY AS A POINT OF REFERENCE IN UNDER-
STANDING THE RISK ESTIMATES PROVIDED. IN ADDITION, THE -
RISK ESTIMATES SHOULD NOT BE INTERPRETED AS REPRESENTING
THE TOTAL UPPER-BOUND CANCER RISKS FROM ALL POLLUTANTS IN
ANY PARTICULAR MEDIUM. THEY DO NOT TAKE INTO ACCOUNT ALL
POLLUTANTS THAT MAY BE PRESENT IN THE MEDIUM, ALL SOURCES
OF THESE POLLUTANTS, AND ALL EXPOSURE SCENARIOS INVOLVING
PATHWAYS OR EXPOSURES OF SHORT DURATION TO RELATIVELY HIGH
DOSES.
3EPA weight-of-evidence classifications: A = human carcino-
gen; B2 = probable carcinogen; C = possible carcinogen.
(See Appendix A for more detail.)
*The incidence estimates listed in this column were calcu-
lated using cancer- unit risk values developed in 1985.
^The incidence estimates listed in this column were calcu-
. lated using current (5/86) cancer unit risk values.
-------
Table ES-2
BALTIMORE IEMP PRELIMINARY
RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF ANNUAL
EXCESS CANCER INCIDENCE:
METALS IN THE AMBIENT AIR
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
Upper-Bound Annual Cases*,
Pollutant 1985 Revised
(weight of evidence)3 Analysis^ 19865
Chromium (A)
Total hexavalent 4.2 4.2
50% hexavalent 2.1 2.1
105 hexavalent 0.4 0.4
1% hexavalent 0.04 0.04
OK hexavalent 0.00 0.00
Cadmium6 (Bl) 0.00 to 0.05 0.00 to 0.04
Total 0.00 to 4.25 0.00 to 4.24
*THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED
ON CONSERVATIVE ASSUMPTIONS THAT GENERALLY PRODUCE
UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA
AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK
ESTIMATES WERE CALCULATED AS AIDS TO POLICY DEVELOP-
MENT, NOT AS PREDICTIONS OF ACTUAL CANCER RISKS IN
BALTIMORE. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER;
IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF
THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND
EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS
EXAMINED.
2RID'S ESTIMATE OF THE ACTUAL NUMBER OF CANCER CASES
IN THE STUDY AREA IN 1984 IS 8,000 CASES. (SEE II-8
AND 9.) THIS NUMBER SHOULD SERVE ONLY AS A POINT OF
REFERENCE IN UNDERSTANDING THE RISK ESTIMATES
PROVIDED. IN ADDITION, THE RISK ESTIMATES SHOULD
NOT BE INTERPRETED AS REPRESENTING THE TOTAL UPPER-
BOUND CANCER RISKS FROM ALL POLLUTANTS IN ANY PARTICU-
LAR MEDIUM. THEY DO NOT TAKE INTO ACCOUNT ALL POLLU-
TANTS THAT MAY BE PRESENT IN THE MEDIUM, ALL SOURCES
OF THESE POLLUTANTS, AND ALL EXPOSURE SCENARIOS
INVOLVING PATHWAYS OR EXPOSURES OF SHORT DURATION TO
RELATIVELY HIGH DOSES.
'EPA weight-of-evidence clasaificationa: A = human
carcinogen; 81 = probable carcinogen. (See Appendix A
for more detail.)
*The incidence estimates listed in this column Mere
calculated using cancer unit risk factors available in
1985.
'The incidence estimates listed in this column were
calculated using current (5/86) unit risk factors.
^Measured cadmium concentrations were below detection
limits. For screening purposes only, we calculated
risks to human health assuming a range in ambient
concentration from zero to the detection.limit.
-------
Human health issue; Trihalomethanes in drinking water
The subcommittee examined the health risks from ingestion of
trihalomethanes (particularly chloroform) in the water from public
drinking water treatment plants.
The upper-bound estimate (assuming the EPA standard of 70
years of exposure) of cancer risk to individuals is 5 chances in
100,00 for either plant (using the 1985 unit risk factor for
chloroform). The chloroform levels correspond to an upper-bound
estimate (assuming the EPA standard of 70 years of exposure) of
less than one annual excess cancer case from ingest-ion for each
plant (using 1985 cancer potency data). No other monitored
pollutants of carcinogenic concern were detected. Table ES-3
summarizes these results.
None of the pollutants examined in drinking water appear to
pose noncancer health effects at the concentrations found, with
the possible exception of lead. As mentioned earlier with re-
gard to lead in the ambient air, total exposure from all path-
ways is essential in estimating the health risks posed by this
substance. Though lead in the public water supply is present at
very low concentrations and meets the current standard for drink*-
ing water quality, it can, nevertheless, leach out of plumbing
from residences and the distribution system and hence be present
at higher levels in tap water. The subcommittee could not esti-
mate the actual contribution of lead in drinking water at the tap
to health risks from lead.
Comparison of risks in ambient air and drinking water
Using the above analyses of the upper-bound estimates (assum-
ing the EPA standard of 70 years of exposure) of the annual
excess cancer incidence, we conclude that the magnitude of the
risks from the examined organics and metals in the ambient air
and from THMs in drinking water is roughly comparablean upper-
bound estimate of 3 excess cancer cases per year.. If all airborne
chromium were hexavalent, an unlikely assumption, the upper-bound
estimate for the ambient air risk would be roughly twice that of
THMs7 excess cases per year versus 3 excess cases per year.
Table ES-4 summarizes these results.
Human health issue; Lead in the General Environment
Exposures to lead in the amb'ent air could also contribute
to the total intake of the metal from all pathways. Low levels of
lead exposure can lead to hypertension in adult males and blood-
related problems and neurological dysfunctions in children and
the unborn. However, the exposure to the levels of lead in the
ambient air of Baltimore alone are unlikely to lead to these
health problems.
ES-9
-------
Table ES-3
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF ANNUAL EXCESS CANCER INCIDENCE:
POLLUTANTS IN BALTIMORE DRINKING WATER1,2
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
Ashburton3 Montebello3
Upper-Bound Annual Upper-Bound Annual
Cancer Cases Cancer Cases
Pollutant Average Average
(weight of Concentration 1985 Revised Concentration 1985 Revised
evidence)4 (ug/1)5 Analysis6 19867 (ug/1)5 Analysis6 19867
Chloroform8 (B2) 54.3 0.7 1.6 49.3 0.5 1.1
Total 0.7 1.6 0.5 1.1
XTHE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE ASSUMPTIONS THAT GENERALLY PRODUCE
UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE CALCULATED AS AIDS TO POLICY DEVELOPMENT,
NOT AS PREDICTIONS OF ACTUAL CANCER RISKS IN BALTIMORE. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER; IN FACT,
THEY COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE
ISSUES AND SET PRIORITIES FOR THE TOPICS EXAMINED.
2RID'S ESTIMATE OF THE ACTUAL NUMBER OF CANCER CASES IN THE STUDY AREA IN 1984 IS 8,000 CASES. (SEE II-8 AND
9.) THIS NUMBER SHOULD SERVE ONLY AS A POINT OF REFERENCE IN UNDERSTANDING THE RISK ESTIMATES PROVIDED. IN
ADDITION, THE RISK ESTIMATES SHOULD NOT BE INTERPRETED AS REPRESENTING THE TOTAL UPPER-BOUND CANCER RISKS FRC
ALL POLLUTANTS IN ANY PARTICULAR MEDIUM. THEY DO NOT TAKE INTO ACCOUNT ALL POLLUTANTS THAT MAY BE PRESENT IN
THE MEDIUM, ALL SOURCES OF THESE POLLUTANTS, AND ALL EXPOSURE SCENARIOS INVOLVING PATHWAYS OR EXPOSURES OF
SHORT DURATION TO RELATIVELY HIGH DOSES.
'Ashburton serves a population of 900,000; Montebello serves a population of 700,000.
*£PA weight-of-evidence classification: B2 a probable carcinogen. (See Appendix A for more detail.)
^Measure of pollutant concentration in finished water over 1981-1983, City of Baltimore Drinking Water Quality
Data (Versar, 1984).
6Chloroform risk calculation based on the unit risk factor available in 1985.
7Chloroform risk calculation based on the most current unit risk factor (5/7/86).
"Analysis assumes total trihalotnethane concentration is chloroform.
-------
Table ES-4
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS1,2
COMPARISON OF UPPER-BOUND EXCESS ANNUAL CANCER
INCIDENCE ACROSS ISSUES AND POLLUTANTS IN BALTIMORE
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
(1986 analysis)
Drinking
Compound (weight of evidence)3 Air Water
Volatile Organic
Compounds
Benzene (A) 1.80
Trichloroethylene (B2) 0.02
Perchloroethylene (B2) 0.10
l,2-Oichloroethane(B2) 0.10
Chloroform (82) 0.40 2.7
Carbon Tetrachloride (82) 0.30
1,2-Dichloropropane (C) 0.10
2.82 2.7
Chromium (hexavalent)4 (A) 0.00 to 4.2
Cadmium5 (Bl) 0.00 to 0.04
Subtotal 0.00 to 4.24
TOTAL6 2.8 to 7.1 2.7
UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON
CONSERVATIVE ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND
ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS IN
SEVERAL AREAS OF THE ANALYSIS, SUCH AS EXPOSURE CALCULATIONS
AND POLLUTANT SELECTION, RISK ESTIMATES WERE CALCULATED AS
AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL
CANCER RISKS IN BALTIMORE. ACTUAL RISKS MAY BE SIGNIFICANTLY
LOWER; IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF
THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE
ISSUES AND SET PRIORITIES FOR THE TOPICS EXAMINED.
2RID'S ESTIMATE OF THE ACTUAL NUMBER OF CANCER CASES IN THE
STUDY AREA IN 1984 IS 8,000 CASES. (SEE II-8 AND 9.) THIS
NUMBER SHOULD SERVE ONLY AS A POINT OF REFERENCE IN UNDER-
STANDING THE RISK ESTIMATES PROVIDED. IN ADDITION, THE RISK
ESTIMATES SHOULD NOT BE INTERPRETED AS REPRESENTING THE TOTAL
UPPER-BOUND CANCER RISKS FROM ALL POLLUTANTS IN ANY PARTIC-
ULAR MEDIUM. THEY DO NOT TAKE INTO ACCOUNT ALL POLLUTANTS
THAT MAY BE PRESENT IN THE MEDIUM, ALL SOURCES OF THESE
POLLUTANTS, AND ALL EXPOSURE SCENARIOS INVOLVING PATHWAYS OR
EXPOSURES OF SHORT DURATION TO RELATIVELY HIGH DOSES.
'EPA weight-of-evidence classifications: A = human carcino-
gen; Bl, 82 = probable carcinogen; C = possible carcinogen.
(See Appendix A for more detail.)
^Chromium incidence calculations indicate a range of possible
ambient levels of hexavalent chromium from 0 percent to
100 percent.
^Cadmium incidence calculations indicate a range of possible
ambient levels from 0.0 ug/ra3 to detection limits (between
.001 and .002 ug/m3).
^Numbers have been rounded to one significant decimal.
-------
Also, lead in the first few minutes of flow of tap water can
contribute to the Baltimore area residents' overall intake of
lead; it too may put the unborn child and young children at
greater risk of adverse neurological effects and adult males at
greater risk of hypertension. Finally, ingestion of lead in
household dust is clearly a very significant threat to the health
of young children in older homes which have been painted with
lead-based paint.
Human health issue; Indoor air pollution
The subcommittee reviewed existing studies on indoor pollu-
tion. The studies, which were not specific to the Baltimore area,
suggested that indoor levels of many pollutants generally associ-
ated with the outdoors may be high enough to warrant public
health concern. In addition, the indoor environment has its own
pollutants of unique concern, such as radon and tobacco smoke.
Few data were found on actual exposure levels to indoor pollu-
tants in the Baltimore area.
Ground-water resource issue
The ranking system for determining the relative relationship
of potential threats to ground-water resources was designed to
take advantage of the expertise and professional judgment of sub-
committee members. This was necessary because of the lack of
data needed for modelling purposes or for validating models that
are or could be developed.
The ranking system takes into account both pollution impact
and economic impact. Pollution impact took into account such fac-
tors as number of sources, release volume, the present and future
rate of contamination incidents, and the potential extent of
damage. Economic impact included assessment of the relative mag-
nitude of costs to prevent or reduce contamination, and the cost
of response to contamination, again using best professional judg-
ment. Table ES-5 shows the results for the top eight source
types examined by the ground-water subcommittee.
The two issues that consistently ranked highest as potential
threats were underground storage tanks and multimedia metals
(toxic metals in the various media). They represent the sources
that are relatively the most important potential threats to
ground-water resources. The methodology does not allow us to
conclude they are problems. Chapter VI describes the analysis in
detail.
Ecological Issue; The Harbor
The harbor subcommittee examined a number of different
methods for assessing the relative significance of different
pollutants. For Phase I, they used a method, eco-scoring, which
compares ambient levels of toxics in the harbor and its tributa-
ES-10
-------
Table ES-5
Baltimore IEMP
Results of Relative Ranking of Sources with Potential
Mverse Impact on Ground-water Resources1'2
Relative Ranking
Based on Equal Relative Ranking # of Times
Weighting of Based on Pollution Workgroup
Pollution Impact and Impact Weighted as Members Scored
Economic Impact-* Twice Economic^ Top
Underground storage Underground storage
tanks tanks 4
Multimedia metals Multimedia metals 3
Benzene Benzene 3
Pesticides/herbicides Pesticides/herbicides 2
Pollution from farming Pollution from fanning 1
Landfills Landfills 1
Septic Tanks Septic Tanks 1
Chromium in Harbor Chromium in Harbor 1
ifiased on a system developed by the ground-water workgroup, Which ranks
sources for potential for damage to ground-water resources. These rankings
are for the purpose of setting priorities for further study; they do not
apply to specific sites within the study area, but rather, provide results
of the workgroups' deliberations regarding the relative ranking of potential
threats to the ground-water resource.
^Chly the top eight sources are shown; the other five sources can be found in
Chapter VI.
^The first scoring system used by the ground-water subcommittee weighted pol-
lution impact and economic impact equally.
^The ground-water subcommittee changed the scoring system slightly, weighting
pollution impact twice as heavily as economic impact, to determine how sensi-
tive the scoring system was to variation.
^This counts the number of times the source scored in a ground-water workgroup
member's top five sources. As there were 4 members in the workgroup/ the top
score was 4.
-------
ries to EPA's Water Quality Criteria.
For metals in the ambient water, mercury, lead, nickel, and
copper scored high, though lead was of concern only in an older
set of data. Mercury was the metal of greatest concern in the
tributaries. In the sediments, chromium had a higher score than
the other metals. In comparing the metals in the Harbor with
those in the Bay, all metals were higher in the Harbor than the
Bay with the exception of copper. These metals represent the
relatively most important potential threats to the Harbor. The
methodology does not allow us to conclude they are problems. The
indexing results are summarized in Table ES-6. Chapter VII de-
scribes the analysis in detail.
CONCLUSIONS ABOUT THE PROJECT TO DATE
The EPA, Maryland State government, and local government
officials have established the organizational framework at the
State and local level for setting priorities for government
action on environmental issues in the study area. The Management
Committee MC, with the assistance of the Technical Advisory Com-
mittee (TAG) effectively identified and set priorities among a
wide-ranging and diverse set of environmental issues.
The Baltimore IEMP has helped State and local governments
develop a working understanding of methods for analyzing issues.
Priority-setting in the Baltimore IEMP was a hands-on process.
The TAG played an active role through its provision of expert
judgment while it used analytical tools to identify important
environmental issues and compare and rank them against evaluative
criteria. The success of representatives of State and local
jurisdictions in reaching consensus on questions of environmental
priorities that unevenly affect them testifies to the usefulness
of these tools and these governments' ability to use them. Also,
EPA held workshops for both government officials and the general
public to familiarize them with the use of risk assessment.
The Baltimore IEMP has helped State and local governments
address a high-priority problem. The State and counties drasti-
cally reduced the standards for the amount of lead used in solder
and flux in the plumbing of residential drinking water systems
after work during Phase I identified this as a potentially seri-
ous health problem. However, most of the tangible progress to-
wards solving other environmental problems is generally not ex-
pected until completion of the second phase of the project.
ES-11
-------
TABLE ES-6
Baltimore IEMP
Comparison of Index Values For Harbor Priority-Setting,
Using Different Types of Indexing Techniques 1
Ambient Water Quality
Harbor
Ambient Water
Quali ty-Tr ibu taries
Ambient Sediment Ambient Sediment
Quality-Baltimore Quali ty-Chesapeake
Harbor Bay
Trident
' Data1 .
Zinc °
Nickel +
Mercury O
Lead +
Copper O
Chromium °
Cadmium °
I HMD
Data2
+
+ N/A
0 O
0
+ 0
+
N/A
+
+
N/A
0
0
O
N/A
N/A
+
N/A
N/A
N/A
+
N/A
O Index value greater than 2.
Criteria)
+ Index value greater than 1.
* Index values less than 1.
N/A Not Available.
(Ambient values are more than twice the level of EPA
(Ambient values are greater than EPA criteria)
(Ambient values are less than EPA criteria) '
1. See Figures VII-6 to VII-10 for a presentation of the indexing scores and the
sources of data used to generate this table.
-------
The Baltimore IEMP has provided information that will help
EPA conduct its programs. EPA has lacked analytical methods and
procedures for setting priorities among issues that do not direct-
ly relate to health. In Phase I of the Baltimore IEMP, we have
made progress in developing priority-setting tools for ecologi-
cal issues relating to the aquatic environment and a procedure
for achieving consensus on issues of importance to ground-water
resources.
PHASE II STUDIES
We present below a brief summary of the five issues selected
for Phase II study. The order of presentation does not reflect
their relative importance. Importantly, each study area is tai-
lored to Baltimore's needs and with the recognition that local,
state and EPA regional staff are also working in these areas.
In effect, our work fits into the overall area environmental
agenda to maximize what all levels of the governments are learn-
ing about issues in Baltimore.
Air Toxics
Air toxics were found to be a potentially significant though
undefined threat to public health in the Baltimore area. The
State air toxics program is being designed to address area-wide
problems from industrial emissions of air toxics. To complement
this effort, the IEMP air toxics study is designed to address the
so-called "urban soup" where toxic emissions from both point and
area sources combine to form elevated concentrations of pollutants
in localized areas. The goals are to estimate human health risk
from selected air toxic emissions from both industrial and area
sources and to analyze control strategies to reduce the adverse
health effects.
An important part of this effort is the Baltimore Total
Exposure Assessment Methodology (TEAM) study to be conducted in
conjunction with EPA's Office of Research and Development. This
study will help provide information on the relative risks of in-
door versus outdoor pollution. The objectives of the study are:
- to apply "modified" TEAM methodology to Baltimore to
estimated exposure of Baltimore area residents of specific
geographic areas to selected volatile organic compounds;
- to compare modeled concentrations with measured ambient
levels for selected volatile organic compounds.
- to compare indoor concentrations, outdoor concentrations,
and personal exposures.
ES-12
-------
Multimedia Metals
The MC and the TAG found metals from a wide variety of
sources and in all media to pose potentially significant health
and environmental risks. The Phase II will focus our limited
resources on one major issue and allow the IEMP the best chance
of assisting local officials in managing it.
The major task is to develop cost-effective techniques for
removal and abatement of lead paint and dust. An extensive num-
ber of studies dealing with various aspects of exposure to lead
indicate that current techniques for abating lead paint and dust
are not very effective and, in some cases, actually increase the
levels of lead dust in housing.
Indoor Air Pollution
While data from other cities suggest that indoor pollution
may pose significant risk to human health, there is little local
data on whether indoor pollution is also a problem in the Balti-
more area.
The goals of the workplan are to learn more about indoor air
quality in Baltimore; to investigate possible programs to reduce
exposure from indoor air pollution; and to recommend their imple-
mentation where appropriate.
Underground Storage Tanks
The TAG and the MC chose leaking underground storage tanks
(USTs) as an issue to be studied in Phase II because USTs were
highly ranked relative to other potential sources by the ranking
system used to assess potential damage to ground-water resources.
Because Maryland already has regulated underground storage tanks,
the UST workgroup members perceived a unique opportunity to
develop a study which would help the state and local government.
The analysis focuses on developing an approach to help establish
priorities for inspection and enforcement activities, given the
governments' limited resources.
Baltimore Harbor
The TAG and the MC chose Baltimore harbor as an issue for
Phase II because of the importance that pollution of the harbor
can have on the current and future uses of this vital resource.
Our objectives are to define the possible future uses of the har-
bor and to identify additional research and institutional arrange-
ments that should occur to help environmental decision-makers
understand how to achieve any set of goals they have to ensure
that those uses can occur. The work group will also explore
methods to assess the effects of pollutants on aquatic life.
ES-13
-------
CHAPTER I
INTRODUCTION
-------
I. INTRODUCTION
This report describes the first phase of the two-phase
Baltimore Integrated Environmental Management Project (IEMP) ini-
tiated and sponsored by the Regulatory Integration Division of
the Environmental Protection Agency (EPA). The purpose of the
IEMP is to identify and assess the significance of environmental
issues of local concern, to set priorities for action among these
issues, and to analyze appropriate approaches to manage the prob-
lems we assess.
EPA initiated the project in Baltimore as cart of its pur-
suit of new approaches to environmental management and policy.
The project is EPA's second of five larae-scale, multi-media stu-
dies on the effect of toxic pollutants on human health and the
environment in a single geographic area. Like the other IEM pro-
jects, the Baltimore IEMP differs from earlier EPA efforts to
control toxics because it uses an integrated approach to measure
the complex, interactive effects of pollutants in the environment
and to identify cost-effective methods of control of these air,
water, and' soil pollutants. The Baltimore IEMP also reflects a
shift in public interest by focusing on the effects of toxic
chemicals, rather than "conventional" pollutants. Unlike the
other TEMPs, the management structure of the Baltimore project
is unique; State and local authorities make the key management
decisions and have responsibility and authority for project direc-
tion and policy. EPA provides the financial resource, and EPA
staff provide technical and administrative support.
After briefly comparing the IEM approach with traditional
approaches, we explain in this introduction the rationale for the
integrated approach, the objectives of the Baltimore- IEMP, and
the contents of this report.
TRADITIONAL APPROACH
The United States has' traditionally responded to its
environmental problems by passing new laws that focus on aNsingle
environmental medium, such as the Clean Air Act, Clean WateNr Act,
or Safe Drinking Water Act. Accordingly, each major office at
EPA handles the problems associated with a specific medium, such
as air or water pollution and hazardous waste disposal practices.
Most state and local environmental agencies mirror this medium-
specific organizational structure.
EPA has also traditionally focused on one pollutant or class
of pollutants within each medium at a time. When applied to
toxics, this pollutant-by-pollutant, single-medium aoproach has
sometimes led to environmental programs and regulations characte-
rized by the inefficient use of resources and ineffective manage-
1-1
-------
ment. In particular, there has been-concern that the traditional
approach could contribute to a number of unfortunate consequences:
(1) The solution to one' single-pollutant, single-medium
problem might simply transfer the problem to another
medium (e.g., from water to air), perhaps resulting in
greater risks and costs of control.
(2) Problems, such as hazardous waste, involving several
environmental media may not be addressed sufficiently
by an environmental agency that generally examines
each medium separately.
(3) Policy-makers have no way of setting priorities across
pollutants, sources of pollutants, and pathways of ex-
posure in different media. Consequently, environmental
policies and regulations may not be the most efficient:
too many resources may be spent on some problems and
too little on others.
(4) Narrowly focused studies of pollutants within a parti-
cular medium may not consider total or cumulative
environmental exposure, either within the medium or
across all media.
(5) Laws and regulations developed by separate EPA divi-
sions or on the basis of separate analyses may involve
different and sometimes inconsistent objectives, meth-
ods, and standards.
(6) The single-medium approach frustrates attempts to
understand fully the relationships among pollutants,
media, and control alternatives.
Industrial wastewater treatment, for example, can shift
pollutants from water to land or air. Though the process pro-
duces a relatively clean release to a nearby body of water, it
also produces a residual sludge containing many of the original
pollutants in concentrated form to be disposed of on land. If
not properly managed, these more highly concentrated pollutants
can leach from land disposal sites into drinking-water supplies,
where they may represent greater environmental risks than in
their original context. Alternatively, the more volatile com-
pounds within the sludge can evaporate into the air to create
an air pollution threat. Similarly, air pollution control de-
vices, such as scrubbers, generate sludges that require land
disposal which can also pose environmental hazards.
EPA's traditional approach has also been directed toward
establishing policies and standards for the nation. Though gene-
rally a reasonable and effective strategy, addressing pollution
on a national basis can lead to situations where particular
1-2
-------
areas receive inappropriate levels of control relative to their
needs. Of special concern are issues that may be overlooked or
placed low on EPA's agenda because they have regional, but little
national, significance.
The risks from toxic substances pose a more complex and
difficult problem for government regulation than the previously
addressed conventional pollutants, such as the oxides of sulfur
in air or fecal coliform in water. Even minute quantities of
some toxic substances can threaten public health and biological
systems. And sophisticated monitoring devices can detect these
chemicals at levels at which science cannot, as yet, provide
definitive answers regarding their effects on health or ecosys-
tems. Toxic substances can move through the environment in com-
plex patterns, even migrating from one medium to another (by eva-
porating to the air from water or land and leaching from land to
ground or surface waters). Our ability to monitor these move-
ments from the initial source of contamination to the eventual
public exposure or ecological target is limited. We know little
about migration pathways, chemical transformations, and meteoro-
logical and other factors that affect these; pathways and trans-
formations. Finally, even should we identify pollutants of con-
cern and their sources and pathways of exposures, we may still
lack techniques for their abatement and control. The consequence
is that we must develop new strategies for addressing the prob-
lems created by the introduction of toxic substances into the
environment.
j,
INTEGRATED ENVIRONMENTAL MANAGEMENT
EPA adopted the concept of integrated environmental
management as a response to the growing challenge of toxic pol-
lution and as a potential solution to the shortcomings of the
traditional approaches for pollution control. The term "inte-
grated environmental management" covers both the evaluation of
public health and environmental risks from various pollutants
and the development of the appropriate control strategies from
a multi-media perspective. The approach is designed:
o to account for multi-media phenomenathat is, to identify
and to track situations where pollution is transferred
from one medium to another;
o to focus attention on the cost-effective allocation of
resources to program areas where we see. the largest prob-
lems; and
o to address other issues that once lacked an appropriate
forum. An integrated approach to environmental management
also calls for a focus on a specific geographic location.
1-3
-------
The geographic basis allows the IEMP to consider all major path-
ways and sources, as well as local meteorology, hydrology, indus-
try, population patterns, and social and political factors in
designing government regulatory strategies.
The integrated approach incorporates a "systems" view of
environmental managementa recognition that media are linked and
that natural processes and abatement strategies are interrelated.
The systems approach can have a profound effect on environmental
management in a number of ways: engineers and planners can craft
better strategies by respecting and exploiting the linkages, and
decision-makers can appreciate better the implications of their
decisions.
The IEM projects are intended to involve all local respons-
ible parties and agencies in managing and coordinating the pro-
jects. Broad local involvement ensures that issues of greatest
local concern are adequately addressedregardless of their per-
ceived national importance. And, by pooling informational and
analytical resources among all levels of government (federal,
state, regional, and local), responsible agencies under the au-
spices of EPA are able to undertake efforts that, because of
limited data, manpower, and funds, none would be able to under-
take alone.
Bringing all responsible agencies and parties together at
the same table has additional benefits. For example, environmen-
tal problems can persist because the entity that discovers the
problem and has the best information regarding its nature is
different from the entity that has both the authority and the
resources for solving it. The control agency may not respond
because its management priorities are inconsistent with or do not
take into account the issue at hand. The remedy may necessitate
both adopting a method for making priorities more consistent and
establishing the organizational link between the two agencies.
The IEM strategy of bringing the key agents together at the out-
set of the project is the first step in the remedy.
The shift in orientation from largely uniform national
policies to a policy that is peculiar to a region or locality, as
embodied in an IEMP, also signals a new emphasis with regard to
assessing and managing risks. These concepts are discussed in
Appendix A. But here we should note a theme recurrent throughout
the project and therefore throughout this report. In the first
phase of the Baltimore IEMP, we concentrated on assessing risks
to the environment and human health to define issues and to rank
them in importance. The methods we used for assessing risk and
environmental impacts are intended to help regional and local
environmental managers make better environmental decisionsa key
objective of the IEMP. These decisions, in turn, should serve
to focus management efforts in the second phase of the project in
1-4
-------
the management of these risks. Thus, in terms of progress to-
wards protection of environmental resources and human health, the
real rewards from the adoption of these methods are not expected
until the conclusion of Phase I.
Finally, the integrated environmental management approach
allows for greater consistency in the assessment, and greater
flexibility and efficiency in the management, of risk to the
environment and human health. As noted above, laws and regula-
tions developed in the early stages of EPA's efforts involved
such diverse criteria as technology standards, air or water
quality, health effects, and costs. To address human health as
well as ecological and resource issues involving different pollu-
tants , sources, and exposure pathways, and to account for inter-
media transfers, lEMPs use the reduction of adverse effects on
human health, ecosystems, or resources as common measures of
effectiveness. Whenever feasible, use of such common measures
allow managers to compare different environmental problems and
solutions and to set priorities for further study or action.
Comparisons and priorities in turn make possible the coordination
of efficient control strategies that can achieve the greatest
reduction of risk at a given cost of control^.
Though the projects are intended to be "integrative," that
is, they look at pollutants and problems across media, they are
not comprehensive.Comprehensiveness requires that all potential
environmental problems be identified, evaluated, and ranked in
importance. Yet the sheer number of potential environmental and
toxic issues exceeds the resources available to the project. We
are limited by time and the size of our budget, which is not
enough to inventory all chemicals and their sources or to fill
all scientific gaps in knowledge. Instead, we must rely heavily
upon existing chemical inventories and scientific data that gene-
rally only suffice to characterize a limited subset of issues.
On the other hand, we must qualitatively estimate the social
"cost" of overlooking important issues because of absence or pau-
city of data. Hence, we weigh the adequacy of available data for
setting management priorities againsj; the expenditure of valuable
and limited resources in generating new information. Furthermore,
in the step that we call "scoping"--that is, defining the scope,
or comprehensiveness, of the projectand to a lesser extent
throughout the project, project participants make value judgments
regardingvthe types of issues (such as health-related, ecological,
or resource-related) that we will subject to our screening metho-
dology and the amount of resources we will devote to each.
At the completion of Phase II, EPA's involvemejnt in the
project ends. Nevertheless, the legacy of the IEMP will be
improved institutional structure and coherent management techni-
ques for making better decisions on future environmental issues.
The process that has led to these changes will hopefully continue.
1-5
-------
THE BALTIMORE IEMP
The Baltimore IEMP is a cooperative effort with the
governments of the State of Maryland, the City of Baltimore,
Baltimore County, and Anne Arundel County. Since EPA wishes to
help local governments manage environmental problems, it is
seeking out counties and cities that are likely to benefit from
the integrative approach. Baltimore is one of a number of older
Eastern Seaboard cities with a large number and variety of indus-
tries traditionally associated with environmental pollution. EPA
chose Baltimore because of the cooperative relationships that
existed between EPA and Maryland officials through such activi-
ties as the Chesapeake Bay Program, the progressive nature of the
State's and local jurisdictions' environmental programs, the
availability of data on toxics, and the presence of high quality
environmental staff with both monitoring and modelling capabili-
ties. Also not to be overlooked was the willingness of local
and State governments to work with EPA to address environmental
issues primarily of local concern.
The Baltimore IEMP represents EPA's second full-scale
geographic project. Most important, the Baltimore IEMP is diffe-
rent from all other geographic projects in that the authority for
project direction and policies and resource allocation resides
with State and local officials. EPA has set the overall study
objectives and the levels of resources for the project. EPA also
plays a support and advisory role, leaving project management
decisions and control to local authorities. To our knowledge,
such a working relationship with local authorities has never
been attempted before by EPA on such a scale.
Through this experiment in local participation in project
management of the Baltimore project, EPA hopes to achieve greater
local commitment to implement EPA's IBM concept. EPA, State, and
local officials have therefore focused our efforts in this first
phase of the project primarily on making the processthat is,
how management decisions are madework.
The primary objectives of the Baltimore IEMP are, in order
of priority:
o To establish the organizational framework at the regional
and local levels for identifying environmental issues and
setting priorities for government action oh them. The
project structure includes decisionmakers at the appropri-
ate governmental level of each pertinent agency.
o To help state and local governments develop a working
understanding of EPA's methods for analyzing issues and
evaluating control options.
1-6
-------
o To help State and local governments address high-priority,
site-specific multi-media and multijurisdictional problems
through risk assessment and management techniques. Pro-
gress toward this goal can be measured in tangible reduc-
tions in environmental risk.
o To help EPA set its national agenda and design and conduct
its programs. By understanding how Federal programs
affect conditions at the local level, EPA is better able
to redress inefficiencies and identify unaddressed prob-
lems. The Regulatory Integration Division, in particular,
needs to develop and improve methods for use in IEM, to
address shortcomings of a non-integrated approach to pol-
lution, and to test hypotheses regarding policies for
environmental management.
o To provide the analytical methods for determining manage-
ment priorities and to demonstrate their use on a limited
scale. These methods may include quantitative ri.sk assess-
ments, cost-effectiveness analysis, and other analytical
techniques.
EPA's choice of the Baltimore metropolitan area for this
study was not based on a belief that the area has a significant
toxics problem. Rather, EPA chose the Baltimore area because
both EPA and local officials wanted to explore better ways to
identify, assess, and manage the human health risks of environ-
mental pollutants In the area.
WORK PLAN FOR THE BALTIMORE IEMP
The work of the Baltimore project is broken up into two
phases. In Phase I, State, local and EPA staff conducted the
following activities:
o We developed institutional arrangements necessary for
running and sustaining the project and for determining
the scope of the project. .. We established cooperative
working relationships with representatives of the politi-
cal jurisdictions in the study area. We also formed two
committees: the Management Committee (MC) to make manage-
ment decisions about which problems should be addressed
and the Technical Advisory Committee (TAG) to advise the
MC and EPA on technical matters.
o We conducted a scoping process where we identified many
potential issues for study and established the criteria
to be used in the initial screening and assessment.
o We conducted an initial screening and assessment of issues
so as to set priorities among them for further analysis
in Phase II.
1-7
-------
The products resulting from this phase of the project relate
principally to process; we expected no real reductions in health
or environmental risk during Phase I. In addition to an effec-
tive institutional structure, the products are definitions of
the environmental issues and work plans for more detailed analy-
sis and for development of control strategies. These work plans
will be carried out in Phase II of the IEMP. Though we did not
intend to solve specific environmental problems in Phase I, the
existence of the IEMP and its committee structures did spawn
activities that have already led/to tangible health and environ-
mental benefits. These are detailed in Appendix C.
Comparison with Other Integrated Environmental Management
Projects
The EPA's Regulatory Integration Division (RID) has initi-
ated four other geographic projects to date. Only the first
project, conducted in Philadelphia is complete. The Baltimore
geographic project has been conducted concurrently with a pro-
ject in the Santa Clara Valley ("Silicon Valley") of northern
California. A project in the Kanawha Valley (West Virginia)
is drawing to a close. In 1986, RID initiated a new study in
Denver.
The Baltimore IEMP differs in two key ways from other lEMPs.
First, as mentioned above, it experiments with an innovative
means of managing an EPA project: management has been delegated
entirely to local authorities, with EPA in a support and advisory
role. Second, the initial scope of the project is much broader
than either the Philadelphia or the Santa Clara Valley project.
The unique management arrangement explains, in part, the broader
scope of the Baltimore project.
For both the Santa Clara Valley Phase I and the Philadelphia
studies, EPA intended only to study human health-related issues.
In the Santa Clara Valley, EPA and local authorities shared
common concerns about a specific potential public health risk.
The public had become concerned about drinking water, following
reports of contamination of the ground-water from which drinking
water was drawn. A similar commonality of interests existed in
Philadelphia, the first large-scale IEMP. Before the start of
the IEMP, Philadelphia had developed ambient air guidelines for
99 toxic air pollutants* These guidelines were the result of the
public's and Philadelphia health officials' concern about expo-
sures to toxic air contaminants. In particular, Philadelphia
health officials were interested in determining pollutants and
sources of those pollutants that violate ambient air guidelines.
EPA, on the other hand, specifically wished to test its methods
for assessing potential human health risks from toxic chemicals
released into the local environment.
In contrast, the Baltimore project began with a blank
«-.
1-8
-------
managers were potential candidates for inclusion in the project.
Initially, the number of potential issues and pollutants was
extremely largefar larger than the resources available for the
project. Thus, in Baltimore we first had to apply and, in some
cases, to develop the methods for narrowing down the number of
Phase I issues from the universe of potential environmental prob-
lems confronting the study area. [We describe this method in
Chapter IV.]
The Baltimore IEMP does not attempt to be a comprehensive
qualitative or quantitative assessment of all risks from toxic
chemicals in the Baltimore metropolitan area.Tomakesuchan
assessment would have requiredfar more than the resources avail-
able to the project. This report," then, is a review of the
methods developed and approaches taken in setting priorities
among a limited number of issues which included human health
risks, ecological damage, and resource protection. The initial
list of environmental issues was developed by local participants
in the study.
Adopting an approach to defining the scope of the project
that allowed for a more inclusive definition of study boundaries
and issues had far-reaching consequences. In addition to the
issues studied in other lEMPs, such as human exposure to toxics
through drinking water and the ambient air, Phase I of this
study also examined indoor air pollution, conventional pollutants
of water, and ecological damage to Baltimore Harbor. Greater
study breadth demanded a tradeoff, however. We had less time and
resources available for screening* issues within any broad cate-
gory of issues (that is, a group of issues that can be compared
against a common metric) and for developing risk estimates.
OUTLINE OF THIS REPORT
In the following chapters, we provide a brief introduction
to the Baltimore study area and how the project is managed,
describe how we conducted Phase I of the project, and present its
findings and conclusions.
o Chapter II, "Background on the Baltimore Study Area,"
provides an introduction to the climate, the geographic
boundaries, the physical setting, the demography, the
economy, the history, and the structure of local govern-
ment in the Baltimore region.
o Chapter III, "The Institutional Structure of the
Baltimore IEMP and Its Evolution," describes the manage-
ment of the project and how it came about.
*ln this context, screening means making a rough comparison of
issued against one or more criteria.
1-9
-------
Chapter IV, "Overview of Priority Setting Process,"
reports on how the priority-setting procedures were
developed and provides a framework for the following four
chapters.
Chapter V, "Analysis of Human Health Risks," reviews the
screening process for human health risks.
Chapter VI, "Analysis of Sources with Potential Adverse
Impacts on Ground-Water Resources," reviews the screening
process for ranking sources with potential adverse im-
pacts on ground-water resources.
Chapter VII, "Analysis of Ecological Impact," reviews the
screening process for ecological impacts.
Chapter VIII, "Completion of the Priority-Setting Pro-
cess," reports on the issues chosen for each type of im-
pact, the choice of issues based on the primary and then
the secondary criteria, and the results of the priority-
setting process.
Chapter IX, "Phase II Work Plans," presents the work
plans for the issues that are to be pursued in Phase II
of the IEMP.
Chapter X, "Summary and Conclusions," summarizes the con-
clusions the specific findings from Phase I activities,
and the limitations on interpretation of the results.
Appendix A, "General Methodology," discusses the organi-
zation of IEMP project, the concepts of risk assessment
and risk management, and the process of risk assessment.
Appendix B, "Sample Issue Paper," presents an example
of the kinds of issue papers that were written to provide
background for and to define the issues that had been
identified by the TAG.
Appendix C, "Epilogue," describes the activities that the
Baltimore IEMP has spawned and that have led to tangible
health and environmental benefits since the start of the
project.
1-10
-------
CHAPTER II
BACKGROUND ON THE BALTIMORE STUDY AREA
-------
II. BACKGROUND ON THE BALTIMORE STUDY AREA
In this chapter, we provide a general introduction to the
Baltimore study area. We describe the climate, ground and sur-
face waters, the water supply, general air quality, housing, the
general history of population growth and industrial development,
and the economic and population base. We also present mortality
statistics for the three political jurisdictions. We conclude
with a description of State and local government institutions
which deal with environmental problems.
BOUNDARIES OF THE STUDY AREA
The Baltimore study area covers Baltimore City, which
includes the Port of Baltimore, and Baltimore and Anne Arundel
Counties. The primary focus of the first phase of the IEMP,
however, was on a core area consisting of Baltimore City and the
urbanized, industrialized areas immediately adjacent. The total
land area representing the core study area amounts to approxi-
mately 300 square miles, lying in east-central Maryland on the
western shore of the Chesapeake Bay, the nation's largest estuary.
Also included are the reservoirs that are sources of drinking
water to the area. Figure II-l shows a map of the general study
area. Figure II-2 shows the study area in relation to Maryland.
CLIMATE
The climate of the Baltimore region represents the intermedi-
ate between the cold of the Northeast and the mild climate of the
South. The Appalachian Mountains to the. west and the Chesapeake
Bay and Atlantic Ocean to the east also cooperate in creating a
milder climate than that of inland areas of the same latitude.
In the winter, cold weather is moderated by the ocean. Summers
are warm, with several periods of hot, humid weather. The
monthly average temperature in January is 33 degrees F, and in
July, 76 degrees F. The relative humidity in February, March,
and April is 60 percent and in August, September, and October,
75 percent.
Annual average precipitation, mostly as rainfall, is about
45 inches and is relatively evenly distributed over the year.
Seasonal weather patterns are well defined: extended, low-
intensity frontal storms during the winter and spring monthst
and short-duration, high-intensity convective storms during the
summer and early fall.2
II-l
-------
Figure II-l
BALTIMORE I.E.M.P. STUDY AREA
PENNSYLVANIA
-------
Figure II-2
BOUNDARIES
OF THE
BALTIMORE IEMP CORE STUDY AREA
-------
GROUND-WATER RESOURCES
Productive aquifers (that is, aquifers that represent a
potential source of drinking water) underlie much of the coastal
area immediately surrounding Baltimore Harbor.3 Because of geo-
logical conditions, many of these aquifers are vulnerable to con-
tamination, though they are not currently being used for drinking
water.4 Furthermore, numerous industrial and municipal surface
impoundments and landfills and dumps located within the Baltimore
area are potential sources 'of ground-water contamination. The
extent of any damage to ground-water in the Baltimore area is un-
known. ^ Most of Anne Arundel County's and much of Baltimore
County's drinking water needs are served by ground-water.
SURFACE WATERS
The study area encompasses the Jones Falls, Back River,
Middle Run, Back Run, Gunpowder Run, Bodkin Creek, and Gwynn
Falls watersheds; the Lower Patapsco River; Baltimore Harbor
(including, the Inner Harbor, Curtis Bay, Northwest Branch, and
Middle Branch); and a number of small creeks. All of the water-
sheds fall within the Patapsco River Basin and drain into Chesa-
peake Bay. The total shoreline included within the Port of
Baltimore amounts to approximately 42 miles, with 28 miles in
Baltimore City and 14 miles in Baltimore and Anne Arundel
Counties.
Streams within the area are generally small, shallow, and
rapidly flowing. Each drains an area of a few square miles into
estuarine embayments.^
Jones Falls and its watershed (Figure II-3) cover some of
the more congested urban areas of the Baltimore Metropolitan
Region. The stream originates in Baltimore County; flows towards
the south into Lake Roland, an old, man-made water supply impound-
ment near the city/county jurisdictional boundary; and passes
through Baltimore City into a large conveyance tunnel before emp-
tying into Baltimore Harbor. The lower urbanized portion of the
watershed is severely degraded and representative of the variety
of water quality conditions in the region. Other important
streams are Herring Run, which flows through Baltimore City and
empties into the Back River, and Gwynn Falls, similar in size,
land use, and water quality to Jones Falls, which drains into
the Middle Branch of Baltimore Harbor. Drinking-water impound-
ments include Druid and Lake Montebello in Baltimore City.
Rivers and reservoirs not within the core study area are the
Upper Patapsco River, Loch Raven in Baltimore County, Prettyboy
Reservoir in Baltimore County, and Liberty Reservoir, which
straddles Baltimore and Carroll Counties.
II-2
-------
STATE OF MARYLAND
FIGURE E-3
THE BALTIMORE METROPOLITAN AREA
AND
JONES FALLS WATERSHED
-------
WATER SUPPLY
Water for residents of the core Baltimore study area comes
primarily from the City of Baltimore's municipal water system.
The sources for the municipal water supply are:
a. the Gunpowder River, which provides water to the Loch
Raven and Prettyboy Reservoirs--the Loch Raven Reservoir
has a primary water treatment plant;
b. the North Branch of the Patapsco River, which provides
water to Liberty Reservoir, which also has a treatment
plant; and
c. the Susquehanna River, which serves as a secondary
source, and only for emergencies or special circum-
stances.
A small percentage of total water consumption in the core study
area comes from ground-water, and only in areas not served by sur-
face water supplies.8
GENERAL AIR QUALITY
With the passage of State and Federal air pollution
legislation in the late 1960s and early 1970s, the Baltimore
area has made much progress in eliminating the spot, smoke, and
fly ash that clouded industrial areas. Nevertheless, the
Baltimore area does not meet National Ambient Air Quality
Standards for ozone, carbon monoxide, and particulate matter.'
It does meet the other air quality standards.
HOUSING
Like most older, industrialized cities of the East Coast,
the Baltimore area displays large gradations in the age and qual-
ity of its housing stock. Within the densely populated urban
core located primarily in Baltimore City, much-of the housing
stock consists of brick rowhouses built before World War II.
Many of these units, which had fallen into neglect, are now being
renovated.
Many of the older residential structures in Baltimore City
are suspected of having been painted with many layers of lead-
based paint. In addition, most houses in the study area have
copper pipes with solder in the plumbing systems.
With the recent redevelopment of the "Inner Harbor," there
is renewed interest in downtown living and new housing develop-
ments are cropping up close to the harbor. Ringing the densely
populated areas of Baltimore City are the newer suburban areas
of Baltimore and Anne Arundel Counties. In these areas, the
greater portion of the housing stock was built after World War
II, primarily since the 1960s.
II-3
-------
POPULATION GROWTH AND INDUSTRIAL DEVELOPMENT
The natural harbor formed by the 14-mile long estuary of the
Patapsco River where it meets the waters of the Chesapeake Bay
has historically presented an ideal setting for the growth of
commerce. The waters of the harbor facilitated the transport of
agricultural goods, particularly wheat, grown in the rich soils
of the Western Shore and Piedmont. By 1729, the land along the
rivers, streams, and inlets could claim a population of almost
1,000 people. With the uniting of Jones and Baltimore Towns in
1743, population growth accelerated and led to the first signs
of adverse population impact on the environment of the regiona
deterioration of the condition of a large marsh on Jones Falls.
The end of the 18th century saw Baltimore emerge as the most
important commercial port in Maryland, boasting a wide diversity
of industries. The many rapidly flowing streams invited the con-
struction of mills. In 1825, it became the country's leading
city for the milling of flour. The processing of .oysters became
an important industry with the development of the steamboat and
the discovery of the process for hermetically sealed metal cans.
By 1870, 10,000 workers were employed in the annual processing of
ten million bushels of oysters. The city soon became the nation-
al leader also in food processing and canning.
The construction of the railroads in the early 19th century
led to the development of the area as a key center for heavy
industry, particularly iron and steel. First came iron works,
then locomotive works and steel mills to construct the iron rails
and steel bridges. Sparrows Point steel works began with the
founding in 1887 of Maryland Steel Company, the progenitor of
Bethlehem Steel Corporation.
With World War I came large chemical plants on Curtis Bay.
Shipbuilding, large copper smelting plants, and oil and sugar
refineries located after the war also along the shores of the
lower Patapsco River.
With the city's rapid growth in the 19th century began
Baltimore's quest for an adequate supply of clean water.
Baltimore formed the nation's first public water supply company
and shifted its reliance on supplies from ground-waters and
springs to the surface waters of Jones Falls. Lake Roland on the
Jones Falls was constructed in 1861. Water from the Gunpowder
River provided the area's first interbasin transfer to the City
beginning in 1874.
As in most densely inhabited environments, streams in the
Baltimore area served a purpose other than just as a source of
drinking water and as an energy source; they were also used as a
means for disposing of waste. Throughout the 18th century and
II-4
-------
continuing until today, industries discharged their wastes into
the streams and creeks flowing to the harbor. Urban runoff
became a problem, as indoor plumbing introduced in the 188.0s
emptied into street gutters. To deal with these early water
pollution problems, in 1911 Baltimore built one of the nation's
first waste treatment plants, Back River Treatment Plant. There
are now plans to spend some additional $400 million to upgrade
this plant in the next few years. Baltimore City was also one of
the first major East Coast cities to build separate sanitary and
storm sewer systems.
Also during this period, growth upstream at the existing
sanitary sewer systems caused massive^ overflows of raw wastes
into nearby streams. These incidents forced the State Health
Department to impose building moratoria on certain watersheds
which, in some cases, lasted a decade. Today, problems still
exist such that septic tanks occasionally fail and sewer systems
60 to 70 years old overflow.
In the decades immediately after World War II, the port has
generally brought prosperity to the region. Stimulus to growth
of the port has come from many quarters, including grain and coal
export under the United N'ations Relief and Rehabilitation Program
on behalf of war-torn nations, expansion of the Sparrows Point
steel plant, and conversion of the Fairfield building yard into
shipbuilding and salvage operations* Yet the port's increasing
reliance on bulk cargo serving large industries, rather than the
highly profitable general cargo business, hinted at the larger
Baltimore area's vulnerability to changes in the fortunes of
these large, generally smokestack industries.
Baltimore Today and Tomorrow
In the 1980s, Baltimore remains one of the nation's most
important commercial and industrial centers. At .center stage is
Baltimore Harbor, the Baltimore area's most important economic
resource, and the commercial and shipping port it makes possible.
The port itself is estimated to have created some 79,000 jobs and
generates each year $300 million in State and local taxes. And
the port is in some way tied' to one-tenth of the value of all
goods and services produced by the State. Some 4,000 ships
from around the world use its facilities and nearly 37 million
tons of cargo, of which two-thirds is foreign cargo, are loaded
and unloaded here.
In recent years, urban renewal has revitalized portions of
the harbor, particularly the Inner Harbor and Fells Point. The
Inner Harbor is now one of the region's most desirable locations
for retail and tourist trade. Fells Point has become a focal
point for the new interest in downtown living.
II-5
-------
In addition to commerce associated with the port, manufac-
turing, despite its decline in relative importance to the
Baltimore area's economy in recent years, is an important source
of employment and revenue. Table II-1 lists the area's largest
manufacturing employers. The most important industrial employ-
ers are the primary metal and electrical/electronics industries.
Next is food processing, followed by the chemical industry.
Table II-1; Baltimore Area's Largest Manufacturing Employers(1977)
Employers
Number of
Establish-
ments
Value of
Number of Payroll Shipments
Employees (millions) (millions)
Chemical & Allied
Products 117
Food & Kindred
Products 192
Primary Metal
Industries 44
Electrical/Elec-
tronic Equipment 92
9,900
14,200
24,000
24,700
$150.7
$178.4
$457.1
$424.7
$1 ,192.1
$1 ,567.7
$1,860.2
$1,215.4
Source:U.S.Bureau of the Census, 1977 Census of Manufacturing
for Maryland, as reported in Regional Planning Council, Economic
Indicators; Baltimore Region, 1983 edition.
The largest single manufacturing employer in the study area,
Bethlehem Steel Corporation, employs approximately 20,000 people.
Like many other large manufacturing employers in the Baltimore
area, it is located within the Port of Baltimore.12
The Regional Planning Council predicts that most new growth
in the three jurisdictions will occur in the service and trade
sectors. In the past five years, employment in Baltimore City
declined by an estimated 35,000 jobs--half in manufacturing and
half in government. As Table II-2 (see next page) shows, antici-
pated growth in the service sector will more than offset contin-
ued declines in manufacturing. Service sector and trade growth
will also offset declines in the number of manufacturing jobs in
Baltimore County. The greatest employment growth will occur in
Anne Arundel County, primarily in the service sector.
POPULATION TODAY AND IN THE FUTURE
After hundreds of years of sometimes exponential growth,
population in the two counties and Baltimore City began to stabi-
lize in the 1980s. Table II-3 presents population statistics
for the Baltimore study area. The population of the three juris-
dictions in 1980 was 1,813,000. In 1985 it is estimated to be
II-6
-------
Table II-2: Changes in Employment in the Baltimore Study Area
Change
over
1980-851
% of Total
Employment
in 1985
Change
over
1985-901
% of Total
Employment
in 1990
Baltimore Study Area
10.5
31 .5
Baltimore City
Infrastructure
Manufacturing
Trade
Services
Government
Baltimore County
Infrastructure
Manufacturing
Trade
Services
Government
Anne Arundel County
Infrastructure
Manufactur ing
Trade
Services
Government
34,
8,
16,
-1 ,
10,
8
6
5
9
4
10,
12,
18,
37,
3
4
9
7
-18.2
11 .0
3.2
-6.9
6.6
13.9
-5.8
34.3
2.0
2.3
7.2
.7
13.1
20.6
10.1
14.2
26.4
29.7
19.7
12.2
10.3
18.8
18.8
40.0
3.0
1.8
-5.1
-0.4
7.2
-0.5
11.9
0.8
-3.5
4.5
9.9
0.2
16.6
1.3
-0.1
4.1
7.9
3.4
10,
11
18
39
10
12
26
31
20.4
,0
7
7
4
19.1
11 .9
9.5
19.3
21 .0
38.5
1 Changes in the number of jobs over the indicated period, in
thousands, by place of work.
Source: Regional Planning Council, December 1985,;
1,822,500 and is projected to increase to 1,838,600 by 1990--
about a 1 percent increase over the decade. Like most older,
industrialized cities of the Eastern Seaboard, Baltimore City's
population, however, is declining. It is projected to fall
from 786,800 in 1980 to 749,500 in 1990--an almost 5 percent
decrease over the decade. Most of this decrease will be offset
by growth in population in Anne Arundel County.
The population density of Baltimore City is 9,620 people per
square milea density almost ten times greater than than of its
surrounding counties. However, the populations of the counties
are concentrated around the jurisdictional boundaries shared with
Baltimore City.
II-7
-------
Table II-3: Population Statistics for the Study Area
Past and Projected Population:
Baltimore City
Baltimore County
Anne Arundel County
Total
T980
786,800
655,400
370,800
1 ,813,000
1985
760,000
665,200
397,300
1 ,822,500
1990
749,500
669,100
420,000
1 ,838,600
Population Density (*85) Persons per Sq. Mile
Baltimore City 9,620
Baltimore County 1,094
Anne Arundel County 952
Source:Regional Planning Council, February 1986.
MORTALITY STATISTICS
Table II-4 (see next page) shows total deaths for selected
causes in the two counties and Baltimore City. The data in the
table serve only to give a very rough idea of the relative impor-
tance of the various causes of death in the three jurisdictions.
They are not necessarily good indicators of the general health
of the population.
STRUCTURE OF STATE AND LOCAL GOVERNMENT
In this section of this chapter, we describe the institution-
al context in which the Baltimore IEMP is taking place. Under-
standing the agencies, the formal relationships among them for
making decisions of mutual interest, and their legal responsibi-
lities is crucial for developing effective and implementable
recommendations. Solving environmental problems requires coope-
ration among agencies and jurisdictions.
The existing structure of state and federal environmental
programs is the direct consequence of piecemeal enactment of fed-
eral law. Over a number of years, laws were passed to address
environmental problems by the medium immediately affected by
pollution, such as air pollution or water pollution. With each
new federal law came the governmental apparatus to implement it.
At the federal level, the result has been departmentally separate,
medium-specific programs, such as air, water, and hazardous
waste (land and ground-water ) programs. States have generally
adopted analogous institutional structures to carry out their
responsibilities under these federal laws or to establish state
programs not mandated by federal law.
Under most federal laws, the federal government (EPA in this
instance) sets the rules and the states must enforce them under
the threat of losing federal funds. This entails, in part, devel-
oping implementation plans for how enforcement of these rules by
II-8
-------
TABLE II-4: DEATHS BY SELECT CAUSES IN
BALTIMORE CITY, BALTIMORE COUNTY, AND ANNE ARUNDEL COUNTY
Data Are for 1984
Causes
Number of Deaths
Diseases of the heart
Cancer^/3
Cerebrovascular disease
Accidents and adverse
effects
Pneumonia & influenza
Septicaemia
Chronic obstructive
pulmonary diseases &
allied conditions
Homicide
Diabetes Mellitus
Congenital anomologics
Total deaths 4
Baltimore
City
3,440
2,248
603
287
262
215
215
204
178
39
Baltimore
County
2,198
1,379
382
186
147
75
163
36
101
25
Anne Arundel
County
827
658
139
113
52
38
92
20
62
12
9,554
5,638
2,474
Source: Maryland Center for Health Statistics, DMH, Maryland Vital
Statistics; Preliminary Report 1984.
1 Population estimates for 1984 are from the Regional Planning
Council
2 Figures represent actual rather than age-adjusted rates.
3 It is important to realize that there are approximately 2 cases
of cancer for every cancer death, according to 1983 American
Cancer Society Data for the nation (actual figure: 1.92 cases
per cancer death).
4 Total includes deaths from other causes.
state and local agencies will be carried out. States and local
agencies do have the option of doing more than what federal law
requires, however. In the case of toxics, many states (such as
Maryland, California, New York, and New Jersey) have aggressively
developed their own programs. This is because EPA has issued few
rules and regulations addressing the control of toxic substances
as opposed to the more "conventional" pollutants.
i
In describing the institutional framework for the Baltimore
IEMP, we identify the program medium, such as air or water, and
then describe the responsible state and local entities and their
specific responsibilities.
II-9
Will
-------
Institutional Framework for Water Problems
Both State and local governments carry out regulatory
programs relating to water problems. Local governments also have
responsibility for issues not specifically addressed in Federal
or State law.
State regulatory responsibilities for water are divided up
among agencies according to whether the water is considered an
environmental resource or an item of interest to public health.
Program distinctions are also made according to the geological
location of the wateri.e., whether it is ground or surface
water.
Surface Waters. The Office of Environmental Programs (OEP) of
the Maryland State Department of Health and Mental Hygiene is
responsible for all clearly health-related issues regarding water
use. It is charged with managing and regulating the waters of
the State to protect the beneficial uses of
o water contact recreation;
o fish, other aquatic life, and wildlife;
o shellfish harvesting;
o public water supply;
o agricultural water supply; and
o industrial water supply.
Though OEP is nominally responsible for protecting all sur-
face waters, these responsibilities actually pertain primarily
to protecting waters from sources of the pollution that can be
clearly defined (point sources) and only to a subset (agricul-
tural runoff) of the pollution deriving from large numbers of
small sources distributed over a large area (non-point sources).
Point sources fall under the purview of the Federal Clean Water
Act, for which OEP is given the State enforcement responsibility.
These include issuing NPDES (National Pollution Discharge Elimi-
nation System) permits for point source effluent discharges and
enforcing technology-based emission standards. It also issues
permits for land disposal of wastewaters.
Publicly owned water treatment works (POTWs, or sewage
treatment plants) are regulated as point sources. OEP issues
effluent permits and ensures that POTWs are in compliance with
Federal and State regulations and the water-quality management
plans for individual river basins.
The POTWs, and hence the local government operating them,
are responsible, for implementing sewage pretreatment ordinances
established by the State or EPA and for monitoring the quality of
11-10
-------
the influent to sewer treatment plants. They must perform secon-
dary and advanced treatment of sewage as required by State or
Federal law. The three POTWs in the Baltimore core study area
are Patapsco, located in Baltimore City; Back River, located in
Baltimore County; and Cox Creek, located in Anne Arundel County.
Responsibility for State control of nonpoint sources of
pollution, such as urban (stormwater) or agricultural runoff, is
shared by the Water Resources Administration in the Maryland
Department of Natural Resources (DNR) and OEP. DNR establishes
and enforces policies and regulations relating to stormwater
(especially urban runoff) management, and it approves county and
municipal ordinances and implementation plans. DNR is also
charged with preventing oil pollution by setting and enforcing
standards and regulations for oil operations and for managing
watersheds and resources. The latter entails flood management
and issuing permits for watershed use, erosion control, dam safe-
ty, dredging, and water appropriations and licenses for mining.
OEP's share of responsibility for controlling nonpoint source
pollution lies in administering programs, in coordination with
the State Department of Agriculture, to help farmers reduce agri-
cultural runoff and issuing permits for agricultural discharges.
At the local level, the Departments of Public Works for the
City of Baltimore and Anne Arundel and Baltimore Counties are
responsible for administering the regional stormwater management
plan, which includes permitting, inspecting, and collecting per-
formance bonds. In Anne Arundel County, the Department of Public
Works and the Planning and Zoning Department share responsibilty
for the County's innovative watershed management program. These
agencies, along with the Baltimore County Health Department, also
enforce nuisance laws, regulate the discharge of liquids and
other offensive matter, issue permits for disposal sites, require
environmental impact statements for subdivisions within the water-
shed of any reservoir used for the public water supply, approve
or disapprove the subdivisions within the watersheds, and regu-
late sanitary landfills. They also oversee private water and
sewage disposal systems, conduct the day-to-day activities of the
State Water Quality Program, review watershed management agree-
ments, and develop and carry out the master water and sewer plans.
Ground-Water. OEP is charged with protecting underground sources
of drinking water from contamination. It regulates the discharge
or disposal of waters or wastewaters into underground-waters and
issues permits for ground-water recharge systems, wastewater
effluents disposed of on land, and discharges of leachate from
landfills to surface or ground-waters. The Water Resources Admi-
nistration issues permits for ground-water appropriations. The
safety of the drinking water is the responsibility of the County
Health Department in Baltimore County and of the Department of
Utilities in Anne Arundel County. Drinking water for Baltimore
11-11
-------
City and the section of Baltimore County in the core study area
is exclusively from surface waters. Some residents of Anne
Arundel County also rely on the metropolitan water supply for
their drinking water. Residents of areas outside the core study
area draw their drinking water from ground-water supplies.
Surface and Ground-Waters Potentially Affected by Activities
Associated with Generating Energy. The Energy Administration of
the Department of Natural Resources is responsible for protecting
surface and ground-waters from activities associated with genera-
ting energy. Its responsibilities include siting power plants
and writing permits for coal mining operations.
Regional Planning Relating to Water Use. The Regional Planning
Council conducts areawide planning for the Baltimore Region. Its
authority is based partly on the Clean Water Plan for the Balti-
more Region adopted in accord with the Federal Clean Water Act.
It develops policies for implementation of the General Develop-
ment Plan for the Baltimore Region and coordinates intergovern-
mental planning. The latter involves cooperative preparation and
amendment of the General Development Plan and the review of plans
arid projects by State and local governments which may have an
impact on the region.
Institutional Framework for Air Problems
The State Office of Environmental Programs has been delegated
the responsibility for enforcing the Federal Clean Air Act in the
State of Maryland. OEP also has authority for adopting rules and
regulations regarding the control of statewide problems with toxic
air contaminants. To ensure compliance with air quality standards
and regulations, it conducts air monitoring, investigates commer-
cial operations potentially affecting air quality, registers air
dischargers, issues annual operating permits and permits to con-
struct new facilities that may emit pollution to the air, develops
plans for compliance, investigates complaints, and conducts perio-
dic inspections. ~
Monitoring air and enforcing compliance are performed by both
State and local agencies. The responsibility for point sources is
divided between State and local authorities. In general, the
State is responsible for monitoring compliance at complex sources,
such as bulk petroleum storage facilities and major chemical
plants, and for mobile sources, such as cars and trucks.
At the local level, the Bureau of Environmental Services of
the Baltimore County Department of Health and the Department of
Health of Anne Arundel County carry out the mandate for protect-
ing air quality. The responsibilities include monitoring for tra-
ditional air pollutants (so-called criteria pollutants), regulat-
ing operations involving combustion (such as incineration),
11-12
-------
participating in program planning, and monitoring compliance for
sources that are a joint State and local responsibility and for
sources that are designated as strictly local. The State carries
out enforcement responsibilities for Baltimore City.
Institutional Framework for Land Disposal Problems
The division of responsibilities for regulating land dispos-
al of wastes is based primarily on whether or not the waste is
considered hazardous. Regulating activities associated with the
handling and disposal of hazardous wastes is the province of the
State; non-hazardous wastes that can be disposed of in sanitary
landfills are the responsibility of local agencies. Local agen-
cies do, however, have responsibility to inventory the types,
quantities, and destinations of hazardous wastes and regulate the
primary generators of hazardous wastes (chemical processing faci-
lities).
The specific hazardous waste activities and the State agen-
cies responsible for them are:
o hazardous waste siting, which involves reviewing siting
applications and holding public hearingsMaryland Hazar-
dous Waste Siting Board;
o regulation of private hazardous waste disposal sites,
which involves permitting and monitoring and enforcing
the Federal Resource, Conservation, and Recovery Act
(RCRA)OEP; and
o cleanup of abandoned and inactive hazardous waste dumps
and spills, which involves administering the State Super-
fund program, and permitting, manifesting, and licensing
all transporters and facilities that store and dispose of
hazardous wastesOEP.
In Baltimore City, the Bureau of Solid Waste in the Depart-
ment of Public Works has the responsibility for collection of
seasonal wastes and for land disposal of non-hazardous wastes.
This latter responsibility includes the operation of sanitary
landfills. The City also has oversight, through the Northeast
Maryland Waste Authority, over the operation of two privately-
operated incinerators.
In Baltimore County, sanitary landfills are regulated by
the Baltimore County Department of Health. The responsibilities
entail monitoring solid waste facilities and enforcing rules and
regulations, helping industry in proper disposal practices, and
regulating sewage sludges that are not considered hazardous.
In Anne Arundel County, waste management is conducted by
four departments. The Bureau of Solid Waste of the Department of
Public Works is involved with constructing and operating the
11-13
-------
sanitary landfills. The Department of Inspections and Permits
issues permits and conducts inspections for compliance with
State and local laws and regulations. The Department of Health
inspects for health violations. Finally, the Department of Plan-
ning and Zoning issues special exceptions for landfills.
All of these State and local agencies involved in the
control of toxics and the protection of the environment are repre-
sented in the institutional structure of the Baltimore IEMP. The
nature of this institutional structure and how it evolved are
described in Chapter III.
11-14
-------
1. Baltimore City Planning Department, The Baltimore Book, p. 11
(1986).
2. Regional Planning Council, Growth & Impacts in a Maryland
Watershed; Urbanization in the Jones Fall, 1600-1981, conducted
as part of the Jones Falls Urban Runoff Project by Samuel Martin,
et al, p. 3 (1981) .
3. Maryland Air and Water Quality Atlas, p. 43.
4. Maryland Air and Water Quality Atlas, p. 45.
5. Maryland Air and Water Quality Atlas, pp.45-48.
6. Regional Planning Council (1981), p. 25.
7. Ibid.
8. Regional Planning Council, 1982 General Development Plan;
Environment, p. 44 (October 19lJ2T"I '
10. The sources for the historical information are Regional
Planning Council's 1981 report for the Jones Falls Urban Runoff
Project, Growth & Impacts in a Maryland Watershed: Urbanization
in the Jones Falls, 1600-1981, by Samuel Martin, Roberto Arguero,
Jr., Virginia Driscoll, and Karl Weaver and Baltimore City
Planning Department's The Baltimore Book (1986).
11. Baltimore City Planning Department, p. 11.
12. Ibid .
11-15
-------
CHAPTER III
THE INSTITUTIONAL STRUCTURE
OF THE BALTIMORE IEMP AND ITS EVOLUTION
-------
III. THE INSTITUTIONAL STRUCTURE OF THE BALTIMORE IEMP
AND ITS EVOLUTION
This chapter describes the management and the institutional
organization of the Baltimore IEMP, the evolution of this arrange-
ment, and public education and involvement in the project.
RATIONALE FOR THE INSTITUTIONAL ARRANGEMENT
EPA intends that the IEMP be useful to public health and
environmental decision-makers within State and local government
agencies. The EPA understands that it is these individuals who
must directly carry out, and are most affected by, environmental
decisions. Yet, State and local decision-makers are unlikely to
accept the basic IEMP approach and commit themselves to the suc-
cess of the project unless priorities reflect State and local
concerns. EPA has, for this reason, encouraged the active par-
ticipation by all local responsible parties and agencies in
managing and coordinating the project.
The Baltimore project represents an experiment to test the
hypothesis that local project control of the IEMP will lead to
active participation and commitment by State and local authori-
ties. Thus, in a unique partnership, EPA has delegated key
decision-making authority of the project to State and local
officials. In the section of this chapter entitled "Evolution
of the Management Structure," we explain why the EPA decided to
try this approach.
THE INSTITUTIONAL STRUCTURE
The lEMP's institutional structure consists of two merge
committees: the Management Committee (MC) and the Technical
Advisory Committee (TAG). They also represent the vehicles for
State and local participation. EPA provides administrative, tech-
nical, and analytical support. In the second phase of the pro-
ject, additional support will come from the Risk Assessment
Review Panel, consisting of experts from the Johns Hopkins Univer*-
sity. It will advise the committees on scientific and technical
questions regarding human health risk which arise during the
project. Figure III-l illustrates the organizational relation-
ships.
The Management Committee (MC)
The four members of the MC representing the three local
political jurisdictions and the State are drawn from the follow-
ing agencies:
o the Mayor's Office, City of Baltimore;
III-l
-------
Figure ffl-1
BALTIMORE ffiMP
ORGANIZATIONAL RELATIONSHIPS
MANAGEMENT COMMITTEE (MC)
RISK-ASSESSMENT
REVIEW PANEL (RARP)
TECHNICAL ADVISORY
COMMITTEE (TAG)
EPA
REGULATORY
INTEGRATION DIVISION
(RID)
-------
o the Office of the Baltimore County Executive;
o the Qffice of the Anne Arundel County Executive; and
o the Maryland Department of Health and Mental Hygiene,
Office of Environmental Programs.
These local agencies were chosen because they represent the
highest-level policy-making bodies within their political juris-
dictions. The representatives were selected by the Mayor, in the
case of Baltimore City, and by the Chief Executives, in the case
of Baltimore and Anne Arundel Counties. The Maryland Department
of Health and Mental Hygiene, Office of Environmental Programs
is the State agency that customarily handles EPA projects and
that is concerned with issues involving human health. The estab-
lishment of the MC is intended to increase the likelihood that
the work conducted by the project reflects the attitudes and con-
cerns of greater Baltimore and responds to regional and local
needs .
The MC's functions include:
o managing and overseeing the IEMP,
o selecting the environmental issues upon which the pro-
ject focuses,
o providing local guidance to EPA on political realities,
o determining project strategies,
o overseeing the development and implementation of work-
plans , v
o reviewing EPA analyses and recommendations , and
o issuing results and recommending control actions.
The MC does not determine the EPA1 s contribution to the funding
of the project or the extent of other resources the EPA provides.
It does, however, control the allocation of funds among IEMP
activities.
The membership of the MC is presented in Figure III-2. The
MC chose as its coordinator Dr. Max Eisenberg. The MC generally
meets once a month or whenever a key management decision must be
made. For each meeting, EPA staff prepare an agenda and neces-
sary background documents. Decisions are made by consensus.
The Technical Advisory Committee (TAG)
The TAG advises the MC on the technical and scientific
aspects of the project. It oversees and comments on all EPA and
consultant work for technical competence, paying particular atten-
tion to methods, assumptions, monitoring and modeling, and basic
technical approaches used in the project. The TAG recommends and
suggests issues to study within the limits of the resources and
schedule of the project. For instance, it identified the origi-
nal list of prospective issues for detailed IEMP study. Finally,
III-2
-------
Figure III-2
Current Members of the Baltimore IEMP Management Committee
Mr. J. James Dieter, Assistant Director, Bureau of Environmental
Service, Baltimore County.
Dr. Max Eisenberg, Director, Science and Environmental Health,
Science and Health Advisory Group, Office of Environmental
Programs Department of Health and Mental Hygiene, State of
Maryland.
Mr. Robert Perciasepe, Chief of Capital Improvement Program,
Department of Planning, City of Baltimore.
Mr. Claude Vannoy, Assistant to the County Executive for Land
Use, Anne Arundel County.
Former members of The Baltimore IEMP Management Committee
Mr. Daniel Body represented Anne Arundel County.
Ms. Virginia Kearney represented the City of Baltimore.
-------
serving as liaisons to others in the agencies they represent,
members of the TAG enhance coordination of the IEMP with ongoing
projects at their respective agencies.
A specific example illustrates this coordinating role. The
question was raised at a TAG meeting if any risk assessment had
been made of pathogenic problems arising from urban stormwater
runoff in Baltimore. The representative from the Regional
Planning Council indicated that such a study had been done by a
researcher at Johns Hopkins University, who found that the risks
were low, even though the streams do not meet water quality stan-
dards for fecal coliform. This one, relatively small, example is
but one of many instances of communication and coordination that
would not have occurred or would have happened more slowly in the
absence of the TAG. (See Appendix C for other instances of for-
tuitous interplay among the participants.)
The fifteen-member TAG is composed of technical managers
from the Gity of Baltimore, the two counties, the State, as well
as representatives from the Baltimore metropolitan area's Region-
al Planning Council and the academic community. It has been
designed to help the project make maximum use of existing data
and capabilities within State and local governments and other
area institutions. The members are experts in their fields,
which range from geology to toxicology to engineering.. The
members of the TAG and their affiliations are listed in Figure
III-3.
The membership of the TAG was determined in the following
way. The coordinator of MC compiled a tentative list of indivi-
duals whose skills and interests would most likely match the
advisory needs.of the MC. This list was reviewed, amended, and
then approved by the MC as a whole. EPA staff contacted the
individuals on the list and solicited their interest in joining
the project in an advisory role. For individuals who declined or
were unable, for whatever reason, to accept the invitation, EPA
staff identified alternative candidates. Final selection of
alternative candidates was approved by the Management Committee.
The Chairman of the TAG is Dr. Jared Cohon, Vice Provost for
Research and Professor of Geography and Environmental Engineering
at the Johns Hopkins University. Vice-Chairman is Samuel Martin,
formerly with the Regional Planning Council and now a private
consultant. A chairperson from acaderaia was selected to promote
the impartiality of decisions. The TAC meets as often as deemed
necessary, which is generally once per month. Also, for project
coordination, the Chair of the TAC attends MC meetings.
Though the TAC is not a policy-making body, it does
incorporate policy considerations into its deliberations and
makes recommendations on policy to the MC. The following example
III-3
-------
Figure III-3
Members of the Baltimore IEMP Technical Advisory Committee
Don Andrew, Administrator, Engineering & Enforcement Programs,
Office of Environmental Programs.
Charles Billings, PhD, Associate Professor, Environmental Health
Engineering, School of Hygiene and Public Health, Johns
1 Hopkins University.
Philip Clayton, Manager, Cooperative Clean Water Program,
Regional Planning Council.
Emery Cleaves, Principal Geologist, Maryland Geological Survey
(Chairman,Ground-Water Subcommittee).
Jared L. Cohon, Vice Provost for Research, Professor of Geography
and Environmental Engineering, Johns Hopkins University
(Chairman).
N. Singh Dhillon, Director, Environmental Health, Anne Arundel
County Department of Health.
Tom Ervin, Environmental Planner, Anne Arundel County Office of
Planning and Zoning (Chairman, Ecological Subcommittee).
Katherine Farrell, MD. MPH., Chief Division of Environmental
Disease Control, Office of Environmental Programs.
David Filbert, Director, Division of Air Pollution Control,
Baltimore County Department of Health.
Frank Hoot, Assistant Commissioner, Environmental Health,
Baltimore City Health Department (Chairman, Human Health
Risk Subcommittee) .
Paul Keenan, Administrator, Pollution Control Section, Baltimore
City Department of Public Works, City of Baltimore.
John W. Koontz, represented the Enforcement Program, Waste
Management Program Administration, State of Maryland.
Sam Martin, Consultant (represented Regional Planning Council
during Phase I) (Vice Chairman).
Janice Outen, Supervisor of Water Quality, Baltimore County
Department of Health.
Colin Thacker, Director, Northern Environmental Services,
Baltimore County Health Department.
Bill Wolinski, Water Quality Goodinator, Department of Public
Works.
-------
illustrates the functional overlap between technical review and
consideration and recommendation of policy.
Because scientific data vary greatly in their quality and
usefulness in decision-making, decisions must often be made re-
garding whether sufficient evidence exists to support an analyti-
cal conclusion. The TAC, among whose responsibilities is the
review of data developed for assessing human health risks, is
therefore charged by the MG to determine data quality standards.
The TAC identifies situations where science and science policy
converge (that is, where assumptions must be made to permit sci-
entific inferences to be drawn on health issues of concern to the
IEMP) and makes recommendations to the MC on how these science
policy issues should be resolved.
Many of the decisions of both committees must be based upon
cutting-edge research into often controversial health issues.
The MC established the non-governmental Risk Assessment Review
Panel to provide independent evaluations of these data and re-
lated analyses. Scientists for this panel are on the faculty of
the School of Hygiene and Public Health of the Johns Hopkins
University. The panel's function is to review the interpreta-
tions of the data and the assumptions EPA uses in making assess-
ments of risk to human health. The panel is not responsible for
the quality or appropriateness of the technical data. The ser-
vices of the panel will be drawn upon in Phase II. The composi-
tion of the panel is presented in Figure III-4.
EVOLUTION OF THE INSTITUTIONAL STRUCTURE
What began as a traditional EPA project in the summer of
1983 evolved into an experiment in local control of a joint EPA/
State and local government project. To test the hypothesis
that delegating management authority will lead to active local
participation and commitment, EPA decided to shift gears shortly
after starting the project and commit itself to a nontraditional
EPA-State/local relationship. The management structure is the
product of a series of meetings of EPA with State and local offi-
cials working together to find the right institutional formula
for the success of the Baltimore IEMP.
At the start of the Baltimore IEMP, EPA made all the key
management decisions, with limited participation by State or lo-
cal authorities. Yet EPA had not originally intended State and
local involvement to be so limited. Instead, the plan was to
have representatives of all relevant jurisdictions, as well as
EPA's Region III and the Chesapeake Bay Program, participate in
a joint project management team with staff from EPA headquarters.
These representatives would keep the jurisdictions informed and
the jurisdictions in turn would have veto power over EPA's direc-
tion of the project. Two advisory groups were also plannedone
III-4
-------
Figure III-4
Members of the Risk Assessment Review Panel
Robert Frank, (RARP Coordinator), M.D., Professor,
Department of Environmental Health Sciences, Inhalation
iToxicology.
Morton Corn, Ph.D, Professor, Department of Environmental
Health Sciences, Environmental Health Engineering.
Edward A. Emmett, M.D., Professor, Department of Environmental
Health Sciences, Occupational Medicine.
Gareth M. Green, M.D., Professor and Chairman, Department of
Environmental Health Sciences
Charles A. Rohde, Ph.D, Professor and Chairman, Department of
Biostatistics.
Robert J. Ru£in, Ph.D., Professor, Department of Environmental
Health Sciences, Toxicology.
David L. Swift, Ph.D., Professor, Department of Environmental
Health Sciences, Environmental Health Engineering.
Melvyn S. Tockman, M.D., Ph.D., Associate Professor, Department
of Environmental Health Sciences, Occupational Medicine.
-------
for scientific advice, to be established under the auspices of
the State Governor's Council on Toxic Substances and to consist
of nongovernment experts, and the other for economic and regula-
tory review. In contrast to expectations, this initial manage-
ment committee convened only _once--in the fall of 1983--and the
technical committees did not meet at all.
In the first few months after the project started, manage-
ment decisions were nevertheless being made. The IEMP developed
and executed a technical work plan and, by early 1984, produced
a preliminary data base on air emissions, water discharges,
drinking-water quality, and hazardous waste characteristics. The
State's role was limited to providing access to monitoring sites
and local authorities had yet to be brought into the project.
In the winter of 1984, staff at EPA headquarters recognized
that the absence of involvement by the management committee in
the oversight and conduct of the technical work made the Balti-
more project like any other EPA project--EPA-controlled, with
little participation in decision-making by State and local autho-
rities. They saw the situation as presenting a unique opportu-
nity for EPA to experiment with local control, relinquishing the
authority for key decisions regarding project direction and poli-
cies and allocation of budgeted funds. Through a novel institu-
tional arrangement, EPA hoped it could effect an enduring multi-
governmental structure that could effectively address such envi-
ronmental issues as cross-media transfers of toxic substancesan
important IEMP objective.
Through the rest of 1984, management authority was gradually
shifted over to a reconstituted Management Committee, which had
served initially in an advisory role in managing the project,
until the committee acquired the responsibilities consonant with
its present role. To facilitate the transition to local control
of the project, EPA hired a risk assessment and management con-
sultant to assist in the formation of the new institutional struc-
ture and to advise local management on risk assessment issues.
Participation by the State in the management committee was made
contingent upon participation by the local authorities. To this
end, EPA had to assure all entities that major management deci-
sions would reflect State and local concerns. EPA did, however,
reserve the right to terminate the study at any time if the ori-
ginal objectives of the IEMP (as stated in Chapter I) wer"e not
being achieved. By this condition, EPA" made clear that only the
means to reach these objectives, and not the objectives them-
selves, were being left to local control. EPA retained Dr. Erik
Rifkin who is a professional facilitator and expert in risk com-
munication to assist in this transition. Dr. Rifkin provided
assistance to all the parties--EPA, State and local officials--in
clarifying questions, suggesting solutions, and generally resolv-
ing concerns in a positive manner.
III-5
-------
The first order of business of the new MC related to the
structure of the MC itself. EPA suggested that the committee
elect a chair, but the members argued against this because of
concern that the appointed person could exert unequal authority
over the other jurisdictions represented. Committee members
agreed that a chairperson was not needed because of the small
size and closeness of the group. Designation of a chairperson
would also incorrectly connote a position of power or authority
to the jurisdiction represented by the person serving as chair-
person. Any documents produced by the project and bearing the
name of the MC chairperson could be erroneously viewed as
imparting sole authorship. This would not be a true representa-
tion of the equal decision-making status among the participating
governments a condition agreed to by all governments before
their participation. Instead, they agreed upon a State repre-
sentative (Max Eisenberg) as coordinator to facilitate communica-
tion and arrangements within the committee.
The TAG, too, was concerned about one jurisdiction's having
undue influence on decisions of the committee. Nevertheless, it
recognized the need for a chair because of the size of the commit-
tee. To resblve this dilemma, the TAG formed a nominating commit-
tee for the chair. An academic representative (Jared Cohon),
believed to be more impartial in jurisdictional disputes, was
chosen to promote a neutral perspective.
To facilitate coordination with the TAG, the chairman of
the TAG was to attend MC meetings. Finally, the MC decided that
the TAG would provide only advice on technical matters and make
recommendations on key decisions, but not have authority over
direction of the project. Final authority on all key decisions
was reserved for the MC.
The TAG did not always meet as one group. The TAG eventual-
ly spawned subcommittees to improve the TAC's ability to review
and discuss a wide range of issues. Prior to forming the subcom-
mittees however, the TAG spent several months working .together
to identify issues, agree on a procedure for evaluating issues,
selecting issues for further study and creating the appropriate
subcommittees. The subcommittees analyzed, discussed in detail,
and ranked issues that had been grouped according to the kind of
environmental impact that they implied. The whole TAG did a
final ranking to recommend issues to the MC. [In Chapter IV and
Chapter VIII, we describe the priority-setting process.] TAG
members volunteered for one or more of the subcommittees. In
his function as chairman, Jared Cohon chose subcommittee chairs.
The "issue" subcommittees met for four months during the Phase I
analysis.
III-6
-------
PUBLIC EDUCATION AND INVOLVEMENT
The general public was kept informed of the progress of the
project through educational outreach efforts. Status reports were
periodically sent out to over 250 groups and individuals on a
local mailing list. Also, in June 1985, the IEMP held a public
seminar attended by roughly 150 people to inform the public
regarding project methodology, status of the project, and the
issues under review for Phase II study.
In the concluding stage of Phase I, the general involvement
in the IEMP widened through membership in workgroups established
to develop workplans for Phase II issues that is, after the ini-
tial 32 issues had been screened and Phase II issues tentatively
identified. Though chaired by TAG members, the workgroups were
comprised of representatives from industry, public interest
groups, government, and acaderaia, amounting to nearly 60 people.
A list of these individuals and the agencies and interests they
represent (who were not already listed on the TAG list) is pre-
sented in Figure III-5.
The government participants in the project recognized the
need for and the benefits of direct public participation and
other types of public involvement in the project. However, for
many reasons, they were unsure as to when the . public should
participate.
First, the TAG as a committee of the whole, viewed the
project as atypical of other environmental projects. The TAG
had identified 32 issues of potential concern* Before the actual
screening of issues, no one could possibly know which of these
would be selected for Phase II study. Because objectives speci-
fic to issues were as yet undecided, the TAG could not communi-
cate work plan objectives and the expected results. This situa-
tion Tvas completely unlike the development of local waste water
pretreatment ordinances, for example, where there is only one
major issue to be decided and work steps are fairly evident.
Second, the large number of issues for potential study
presented a difficult situation into which to invite public par-
ticipation. The range of identified topics covered almost all
locally important environmental issues. To give equal treatment
to each issue, as well as to be fair to the local constituency
that inevitably develops around specific issues, the TAG would
have had to invite all interested groups to participate. This
obviously would have increased the number of project participants
by.several dozen. The coordination, communication, and logisti-
cal complexities of the project would have increased several-fold.
Finally, the TAG realized that the project could not afford
to carry all 32 issues into Phase II. As a result of priority-
III-7
-------
Figure III-5
Additional Participants in Phase I Baltimore 1EMP
Through their Membership on Workgroups
Anthony S. Bonaccorsi
Eastern Stainless Steel Company
Daryl Braithwaite
Clean Water Action Project
Midhael K. Hettleman
The Southern Galvanizing Company
Genevieve M. Matanoski
Johns Hopkins School of Hygiene and
Public Health
Darryl W. Palmer
FMC Corporation
%-
Susan S. G. Wierman
Office of Environmental Programs,
State of Maryland
Bruce Windsor
American Lung Association of Maryland
r
Joe Macknis
U.S. EPA Cheaspeake Bay Program
Rich Batiuk
U.S. EPA Chesapeake Bay Program
Mary Jo Garreis
Office of Environmental Programs, Maryland
Mary DoIan
Baltimore City
David Carroll
Baltimore City Department of
Planning
Stuart May
Dept. of Natural Resources, State of
.Maryland
Harold M. Cassell
Dept, of Natural Resources, Maryland
-------
Ajax Eastman
State Water Quality Advisory
Council
Daryl Tuckey
Bethlehem Steel Corporation
Bill Burgess
Department of Natural Resources,
Maryland
John Hobner
Baltimore County Health Department
Edwin C. Weber
Department of Natural Resources,
Maryland
Tom Kusterer
Office of Environmental Programs,
Maryland
Bernard Bigham
Office of Environmental Programs,
Maryland
Mark Farfel
Johns Hopkins University
-------
setting, roost issues would not be further assessed after Phase I.
To invite numerous members of the public to become involved in
the project and to ask them to spend much of their uncompensated
time and energy in screening all 32 issues only to have the issue
of most interest to them not be selected for further analysis was
viewed as unfair and frustrating.
The TAG therefore decided first to select the Phase II
issues of most importance, and then to invite members of the pub-
lic to- participate in the workgroups established for each Phase
II issue. Questions of the scope of the study, objectives, and
results could then be adequately addressed.
\
In Chapter IV, we present the 32 issues identified and an
overview of the methods used to set priorities among the issues.
We also discuss how these methods were developed.
III-8
-------
CHAPTER IV
OVERVIEW OF THE PRIORITY-SETTING PROCESS
-------
IV. OVERVIEW OF THE PRIORITY-SETTING PROCESS
This chapter discusses the procedures the Management Committee
and Technical Advisory Committee used to set priorities among envi-
ronmental issues in the Baltimore area and to choose issues for
more detailed analysis in the second phase of the IEMP. The impor-
tance of this priority-setting exercise extends beyond the relative-
ly narrow purpose of defining study topics for this particular
project. The techniques described here are applicable to ongoing
environmental decision-making in the Baltimore area and they could be
applied to other metropolitan areas as well.
N Decisions about resource allocation for environmental protection
are always made within the context set by statutes, public concern,
and limitations of scientific data and understanding. The decisions
made in the Baltimore project were no exception. Within that frame-
work, the primary criteria for the making of such decisions require
the best scientific evidence that governments can bring to bear on
particular problems. Often these data are scarce, poor in quality,
or unavailable. Priority-setting for the purpose of resource
allocation is thus an exercise in decision-making under uncertainty.
V-
It is important to keep in mind that the priority-setting
for the Baltimore IEMP was conducted primarily on the basis
of available information, supplemented by data from a brief ambient
air monitoring effort conducted by the EPA. Although EPA and the
state and local governments who managed the project used scientific
information, the priority-setting was not a strictly scientific
enterprise. Rather, it was an exercise in policy analysis using
analytic tools; reasonable assumptions, where gaps in data existed;
and scientific reasoning. Products included a series of policy
analyses and a set of issues for detailed investigation in the
second phase of the project. It relied heavily on the scientific
and technical judgment of the officials who sat on the project's
committees. And where the project lacked data or objective measures
against which to compare issues and problems, the members attempted
to make their valuations reflect local social concerns and priorit-
ies.
The process for setting priorities developed by the participants
of the Baltimore project straddles the distinction that EPA has
often drawn between risk assessment and risk management. Scientific
considerations played a central role in determining the project's
priorities, but the committees also considered policy and economics
in making their judgments. For the most part, however, the distinc-
tion between use of science and basing decisions on policy was
preserved by the use of "primary" decision criteria which heavily
relied upon scientific data and "secondary" criteria which drew
on pragmatic considerations by local officials regarding the best
use of study resources.
IV-1
-------
GENERAL STRUCTURE OF THE PRIORITY-SETTING PROCESS
The priority-setting process in the Baltimore IEMP which the
TAG developed and implemented (illustrated in Figure IV-1) comprised
the following steps. First, the TAG members defined the geographic
boundaries of the study and set some preliminary, and flexible,
rules regarding its scope. Second, the TAG identified thirty-two
potentially important environmental issues, drawing heavily upon
the members' experience and knowledge of potential problems. tThe
list is shown in Figure IV-2.] Third, the committee agreed on the
use of three separate measures of environmental degradation by
which to evaluate the severity or significance of the thirty-two
issues. These measures (or primary selection criteria)human
health risk, ecological impact, and ground-water resource impact
also would serve as the "common denominators" to define a set of
reference categories into which each of the thirty-two issues would
be placed. The TAG established subcommittees to apply or, where
necessary, to develop the appropriate methodologies for evaluating
the issues in terms of the three measures. It then placed each
issue into the reference category in which the members deemed the
issue most appropriately belonged.
Drawing on their expertise and the analytic resources provided
by EPA, each of the subcommittees assessed the appropriate subset
of the thirty-two issues. [For example, the ecological impact sub-
committee assessed only those issues clearly pertaining to ecological
issues.] From time to time during their application of the three
methodologies, the subcommittees requested EPA to gather more
information or to analyze issues for which the scientific data were
particularly ambiguous.
Following this work, each subcommittee reported to the TAG on
the most pressing problems as defined by primary criterion used by
the particular subcommittee. The TAG as a whole consolidated and
regrouped the topics, and applied a set of secondary criteria to
each of them based on economic and policy considerations. Finally,
the committee drew up a workplan for the project's second phase for
each of the five "study topics" that had emerged from the priority
-setting process.
The charge of each subcommittee was to set priorities among the
issues and recommend before the full TAG the issues warranting
further study in Phase II. To do so, the TAG believed that they must
reduce the list of possible topics from thirty-two to about ten or
eleven. The TAG intended to use meetings of the full TAG to arrive
at a consensus about the most important problems and thus reduce
the number of study topics to about five, the number that the
project's managers felt could reasonably be analyzed in detail
during Phase II of the project. Because of overlaps among the issues
recommended by the various subcommittees, this second step in the
priority setting process turned out to be less difficult than
originally thought (See Chapter VIII for a more complete discussion
IV-2
-------
Figure IU-1
Baltimore IEMP
Priority-Setting Process
Selection of
Study Issues
Selection of
Secondary
Criteria
Definition of Project Scope
and
Geographic Boundaries
Rankiny against
Secondary Criteria
fr Selection of
Phase II Issues
Deuelopment of
Phase II UJork plans
fr Budgets
Redefinition of
Study Issues
Rankiny
ayainst
Primary
Criteria
Re definition
and
Re ran king of
Metals Issues
Negotiation)
of IVorkplan »
Rudyets
Selection of
Primary
Criteria
Data-
Gathering
(by EPfl)
I
Deuelopment of
Priority-Settiny
Methods
Recommendations on
Phase II Issues,
UJorkplans, ff Budgets
-------
of this.).
The Management Committee, although responsible for the
overall direction of the study, gave the TAG members broad authority
in devising and implementing this strategy. The MC reviewed the
list of thirty-two issues, the decision to use. three primary criteria
and their associated analytic methodologies, and the ultimate
decisions on priorities that arose from the process. The chairman
of the TAG attended nearly every meeting of the Management Committee
to discuss the analytic work of the study and the policy implications
of the analysis. The Management Committee1 s role was not exclusively
one of delegation and oversight; the committee took the lead in
matters of public participation in the study, and also coordinated
external reviews of the project's work.
Consensus and professional judgment played critical roles
throughout the priority-setting exercise, both for the Management
Committee and for the TAG. Neither committee, as a rule, took
formal votes,, preferring instead to debate issues until a consensus
became clear. The consensus was not always unanimous, but both
committees were able to arrive at decisions without using formal
mechanisms like voting. Partly because of this open structure and
partly because of limitations in the data and in scientific under-
standing of environmental problems, the MC and the TAG relied
heavily upon best professional judgment to overcome obstacles to
agreements.
PRIORITY SETTING IN PRACTICE
The priority-setting procedure which we described in the preced-
ing section did not exist at the outset of the process. The committee
members developed the methods as they proceeded, taking into account
the particular environmental problems of the area, the project
budget and time constraints, and limitations in the data base. The
committee tried several different ways of setting priorities before
arriving at the methods it ultimately used. These different schemes
are discussed in more detail below. Also the chronology of decisions
did not happen as one might infer from the description of the
priority-setting process. Furthermore, the distinctions between
various stages of the priority-setting process were, in practice,
not clearly drawn. Some issues regarding the analytic scope of the
study (e.g., the inclusion of certain pollutant classes) were not
finally resolved until the end of the project's first phase.
Similarly, the secondary criteria of pragmatic considerations,
rather than being applied all at once at the end of Pha-se One,
actually played a role from time to time throughout the analysis
described in this report. This again was a product of the fact
that the IEMP was not a formulaic academic exercise but a real-world
attempt to set priorities and solve problems. Adherence to a more
rigid decision-making structure would probably have reduced the
amount of debate on these issues, but in doing so it would have
impeded the creation of informed consensus on which the success of
the project rests.
IV-3
IIP
-------
PRELIMINARY DEFINING OF THE SCOPE OF THE STUDY
When EPA originally approached the state and .local govern-
ments in the Baltimore area with the idea of conducting an IEMP,
it had in mind a project devoted primarily to the study of toxics,
focusing primarily on human health effects. Initial discussions
among the various governments strengthened EPA's decision to focus
on toxics. The major considerations were that conventional pollu-
tants (e*g. biological oxygen demand in water, oxides of nitrogen
and sulfur in air, etc.) had already been well studied in the study
area, and IEMP was unlikely to be able to contribute significantly
to their analyses or to strategies for their control. Accordingly,
the decision not to consider conventional air pollutants was made
by consensus of the various parties before the Management Committee
and TAG ever formally met.
The same discussions, however, led EPA to reconsider the
notion of concentrating primarily on human health effects; from
the outset of the project, state and local officials were convinced
that the project needed to include ecological effects as well in
order to be useful. Because of their importance on marine ecosystems,
conventional water pollutants were included in the studyagain by
consensus before the establishment of the committees.
The governments also agreed at this stage to exclude radio-
nuclides and pathogenic micro-organisms, episodic events such as
spills and sewage treatment plant upsets, occupational exposures,
and exposures through the food chain (except those involving locally
harvested seafood). The reasons, in each case, were that the data
were so limited that no useful analysis could be expected within
the time and budget constraints of the project, that none of the
governmental agencies involved in the study had any jurisdiction
over the problems, or that the issues were such that the site-specific
methodologies used in lEMPs were not appropriate for the analysis.
An example of the last case is the food chain: analysis of exposures
to toxics through food would require consideration of the interna-
tional food supply and distribution network and thus seemed an
inappropriate topic for the IEMP to consider.
Some of these decisions on exclusion of certain pollutants,
sources, or exposure pathways were debated again after the TAG
began its deliberations, although none was changed as a result of
those discussions. The decisions on the scope of the study were
not formally made by the TAG until a considerable time after the
project began.
Finally, another important decision that had to be made at
the outset of the study concerned the geographic boundaries of
the study. Pollutants are 'bound to cross arbitrary boundaries.
The problem was this: to define a small enough area such that the
IV-4
-------
project could contribute to the understanding of its problems
without making the area so small that pollution "imports" and
"exports" have a greater impact on the area than the effect of
locally generated pollution.
EPA consulted with the state government to arrive at tenta-
tive boundaries which included the city of Baltimore and its
immediate suburbs. It then discussed these proposed boundaries
with the local governments. [The boundaries that resulted from
the process are shown in Figure II-l in Chapter II.] The major
issue in the selection of boundaries was whether the suburban
areas in eastern Howard County should be included in the study.
After consulting with Howard County, State, and Regional Planning
Council officials, EPA decided not to include those areas. The
sources of possible pollution were not seen to be as concentrated
in these areas as in the other two counties or the city. Every-
one Jagreed to reconsider the decision should a pollutant source
in that jurisdiction become important to the project. Also agreed
to was the decision to include within the project scope of any
area in Anne Arundel or Baltimore Counties which might later be
deemed important to completion of the project. [For issues con-
cerning underground storage tanks, all of these two counties would,
in any case, be included.]
SELECTION OF THE INITIAL LIST OF PHASE I STUDY ISSUES
After having established the geographic boundaries for the
study and deciding to focus primarily but not exclusively, on
toxic pollutants, the Technical Advisory Committee drew up a list
of thirty-two topics for preliminary screening. In doing so, they
relied heavily on their previous experience with pollution problems
and on professional judgment.
The development of the list took many months. The TAC's
chairman asked each member of the committee to identify at least
one issue that he or she thought was especially worthy of considera-
tion, and at least one other issue whose importance had been under-
estimated in the past. Each member was explicitly instructed to
indicate whether or not the issues were in the individual's area of
expertise or under the jurisdiction of his or her agency. Several
committee members raised more than two issues, often from outside
their own agencies' jurisdictions. The chairman himself did not
raise any issues. EPA staff compiled the lists and used them to
develop the list of thirty-two that ultimately emerged.
This process was neither designed nor ever intended to'inden-
tify systematically every environmental problem; no process could
do so given data limitations. There may be other, more severe
problems which have not been acknowledged because data on expo-
sure, toxicity, or both are absent. This is a pervasive problem
in environmental or public health protection, one which will never
be completely solved. For this reason, it is impossible to state
IV-5
-------
that the thirty-two issues are a comprehensive list of the most
pressing environmental problems in Baltimore. The process estab-
lished by the Baltimore IEMP will, however, put agencies in a
better position to appraise and solve any new problems that emerge
as additional data come to light.
The issues raised by the TAG can be grouped into three broad
categories: 1) issues defined by pollutants, 2) issues defined by
source or source category, and 3) issues defined by environmental
medium or exposure route. The list of issues ultimately deve-
loped by the TAG is shown in Figure IV-2.
The MC reviewed the list and concluded that it covered the
most important environmental issues in the study area. It could
thus serve as the basis for the priority-setting that was to
follow. The Management Committee's review was principally con-
cerned with making sure that no potentially important problem
known to them had been omitted and that issues would be of area-
wide concern. The fact that the MC proposed no additions to the
list is evidence of the degree of consensus that exists among
governmental officials about the principal environmental problems
of the Baltimore area.
METHODS FOR SETTING PRIORITIES
While developing the list, the TAG made two preliminary
efforts to set priorities among the issues it was identifying.
After the committee had prepared an early version of the issue
list, its chairman consolidated the issues into four main groups:
toxic air pollution, indoor pollution, drinking water (including
groundwater), and the Baltimore Harbor. Hazardous waste was not
included as a separate issue group, but various impacts attribu-
table to hazardous waste (e.g., on air toxics or water quality in
the harbor) were included in the appropriate group.
The chairman then asked each member to consider three crite-
ria for priority-setting: human health effects, ecological effects,
and economic impacts. He asked each of the committee members to
rate each of the four issue groups as having high, medium, or low
importance when considered according to each criterion. Thus, for
for example, a committee member might say, "Toxic air pollution
has high importance for human health, but low importance for eco-
systems. It has medium importance in terms of its economic im-
pacts. Indoor pollution is of low importance under all three
criteria," and so on.
While this method revealed a great deal about the judgment
of the various committee members, it did not result in consensus
on the importance of issues. Each of the groups was considered
by most members of the committee to be extremely important under
at least one of the criteria. Further, committee members dis-
agreed about the relative importance of the criteria themselves
The issue groups were broad enough that each included many in-
IV-6
-------
FIGURE IV-2
THE INITIAL LIST OF 32 ISSUES
CONSIDERED IN THE BALTIMORE IEMP
ISSUES SPECIFIC TO A POLLUTANT OR POLLUTANT CLASS (8)
ACID RAIN
ASBESTOS
BENZENE
MULTIMEDIA
METALS
PESTICIDES
Acid rain is acid precipitation in the form of
rain.1 Transformation products of oxides of
nitrogen (NOx) and sulfur (SOx), which are emitted
into the air through combustion of fossil fuels by
industrial facilities or even cars and trucks,
render precipitation more acidic. Acid rain
threatens damage to lakes, forests, soils, and
crops.2
Asbestos is the popular name for fibrous minerals
used for fireproofing and heat insulating
material.3 Exposure to asbestos has been linked
to a respiratory disease called asbestosis and to
lung cancer.
Benzene, a volatile aromatic hydrocarbon proven to
be a human carcinogen, is an important industrial
solvent and constituent of gasoline.
Large amounts of metals are used in industrial
processes in the Baltimore area. Unlike organic
compounds, metals do not decompose in the
environment and are found in all environmental
media. Some metals, like lead, have been used in
household materials and are known to contaminate
drinking water and the indoor environment.
Pesticides, especially herbicides, are used in
farming and smaller-scale operations, such as golf
courses and homes and gardens. Pesticides in
agricultural and urban run-off may contaminate
surface and ground water. Aerial spraying may
contaminate the ambient air.
IPOLYCHLORINATED
BIPHENYLS Polychlorinated biphenyls (PCBs) are a highly
toxic, persistent group of organic compounds that
may pollute all environmental media. Their
primary use, until their banning by Congress, had
been as insulating material in such items as
electrical transformers.
-------
TITANIUM Titanium tetrachloride is produced by the
TETRACHLORIDE atmospheric transformation of titanium dioxide, a
white, metallic pigment inadvertently emitted into
the ambient air in manufacturing processes. Clouds
of titanium tetrachloride can be both a traffic
hazard and a respiratory irritant.
VINYL CHLORIDE Vinyl Chloride is a volatile chlorine organic
compound that is used in large amounts in the
production of polyvinyl chloride (PVC). It is a
known human carcinogen.
ISSUES SPECIFIC TO A SOURCE OR SOURCE CATEGORY (16)
SEWAGE TREAT-
MENT PLANTS Sewage treatment plants often receive large
quantities of volatile organic compounds, heavy
metals, and other potentially toxic pollutants
from industrial facilities and commercial users.
These pollutants can be released into the ambient
air, surface, or ground waters.
LEAKING UNDER-
GROUND STOR- Chemicals, such as gasoline, contained in
AGE TANKS underground storage tanks (USTs) can leak out and
contaminate ground-water or even cause explosions.
Of particular concern are the thousands of tanks
connected with automotive service stations.
INCINERATION
OF SEWAGE
SLUDGE
SANITARY
SEWER OVER-
FLOWS
"Toxic substances contained in sewage sludge may be
emitted into the air during incineration.
During periods of high- rainfall, the capacity
of the sewer system is not sufficient to accept
all sewage. Some sewage may flow directly to the
harbor or its tributaries.
SURFACE IM- Surface impoundments may release pollutants to the
POUNDMENTS FOR ground-water. Impoundments that contain volatile
HAZARDOUS & compounds may release them to the air.
NON-HAZARDOUS
WASTES
SEPTIC TANKS
Septic tanks can release large amounts of nitrogen
to the ground water. Nitrogen may form nitrates, a
potential health risk to children from drinking
water. Septic tanks that contain cleaning
solvents may also release these to the ground-
water. Pathogens, which can also move from septic
tanks to ground water, were explicitly excluded
from consideration.
-------
FEEDLOTS
AIR AND WATER
POLLUTION .
FROM FARMING
NON-POINT
WATER DIS-
CHARGES
LANDFILLS
COMBUSTION &
INCINERATION
POWER PLANTS
EMISSION OF
GRAIN FUMI-
GANTS FROM
SHIPS AND
GRAIN ELE-
VATORS
WASTE-WATER
DISCHARGES
FROM SHIPS
HARBOR SEDI-
MENTS
Feedlots, that is, areas where large numbers of
animals are raised and fed, produce great amounts
of animal waste. These wastes, in turn, can
pollute ground or surface waters.
Air and water pollution from farming can include
pesticides in the ambient air from aerial spraying
of pesticides and nitrogen fertilizer in
agricultural run-off.
Stormwater or agricultural runoff can carry high
levels of pollutants to surface waters, such as
streams or the harbor.
Municipal and hazardous waste landfills may
release pollutants to ground-water or surface
waters. Volatile organic compounds contained in
the sites can be released into the air.
Combustion and incineration can release a variety
of toxic compounds into the air. Examples of
where combustion and incineration occurs are wood-
burning stoves, cars and trucks, sludge
incineration, and hazardous waste incineration.
Power plants emit SO2 and heavy metals
to the air and a large variety of compounds to.
surface waters.
Pesticides, particularly grain fumigants, which
are used to kill insects that feed on grain, may
be released and thus pollute the ambient air.
Ship sewage and contaminated water used to clean
ship bilges may be released and thus pollute the
harbor waters.
Pollutants in the sediments of the harbor may harm
organisms that live in or close to the floor of
harborparticularly if the sediments should
become disturbed, such as through dredging.'
-------
ROAD SALTING
Salts from the salting of roads may pollute ground
or surface waters.
ISSUES SPECIFIC TO AN ENVIRONMENTAL MEDIUM OR EXPOSURE ROUTE (8)
TOXIC AIR
POLLUTION
WATER DIS-
CHARGES TO
THE HARBOR
Toxic chemicals may be released into the ambient
air from factories, cars, trucks, and even
neighborhood dry-cleaning establishments. As used
here, the term does not refer to "conventional"
pollutants, such as oxides of nitrogen or sulfur,
carbon monoxide, or ozone.
This issue covers both point and non-point source
discharges to the harbor.
SURFACE DRINK- Drinking water for most of the population in the
ING-WATER study area comes from surface waters. Quality is
QUALITY defined in terms of the concentrations of various
pollutants, including disinfectant by-products,
in the drinking water.
WATER QUALITY
OF CITY
STREAMS
Pollution in city streams may pose risks to both
the biota inhabiting the streams and people who
use the streams for recreation.
BIOACCUMULA- Certain pollutants may accumulate in
TION OF TOXICS fish, crab, and other biota inhabiting the surface
IN FISH waters of the study area. These fish and crabs
may, in turn, pose health risks to humans,
particularly should they be eaten in large
quantities.
EFFECTS OF
POLLUTANTS ON
AQUATIC
ORGANISMS
CHROMIUM IN
THE HARBOR
AREA
Pollutants in the harbor may harm aquatic
organisms.
In past years, large amounts of chromium-
containing waste were used as fill in areas
surrounding the harbor. It is unknown to what
degree this fill has leached chromium to the
harbor water and sediments.
-------
INDOOR AIR Pollutants from a wide variety of sources indoors
PpLLUTION can degrade the indoor environment. The
pollutants include many of the same chemical
compounds of concern in the ambient air, such as
lead, formaldehyde, chloroform, and the products
of incomplete combustion (PICs). A chemical of
unique concern to the indoor environment is radon.
Webster's Ninth New Collegiate Dictionary. Merrriam-Webster,
Inc., Springfield, Massachusetts (1983), p. 52.
Council on Environmental Quality, Environmental Quality-
1980. p.4.
Van Nostrand's Scientific Encyclopedia, fifth edition, Van
Nostrand-Reinhold Company, New York (1977), p.198.
-------
dividual issues, some of which were seen as important, others
as not so important. Finally, the committee members felt .that
they did not have sufficient information to make - judgments a-
bout the importance of various issues, especially for topics
outside their professional expertise.
The committee, therefore, adopted a second approach. It
requested EPA to write short issue papers describing what was
known about each of the 32 issues as background information for
its deliberations. For reasons of time and resources and also
because the importance of a number of issues required no further
elucidation, the EPA prepared drafts for only about a third of
the issues. Upon receipt of these, the committee requested revised
versions of most of the papers, often with specific suggestions
about further information that could be included. The committee
members also pointed out the need for making explicit the assumptions
in the papers, explicitly identifying the populations at risk,
fully documenting analytic methods and sources of data, providing
maps, and preparing the papers in a consistent format. Because of
budget constraints, none of the papers were ever revised upon
subsequent review by the committee members. Nevertheless, comments
on these papers were taken into account in later analyses. A
sample issue paper is provided in Appendix B.
V
Drawing upon the information in the issue papers and what
was generally known about the other issues, the committee attempted
to assign each of the 32 issues to one of four categories: (1) good
data indicating high risk, (2) apparently high risk, but with weak
data, (3) apparently low risk, and (4) unknown. Risk in this
context included both health and ecological risk, as well as risk
to resources such- as aquifers and bodies of surface water.
This approach, like the first, increased understanding of
the issues, particularly those which lay outside each committee
member's expertise. Like the first approach, however, it did not
lead to a consensus on the relative importance of various issues.
The committee found it difficult to compare problems with little
data with problems for which data were abundant, and the "unknown"
category presented further problems. The committee also did not
try to agree about the relative importance of the various criteria
(human- health, ecology, and economics) employed.
These early attempts at priority-setting illustrate three
more general problems that will inevitably arise when any group
tries to set priorities and allocate resources among a set of
problems this large and diverse. The first, and by far the most
important, is that different problems are significant for different
reasons. If problems differ in degree (i.e., severity) but not in
kind (i.e., the nature of their adverse impacts), and ir reasonable
information exists on the severity of the problems, priority-setting
is relatively straightforward. If one substance is predicted to
cause 1000 cancer cases annually and another one case, deciding
IV-7
-------
which is more important is not difficult. There is, however, no
non-controversial way to compare a pollutant that is present in air
and causes cancer in humans with one that affects only fish.
Comparison of unlike objects and issues involves value judgments
and cannot be done purely on the basis of science.
Previous and concurrent lEMPs had not directly .confronted
the tradeoffs between health and ecological problems. The Philade-
lphia study concentrated solely on human health because EPA felt
when it initiated the project that it lacked the methodology and
the resources to analyze both health and ecological problems. The
IEMP in Santa Clara County, California, concentrated on health
issues because of a widespread consensus that those issues were
most pressing.
In Baltimore the consensus was the opposite: that human
health was not the only concern that the IEMP ought to address.
The TAG in Baltimore felt thatv effects of pollution on ecosystems
and the ground-water resource also deserved attention. To solve
the problems arising from the lack of any non-controversial method
for making tradeoffs between health and other environmental problems,
the TAG developed and applied three separate methodologies for
evaluating the impacts caused by each of the potential issues: one
for human health, one for ecological damage, and one for the resource
impact caused by ground-water contamination. The TAG did not make
any attempt to compare directly the impacts measured by each of the
three methodologies. Instead, it used the methodologies to define
potential serious health, ecological, and ground-water problems,
and agreed to conduct further research on the problems in each
category that appeared to be the most severe.
The choice of the three measures of environmental degradation
human health, ecology, and resource impactarose by consensus of the
TAG. The twin mandates to protect human health and the environment
(i.e., ecosystems) are widespread in both Federal and state statutes.
More recently, especially in the light of growing concern about
hazardous waste, enviromental agencies have turned their attention
to impacts caused by pollution to ground-water resources. The
economic costs of switching to alternative sources of drinking
water if ground-water should become contaminated has led to increased
governmental attention to ground-water protection. On the other
hand, ground-water contamination has at most an indirect effect on
the health of ecosystems. The TAG considered , including a fourth
measure, the costs and monetary impacts of pollution, but'decided
to defer most considerations of cost until the second phase of the
project. Thus, the criterion of economic impact is replaced by
"impact to ground-water resources."
The Management Committee reviewed and approved the decision
to include the three impacts proposed by the TAG. The decision
to examine other measures of environmental degradation in
IV-8
-------
to human health risk was an important one: the inclusion of these
other measures is one of the principal factors distinguishing the
scope of the Baltiore IEMP from that of the first phase of other
lEMPs.
Once the three measures were chosen, the TAG chairman selected
TAG members to head each of the three subcommittees. Each TAG
member then volunteered to serve on at least one of the subcom-
mittees. Each subcommittee was charged with developing and applying
a methodology for ranking the issues according to their importance
against one of the primary criteria.
Each of the subcommittees then analyzed only those issues
that it felt to be relevant to the type of impact under its
examination. These included issues for which they had to postulate
effects because they did not have actual data on exposure or toxi-
city. For example, though surface impoundments may affect ground-
water and ground-water contamination may ultimately find its way
into the harbor, no data were available on the timing or magnitude
of these effects.
The human health subcommittee evaluated about two-thirds of
the thirty-two issues. The results of these analyses are described
in Chapter V. Those not evaluated were clearly of primary import-
ance to the aquatic life of the harbor. The ecology and ground-
water subcommittees each evaluated about half of the issues.
Many of the issues were analyzed by more than one subcommit-
tee. For instance, chromium in the harbor was examined by both
the ecology and ground-water subcommittees. Bioaccumulation of
toxics in fish was considered by both the human and ecological
impact subcommittees* . Septic tanks were examined by both the
human health and ground-water subcommittees. Metals was the only
issue studied by all three.
The TAG did not intend, by splitting its analysis into three
subcommittees, to deny the existence of tradeoffs among various
environmental goals. Its intention, instead, was to defer explicit
consideration of those tradeoffs until it had developed more data
on individual issues and had determined, to its satisfaction, the
relative importance of the various issues within each reference
category with regard to the appropriate measure.
It also felt .that the value judgments necessary to determine
the relative importance of these different types of environmental
impacts could be made at a later step. Since the TAC's role was
only advisory, the MC should ultimately make these decisions.,
A second major problem in priority-setting, which 'the TAC's
early efforts also illustrate, is that the information available to
the committee on each problem varies enormously in reliability
and in sheer bulk. For example, emissions from power plants have
IV-9
-------
been exhaustively studied. On the other hand, much less informa-
tion is available on discharges to ground-water that may occur
from surface impoundments. Similarly, the carcinogenicity of
benzene has been confirmed by numerous epidemiological and laboratory
studies, while virtually no information is available on the toxicity
of titanium tetrachloride. Clearly the degree of certainty about a
problem should be factored into the prioritysetting progress, but a
great deal of honest 'disagreement exists about how to take that
uncertainty into account.
All three of the TAC's methodologies included appraisals of the
degree of certainty associated with the analysis of each problem.
These appraisals were qualitative and based on the best judgment of
the committee members. They were not incorporated into the analysis
in any quantitative way. To do so (for instance, by multiplying
quantitative estimates to damage by a numerical surrogate for
certainty) would have itself assumed a degree of certainty which
does not in fact exist.
The third major problem in,priority-setting is that the pollu-
tant-specific, source-specific, and medium- or exposure route-specific
problems are bound to overlap somewhat. For example, power plants
(listed as a source-specific problem) contribute both to acid rain
(a pollutant-specific problem) and water discharges to the harbor
(a medium-specific problem). This sort of overlap is inevitable,-
-indeed, it is at the heart of the IEMP approach but, it makes
priority-setting difficult because the severity of a particular
problem can depend on how it is defined.
To address this problem; the TAG and EPA separately analyzed
each of the thirty-two issues, ignoring overlaps existing among
them. Through the use of a common metric (such as cancer risk) for
issues in a reference category, the severity of a pollutant specific
problem (e.g., benzene) could be compared directly against that of
a source-specific problem (e.g., sewage treatment plants), even if
some overlap exists. Similarly, comparing ambient levels against
one another, as the ecological impact subcommittee did, implicitly
includes the cumulative effects of all sources. Clearly, if the
results of the study were used to estimate the overall magnitude of
environmental problems (measured in terms of human health, or
ecological damage, or ground-water impact), the numerical measures
of severity could not simply be combined, but would have to be
corrected to account for the overlap among problems.
The TAG chairman directed each of the subcommittee (chairmen)
to recommend issues for further analysis in the project's second
phase. The human health subcommittee was asked to recommend up to
five issues and each of the other subcommittees was asked to recom-
mend up to three. In other words, the TAG intended to use the
subcommittees to reduce the number of possible study topics form
thirty-two to eleven, to apply the 'secondary criteria to those
IV-10
-------
eleven topics, and then reduce the list further in full meetings of
the TAG. Once this method of priority setting was agreed upon, the
TAG and its subcommittees worked with^considerable speed. Develop-
ment of the methodologies described below took about six weeks and
the application of the methods to the ranking of the 32 issues
another month.
The first and second steps worked as planned: the health
subcommittee recommended five issues, the ecological impact subcom-
mittee three, and the ground-water subcommittee two. One topic
trihalomethanes in drinking waterwas eliminated because it did
not meet one of the secondary criteria. The last step turned out
to be unnecessary because of overlap among the study topics recom-
mended by the various groups (See Chapter VIII.).
RANKING OF ISSUES AGAINST THE PRIMARY CRITERIA
In assessing the impact of each of the thirty-two potential
problems on human health in the Baltimore area, the TAG used
standard quantitativerisk assessment models which were developed
by EPA and which have been used in other lEMPs. The models have
been used expensively in a number of other applications, as well,
such as in the development of regulatory standards. They are
currently the subject of a scientific peer review. We discuss
them fully in Appendix A of this report and describe their
application in the Baltimore IEMP in Chapter V.
The models are conceptually simple. Ambient monitoring data
or, where these do not exist, estimated ambient levels based upon
source emission estimates, are used in conjunction with exposure
factors to estimate human exposure to each substance under study.
These estimates of exposure in turn are combined with toxicolo-
gical estimates of potency to yield quantitative estimates of
individual risk. These are expressed as the incremental proba-
bility of disease incidence (not death) that would conservatively
be expected to result from that exposure. Combined with data on
population densities, the information on individual risk can be
extrapolated to yield numerical estimates of disease incidence in
the population attributable to exposures to each pollutant.
The models are deliberately designed to yield conservative
estimates (i.e., to be protective of public health) both of
individual risk and of aggregate disease incidence to average
ambient values for the pollutant in question over long periods of
constant exposure. Their primary usefulness is thus in setting
priorities and allocating resources rather than in predicting
absolute risk. For each of the pollutants analyzed, the methodol-
ogy is much more likely to overstate risk than to understate it.
However, the models do not take into account intermittent
peak levels of pollutants to which an individual may be exposed
during his or her daily activities and which, through dilution in
IV-11
-------
the ambient air, are not well represented by ambient average daily
concentrations. [EPA's Total Exposure Assessment Methodology (TEAM)
projects attempt to identify these exposures through what they term
"personal" monitoring. We describe this method in greater detail in
Chapter IX.] Where these are expected to occur frequently, the
models may be of limited use.
The human health subcommittee chose aggregate expected in-
creases in disease incidence as the primary criterion for setting
priorities among potential health problems. In addition, EPA
calculated for the subcommittee's consideration risks incurred by
individuals at sites of. the highest monitored values for specific
pollutants. While maximum individual risks can be important, the
subcommittee felt that aggregate disease incidence was a better
.criterion for general priority-setting since it takes into account
the numbers of people potentially affected by each environmental
problem.
The methodology ignores the possible synergistic or antagonistic
effects of simultaneous exposure to more than one pollutant. With
few exceptions, scientific understanding of those interactions is
so limited that we cannot state with any confidence how much, if at
all, risks from those simultaneous exposures might vary from the
sum of the individual risks. The toxicological scores used by EPA
in general and the IEMP in particular are all substance-specific
and do not take into account any possible interactions among pollu-
tants.
The TAG applied its qualitative estimates of certainty and
strength of evidence to each problem analyzed from the perspective
of human health. The committee characterized the information which
included source characteristics (identity of the sources, release
rates, and durations), ambient levels, and dose-response functions,
as being "good," "fair," "poor," or "unknown."
The TAG did not consider epidemiological data specific to
Baltimore in its appraisal of human health risks. The human
health risk analysis overseen by the subcommittee was an exercise
in prospective risk assessment that attempted to estimate the risks
of contracting diseases in the future which are attributable to
current exposures. By contrast, epidemiological studies show (when
they are successful) the results of past exposures, the health
effects of which may lag the actual exposures by many years-. (A
hazard evaluation of a chemical does, of course, consider epidemio-
logical studies as the basis for judgments about its toxicity.)
The health subcommittee recommended five issues to the full TAG:
trihalomethanes in drinking water, toxic air pollution, benzene,
metals in air, and indoor pollution. More detailed information on
their methodology and deliberations is provided in Chapter V.
IV-12
-------
To measure e co1og ica1 damage, the TAG, nor anyone else for
that matter, did not have well-developed and widely used models
of the sort that have been available for assessing human health
risk. The TAG subcommittee charged with developing methods for
ecological effects relied on surrogate measures for damage to the
Baltimore Harbor and the Chesapeake Bay. EPA examined monitoring
data on sediments and the ambient water, and compared those data
to the Water Quality Criteria (ambient standards set by the
Agency's Office of Water). By constructing simple ratios of
measured concentrations to allowable concentrations, the TAG
subcommittee compared various pollutants. Ratios of greater than
one were considered to be indicators of possible problems. As
with the human health analyses, the subcommittee characterized
their confidence in the underlying data through the use of a
qualitative ranking system.
Though this comparison of ratios does not yield actual
estimates of ecological damage, it does serve the purposes of the
screening analysis. It provides a means for ordinally ranking
specific pollutants. Furthermore, since the EPA standards take
into account the most recent findings on the levels at which
pollutants begin affecting aquatic life (i.e., ecological
thresholds above which ecological damage increases with pollutant
level), the ratios allow a rough estimation of the magnitude of
the effects.
The method does however ignore (as does the human health
methodology) possible synergistic effects of exposure of biota to
more than one pollutant. Little is known about such effects, and
the data that would be necessary to construct more realistic
dose-response curves do not exist.
Besides not measuring actual ecological damage, the method
does not provide a way .to compare ecological damage to the human
health risks discussed above. The extent of actual ecological
damage attributable to pollution in the Harbor is not well
understood, and the relative importance of that damage compared
to human health effects resulting from other forms of pollution
is a . matter of values. These issues were discussed in the
meetings of the TAG and decided, implicitly or explicitly, in the
committee's decisions about resource allocation and priorities
(budgeting).
The ecological impact subcommittee applied its methodology
only to water pollution problems. Effects of air pollution on
ecosystems are attributable primarily to conventional pollutants,
which had previously been excluded from the study (see above).
No ecological standards exist with which to compare the limited
ambient data on 'air toxics. Hazardous waste was considered in
the ecological study to the extent that it affects surface water,
but the TAG subcommittee did not attempt to isolate the effects
of hazardous waste on ecosystems because it felt that the data
were inadequate for this purpose. Nor did it examine ground
water from an ecological
-------
. The ecological impact subcommittee recommended to the full
TAG that toxic metals in ambient water, previously contaminated
sediments as a source of water contamination, and bioaccumulation
of toxics in aquatic organisms be studied further in Phase Two.
A full discussion of the ecological methodology and its applica-
tion is in Chapter VI.
Finally, the ground-water resource subcommittee developed a
methodology for evaluating the impact on ground-water resources
which are caused by releases from various categories of sources
(e.g., landfills, surface impoundments, and underground storage
tanks). The subcommittee agreed on a list of criteria which
included considerations of both physical and economic impacts.
Included as evaluative criteria for physical impact were source
characteristics (e.g., number of facilities in each source
category), release volume, concentration of contaminants in the
potential releases, environmental persistence of the contaminants,
likelihood of contamination, and potential extent of damage.
Economic considerations included the likely costs both of prevention
and of response. Each criterion was weighted equally.
The subcommittee members then ranked each source category
against each criterion on a scale of 1 (not severe or high) to 5
(very severe or high). Thus, for example, underground storage
tanks could receive a rating of 5 against the "source characterist-
ics" criterion because there are numerous underground storage tanks
in the area and because potential release volumes are large. After
each subcommittee member had individually evaluated the source
categories, the subcommittee met to reach a consensus on the import-
ance of each source category against each criterion. Through a
series of meetings, the subcommittee was able to develop a consensus
on the source categories which appeared to be most urgently in need
of further study.
The ground-water subcommittee recommended to the full TAG that
pollution of ground-water by metals and pollution from underground
storage tanks be studied further in the second phase of the project.
The subcommittee decided not to recommend a third issue for Phase
II study because it felt that the two problems it did recommend
were significantly more important than all the others. A detailed
discussion of the ground-water methodology and its application
appears in Chapter VII.
In summary, ten issues, which we list in Figure IV-3,' emerged
from the priority setting process. In the next three chapters, we
explain in greater detail how each priority-setting approach worked
and describe the results of the analyses that were conducted in
support of these approach. The application of the secondary criteria
to those issues, and the work subsequently done to analyze them, is
discussed in Chapter VIII.
IV-14
-------
Figure IV-3
Study Topics Recommended by the TAG Methodology
Subcommittees after the first Priority-Setting Round
Human Health Risk Subcommittee
Benzene
Trihalomethanes in drinking water
Multi-media metals
Toxic air pollution (low molecular weight organics)
Indoor air pollution
Ecological Impact Subcommittee
Metals
Sediments
Bioaccumulation of toxics in aquatic organisms
Ground-Water Resource Damage Subcommittee
Metals in ground water
Underground storage tanks
IV-15
' "Hi1 iui""ii" 'fi" ffll
-------
CHAPTER V
ANALYSIS OF HUMAN HEALTH RISKS
-------
V. ANALYSIS OF HUMAN HEALTH RISKS
Using the priority-setting exercise discussed in the pre-
vious chapter, the Technical Advisory Committee (TAG) sorted through
32 issues to select a subset of topics warranting further analysis
in Phase II. The topics were organized into one of three categories
(human health, ecological damage, and resource damage), and three
subcommittees were established to review and recommend candidate
issues in each of these areas. This chapter summarizes the analyti-
cal findings for the four issues that the Human Health Subcommittee
identified for further analysis in Phase II: organics in the
ambient air (including benzene); metals in the ambient air and
lead.in the general environment (especially lead in dust and water);
indoor air; and trihalomethanes in drinking water.
We present the results in this chapter in terms of the
risks to human health posed by each environmental issue. While
the risk assessment models are fairly straightforward, there are
numerous limitations that must be considered when interpreting
the results generated by these techniques. For that reason, the
first section of this chapter presents a brief overview of EPA's
standard methodology for assessing risks to human health and
discusses the major limitations associated with quantitative risk
assessment. A more thorough discussion of how we approach risk
assessment and its limitations appears in Appendix A.
The second section of this chapter summarizes our findings
on the risks to human health associated with each of the four
issues the Human Health Subcommittee selected. It is important
to note that the results presented are not based on exhaustive eva-
luations of each topic. The analyses performed in Phase I were de-
signed primarily to provide the members of the Human Health Subcomm-
ittee with sufficient information on the relative significance of
different environmental issues. Subcommittee members would use this
information to select topics warranting further investigation and
to set priorities for additional work in Phase II of the IEMP.
The third section provides a comparison of risks across pollu-
tants and routes of exposure. This chapter also provides a discus-
sion of the conclusions that can be drawn from our analyses of
human health risks in Baltimore and the implications that our
findings may have for Phase II.
RISK ASSESSMENT AND ITS LIMITATIONS; AN OVERVIEW
Using standard EPA methods and models, Phase I of the Baltimore
IEMP assessed the human health risk attributable to different
sources, pollutants, and exposure pathways in order to set research
and management priorities for selected issues.
V-l
-------
The purpose of this section is to provide the reader with a
brief overview of risk assessment and its limitations. A more
detailed discussion of risk assessment can be found in Appendix A.
Defining Risk Assessment
Risk assessment is a technique that allows one to estimate
what health effects may result from current and future' environ
mental exposures. While there is some controversy surrounding
EPA's risk assessment techniques, it is important to evaluate these
procedures in light of their main objectives: to permit a comparison
of one risk with another and to give a very general sense of the
risk a substance may present, rather than to make a definitive
statement concerning the absolute risk posed by a particular sub-
stance, pollution source, or exposure pathway.
It is important to understand that risk assessment does
not examine disease incidence in the local population and then
attempt to link it with environmental exposures. Because it is
not an epidemic logical study, the IEMP risk assessment is not
intended to and does not answer questions such as what caused a
statistically higher rate of cancer in one neighborhood or part
of the community.
Generating Risk Estimates
In the IEMP, we define risk to an individual as the in-
creased probability that an individual constantly exposed to one or
more chemicals will experience a particular adverse health effect
during the course of his or her lifetime (the average lifetime
is assumed to be 70 years). The IEMP risk assessment combines
estimates of toxicological potency (derived from laboratory and
occupation studies) with estimates or measurements of local contami-
nation (exposure) to calculate risks to the local population. We
discuss the basic techniques for estimating toxicological potency and
exposure below. We then present how these data are-used to develop
several different measures of risk in the Baltimore IEMP:
o Aggregate excess cancer incidence, i.e., risk to the
population
o Lifetime risks to the most exposed individual (MEI)
o Lifetime risks to the average exposed individual ' (AEI)
o Increased risk of noncancer health effects.
V-2
P'l"
-------
Estimating Toxicological Potency
Cancer. Assessing the toxicological potency for carcinogens
involves: (1) a qualitative "hazard identification',' to determine
where there is evidence that a chemical causes an adverse health
effect and, if so, (2) a quantitative estimate of the "doseresponse
relationship" or potency of the chemical. The carcinogenic assess-
ment methodology we employ in the Baltimore IEMP is consistent
with EPA* s Guidelines for Carcinogenic Risk Assessment, as published
in the Federal Register, September 24, 1984.
The qualitative hazard evaluation, which is based on a review
of human epidemiological studies and animal experiments, determines
the strength of the evidence that a substance is carcinogenic. EPA
has developed a scheme to stratify the weight of evidence into five
groupings: Group A: Human Carcinogen; Group B: Probable Human
Carcinogen; Group C: Possible Carcinogen; Group D: Not Classified;
and Group E: No Evidence of Carcinogenicity for Humans. Appendix
A discusses the groups in more detail.
If there is sufficient qualitative evidence to suggest that
a substance is a suspected carcinogen, a quantitative dose-response
relationship, or unit risk factor, is developed. EPA's Carcinogen
Assessment Group (CAG) develops potency estimates using a number of
simplifying assumptions, which include: (1) interpretation of
laboratory or animal data; (2) extrapolation ,pf data to other
species or populations; (3) extrapolation from data on high-dose
effects to low-dose exposure situations representative of most
environmental conditions (EPA assumes that cancer is a nonthreshold
health effect); and (4) estimates of the effects of exposure to
toxic chemicals.
The greatest area of uncertainty surrounds the mathematical
models used to extrapolate from high animal or human occupational
exposures to low environmental exposures. In estimating potency
from animal studies, CAG uses the most sensitive species observed
and calculates the upper (95 percent) confidence limit on the
estimated low-dose range of the dose-response relationship curve.
We emphasize that use of the upper-bound potency estimate does not
necessarily give a realistic prediction of actual risks to humans.
The true value of the risk is uncertain, and for many substances
the lower-bound estimate is zero.
Noncancer. The noncancer health effects . that we examine in
the Baltimore IEMP included 'systemic effects, such as liver and
kidney toxicity, fetal developmental effects, and neurobehavioral
effects. The crucial distinction between quantifying cancer and
noncancer effects is the assumption that cancer can occur at any
environmental exposure, i.e., there is no threshold for the effect.
V-3
-------
For most other environmentally caused health effects, we assume
that thresholds exist. Therefore, the first step in the quantita-
tive evaluation of a noncancer effect is calculating the threshold
below, which no effects are expected, i.e., the no-effect threshold.
We rely on EPA Reference Doses (RfDs) and on thresholds computed
by Regulatory Integration Division (RID) toxicologists and consul-
tants to determine the threshold dose below, which no observable
adverse effects are assumed to occur. The thresholds developed by
RID employ the same procedures used in estimating RfDs. We refer
the reader to Appendix A for a more comprehensive discussion of the
analytical approach for developing no-effect thresholds.
Estimating Exposure
To calculate exposure, we need to know the concentrations of a
chemical in the medium of concern (either ambient or indoor air or
drinking water). As part of this calculation, we also make assump-
tions about how ambient concentrations relate to actual human
exposures, or dose. We discuss each separately.
Ambient and Indoor Levels. We can estimate ambient concentrations
in twoways:directmonitoring or simulation, i.e., predictive
transport and fate modeling. In Phase I of the Baltimore IEMP, we
relied heavily on monitoring data to calculate risks to human
health. We decided primarily to use ambient data because of the
uncertainties in the available source and pollutant release inven-
tories required for modeling. We did, however, use data on sources
and pollutant emissions for estimating the contribution to ambient
concentrations from sewage treatment plants because of our experience
with such facilities in the Philadelphia IEMP. For indoor air
risks, we have made inferences on likely exposures in Baltimore
based on monitoring data from other cities.
While ambient monitoring data often provide the most direct
measure of environmental conditions, there are two important limi-
tations to their usage:
o Monitoring data are often taken from a few specific
points and must be extrapolated over the area of con-
cern. In contrast, modeling can take into account geo-
graphic variability and can pinpoint the sites of max-
imum ambient concentrations. The latter is especially
important in our assessment of lifetime risks to the MEI
and the increased risk of noncancer effects.
o Monitoring data, without extensive source modeling, can-
not link observed ambient concentrations to sources, which
which is essential for making risk management decisions.
V-4
-------
While Phase II of the Baltimore IEMP will focus on gathering
the data needed to identify sources, our reliance on ambient
data in Phase I precluded us from quantitatively, assessing the
contribution of specific sources,.
Exposure Constants. Once we have calculated ambient levels,
we need to relate these to actual human exposures. EPA uses stan-
dard assumptions, called exposure constants, to account for the
for the amount of air or water a typical person takes in each day,
and the weight of an average person. Specifically, EPA assumes
that an average person weighs 70 kilograms, breathes 20 cubic
meters of air and drinks two liters of water each day.
Risk Measures
Aggregate Annual Cancer Incidence. The aggregate population
risk is the total estimated increase in cancer incidence (number of
cases), above background levels, in an exposed population. We
calculate incidence by first establishing exposure (i.e., ambient
levels multiplied by the exposure constants). We then multiply the
estimated exposure levels by the exposed population and the appro-
priate potency factor. Finally, we divide by 70, the assumed
average lifetime, to generate annual incidence estimates.
Exposure =
amount of air
concentration in air or water x inhaled or water
consumed per day
Increased lifetime =
risk to the
population
Aggregate annual =
cancer incidence
Exposure x unit potency x
estimated exposed
population
Increased lifetime risk to the
population
- 70 years
Lifetime Risks to the Most Exposed Individual (MEI) We
define risks to the MEI as the increased probability that an
individual constantly exposed to the greatest amount of one or
more chemicals over a lifetime (70 years) will contract cancer.
Generally, this calculation is computed by simply multiplying
the maximum ambient concentration by the exposure constant and
the appropriate unit risk factor.
Lifetime cancer risk
to the most exposed
individual (MEI)
= Continuous 70 year exposure
at highest average annual
concentration
x unit
potency
V-5
-------
In the Baltimore IEMP, we could not use the techniques
traditionally employed by EPA to quantify maximum ambient air
concentrations because of data limitations. EPA generally uses
dispersion modeling to locate the point of maximum ambient con-
centration. In the Baltimore IEMP, we based our risk calcula-
tions on ambient air quality data and were, therefore, limited
in our ability to predict the maximum ambient concentration.
For our purposes, we defined the "maximum" ambient pollutant
level as the highest average measured value reported for the
pollutant at any of the monitoring sites.
Lifetime Risks to the Average Exposed Individual* We define
risk to an average individual as the increased probability that an
individual exposed to the average amount of a chemical over a life-
time will contract cancer. We calculate the lifetime risks to the
AEI by simply multiplying the average ambient pollutant level by
the exposure constant and the appropriate unit risk factor.
Increased Risk of Noncancer Health Effects. We determine the
increased riskofnoncancer healtheffects bycomparing pollutant
concentration levels with estimated no-effect human thresholds.
Because a single substance may have many different thresholds for
the different health effects associated with it, we examine each
health effect for each , substance separately. If exposures exceed
the no-effect threshold, we then indicate the possibility of in-
creased risk of that effect. We also attempt to identify the popu-
lation exposed to such concentrations.
Limitations
Although we have already highlighted some of the limita-,
tions of risk assessment, it is important to summarize these
points below to ensure that the reader properly interprets the
estimates of risk presented in this chapter.
o The estimates of individual health risk and aggregate
incidence from exposure to toxics should not be inter-
preted as precise or absolute estimates of future health
effects. The simplifying assumptions and uncertainties
in both the toxicology and exposure components are simply
too great to justify a high level of confidence in the
precision of the results. They are approximations of the
potential risks to human health and are designed to permit
comparison across different sources, pollutants, and expo-
sure pathways.
o The potency and threshold estimates used in this study
are consistently conservative (i.e., upper-bound) in the
direction of overestimating risk for the particular pollut-
ants and exposure pathways we have assessed.
V-6
-------
o We may understate risks to the extent that we have not
considered all pollutants, sources, and exposure .pathways.
o The results from a risk assessment are based primarily
on existing knowledge.. As new scientific data and tech-.
iques become available, these results may change.
ORGANICS IN THE AMBIENT AIR
On the basis of the scoping exercise described below, the
Human Health Subcommittee selected organics in the ambient air,
including benzene, as a topic deserving more attention. To perform
this analysis, RID staff and technical contractors gathered all
available information on a variety of toxic organic compounds known
or believed to be present in the Baltimore area and used these data
to conduct a preliminary exposure and risk assessment. This section
documents the data used in this analysis (pollutants, sources,
ambient levels, exposed populations, and unit risk factors), our
results, and the limitations of the analysis.
We should note that our discussion of limitations in this
section, and in the others to follow, focuses solely on limita-
tions in the scope and in the exposure data used in the analy-
sis. The limitations of the toxicological data are presented
in the previous section and are applicable to all quantitative
risk assessments.
Pollutants
For the purposes of our preliminary analysis of volatile
organic compounds (VOC) in the ambient air, we were limited to
an investigation "of those pollutants for which we either already
had data or that we could easily characterize using existing
information. We gathered information on organics in the ambient
air during Phase I in two ways: (1) sampling and analyzing
ambient air and (2) collecting and developing, where needed,
estimates of toxic emissions from facilities situated in the
Baltimore study area. Because of the substantial uncertainty in
our emissions estimates, as well as the limited source coverage,
we decided to base our risk calculations solely on the ambient
monitoring data.
The VOC monitoring program successfully sampled for and ana-
lyzed the following target pollutants, which were the focus of the
Phase I exposure and risk assessment (Phase II will encompass a
wider range of pollutants):^
V-7
-------
o Benzene
o Carbon tetrachloride
o Chloroform
o Ethyl benzene
o Xylene
o Toluene
o 1,2-Dichloroethane
o 1,2-Dichloropropane
o Trichloroethylene
o Perchloroethylene
The sampling program focused on these pollutants based on a
preliminary data review of sources in the area, as well as
the ubiquity of some of these substances, such as xylene and
toluene in urban ambient air. The sampling program and its
results are discussed in more detail later in this chapter.
Sources
Baltimore's industrial base is very diverse, including such
sectors as iron and steel production, electronics, pesticide for-
mulation, shipbuilding, and petroleum marketing. The major
Baltimore source categories emitting the ten pollutants consid-
ered in our analysis are listed in Table V-l. Phase II will be
directed toward characterizing better the sources of organics in
the ambient air and their release rates.
Of the industries listed in Table V-l, many are well-documen-
ted sources of these pollutants. An important exception to this
traditional array of source categories is volatilization from sew-
age treatment plants. Because of EPA's work in the Philadelphia
IEMP, our attention was initially directed towards Baltimore's
two municipal sewage treatment plants that receive industrial
wastewater (Patapsco and Back River). In Philadelphia, EPA's
analysis indicated that the city's largest industrial sewage treat-
ment plant was also an important source of VOC emissions.
We developed preliminary estimates of air emissions from
the Patapsco and Back River sewage treatment plants (including
both volatilization and sewage sludge incineration at Patapsco)
using data collected by Black and Veatch in 1984^ and a simple mass
V-8
-------
Table V-l
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
MAJOR SOURCES EMITTING SELECTED ORGANIC COMPOUNDS IN THE BALTIMORE AREA
Pollutant
Benzene
Carbon Tetrachloride
Chloroform
Ethyl Benzene
Xylene
Toluene
1,2-0ichloroethane
1,2-Oichloropropane
Trichloroethylene
Perchloroethylene
Sources
Bulk petroleum terninals; chemical manufacturing; sewage treatment plant
volatilization; agricultural chemical production; iron and ateel produc-
tion; utilities; municipal incineration; electronics; metal can fabrica-
tion; road vehicles; boilers;-and service stations (gas marketing)
Sewage treatment plant volatilization and chemical productions
Chemical production; cooling apice manufacturing
Utilities; gas narketing; road vehicles, solvent surface coating; and
solvent printing and graphing
Bulk petroleum terminals; chemical manufacturing; pesticide formulation;
shipbuilding; can production; phone manufacturing; road vehicles; car
body manufacturing; and service stations (gas marketing)
Sewage treatment plant volatilization;' shipbuilding; bulk petroleum
terminals; commercial printing; can production; car body manufacturing;
phone manufacturing; and aervice stations (gas marketing)
a
Gas marketing; paint manufacturing
No reported sources
Sewage treatment plant volatilization; phone manufacturing; cooking
equipment production; electroplating; and areawide industrial solvent
usage
Sewage treatment plant volatilization; can production; glass container
fabrication; and dry cleaning
Source: Baltimore IEMP database compiled by RID.
-------
balance algorithm. The algorithm assumes that the amount of a
compound that volatilizes is the difference between its influent
and effluent loadings, minus the removal efficiency of the volatile
compound. Table V-2 presents our rough calculation of volatile
organic emissions from the two industrial sewage treatment plants.
Benzene accounted for nearly 94 percent of the total air
emissions we estimated for the pollutants of concern at the
Patapsco plant. At the Back River treatment facility, benzene
and methylene chloride emissions combined accounted for roughly
85 percent of the total releases of those toxic pollutants considered.
An additional 10 percent of the estimated emissions at Back River
were attributable to chloroform.
Ambient Air Levels
From November 1983 to February 1984, RID supported a three
month ambient monitoring program for volatile organic compounds
in the Baltimore metropolitan area. EPA collected 24-hour inte-
grated air quality samples from ten monitoring sites every three
days. EPA also constructed two meterological wind stations to
obtain data on wind direction and speed and vertical dispersion.
Figure V-l shows the locations of the monitoring sites.
The air quality samples were collected on Tenax GC at a
flow rate of about 20 milliliters per minute. Forty percent
of the samples were analyzed by gas chromatography/mass spectros-
copy (GC/MS). The remaining samples were analyzed using gas
chromatography with a Hall detector for chlorinated compounds,
and a photoionization detector for aromatics.^
The results from the monitoring program are summarized in
'fable V-3. The highest observed concentrations were for xylene,
benzene, toluene, and ethyl benzene. It is also interesting to
note the variation in the concentrations of some pollutants across
the monitoring sites. Such variation indicates that exposures, and
thus risks, to these compounds are not uniform throughout the study
area. This is an important factor to consider during the Phase II
analyses to identify strategies for control.
Exposed Population
As mentioned in our discussion of risk assessment, the number
of people exposed to ambient pollutant levels is a necessary'input
to the calculation of excess aggregate cancer incidence. Similarly,
in considering risks to the MEI, one must weigh the calculated
individual risks against the number of people exposed at the site
of maximum ambient concentration.
V-9
-------
Table V-2
/"
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
ESTIMATED EMISSIONS OF VOLATILE ORGANICS
AND METALS FROM THE PATAPSCO AND BACK RIVER
SEWAGE TREATMENT PLANTS
Patapsco Sewage
Treatment Plant
Pollutant
Benzene
Chloroform
1,2-Oichloroethane
T r ichloroethylene
Perchloroethylene
Methylene Chloride
Chromium (VI)
Nickel
Cadmium2
Copper
Zinc
Mercury
Lead2
Total
Pollutant
Benzene
Chloroform
Carbon Tetrachloride
Methylene Chloride
Perchloroethylene
1,2-Dichloroethane3
Trichloroethylene'
Total
Emissions
(kkq/year)1
45.3
Back River Sewage
Treatment Plant
Emissions
(kkg/year)1
27. A
^Emission estimates have been rounded to one sig-
nificant decimal; totals may not sum because of
rounding, kkg = 1,000 kilograms.
^Emission estimates for cadmium and lead are .009
kkg/year and .02 kkg/year, respectively.
^Emission estimates for 1,2-dichloroethane and
trichloroethylene are <.001 kkg/year.
Source: Regulatory Integration Division, U.S. EPA,
Issue Paper; Publicly Owned Sewage
Treatment Facilities.
-------
FIGURE V-l LOCATIONS OF AIR QUALITY MONITORING
SITES FOR THE BALTIMORE IEMP
BALTIMORH
AND VICINITY
KFf:
AI*
-------
Table V-3
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
AVERAGE MEASURED AMBIENT CONCENTRATIONS : ORCANICS IN THE AMBIENT AIR1
(micrograns per cubic meter-ug/ra')
Average
Compound 1 2 34 5
Benzene 12.0 9.5 10.2 12.9 12.6
Carbon Tetrachloride 1.1 0.9 0.6 1.3 0.9
Xylene 30.4 21. i 20.8 20.1 13.7
1,2-Oichloroe.thane ' 0.2 0.2 0.3 0.5 0.4
1,2-Oichloropropane 0.3 0.2 0.2 0.7 0.3
Trichloroethylene 0.9 0.5 0.5 3.9 1.4
Perchloroethylene 4.8 5.4 7.0 6.0 9.3
Site
1 Guilford
2 Northeast Police Headquarters
3 Southwest Police Headquarters
4 Holabird Elementary
5 Dundalk
6 Cheaapeake Terrace Elementary
7 Sun & Chesapeake
8 Ft. McHenry
9 U.S. Coast Guard Yard
10 Riviera Beach
^Concentrations are simple arithmetic averages.
Source: Versar, Inc., Draft Data Report for Monitoring
1111
5.5 12.1 10.6 7.8
0.9 1.2 1.4 0.7
5.7 17.2 16.4 13.6
2.6 0.7 0.2 1.0
0.1 2.0 0.4 0.4
0.4 1.1 1.0 0.2
1.5 3.2 3.9 2.9
Across
10 Sites
10.3 10.4
0*6 1.0
6.3 6.2
19.2 17.8-
0.2 0.6
0.2 0.5
0.3 1.0
2.4 4.6
and Analytical Activities to Determine
Ambient Air Concentrations of Selected Toxic Pollutants in Baltimore,
Maryland,
April 18, 1984.
-------
We determined the total population in the Baltimore study
area, approximately 1.5 million, using 1980 census data provided by
the Baltimore Regional Planning Council. [We describe the boundaries
of the study areathe contiguous urbanized areas of greater Balti-
morein Chapter II.] We then subdivided the data and assigned
subpopulations to each of the 10 monitoring locations (see Table
V-4) considering land usage and proximity to the monitoring site.
The population apportionment was needed to calculate aggregate
incidence. We estimated exposure by multiplying the ambient levels
at each monitoring location by the subpopulation assigned to each
site.
In traditional air analyses, dispersion models are used to
calculate ambient levels and the resulting population exposures.
However, we could not utilize dispersion modeling in Phase I because
of limitations in our source and emissions data.
Unit Risk Factors
For each of the organic compounds considered in the analysis,
with the exception of 1,2-dichlorbpropane (which we discuss below),
we used EPA's GAG potency data for cancer effects. For non-cancer
effects, RID toxicologists developed thresholds for health effects.
We summarize our findings in each area.
Cancer
Table V-5 lists the cancer unit risk factors available for
the pollutants considered in the analysis. We present the unit
risk values that were used at the time Phase I was completed
(1985), as well as the values that are currently available.
The most notable changes are for 1,2-dichloroethane and chloro-
form, for which the estimated potency increased nearly three-
fold and slightly more than twofold, respectively; and trichloreth-
ylene and perchloroethylene, for which the estimated potency decre-
ased by more than half. CAG revised the cancer unit risk factor
for benzene upward by a modest amount.
In this field, scientists generally do not consider such
revisions significant. Given the considerable uncertainty associated
with many steps in the process of estimating cancer potency, small
changes in the order of two or three are not considered major-
As shown in Table V-5, EPA does not currently regard as
carcinogens all the pollutants that are considered. In particular,
EPA does not consider ethyl benzene, toluene, and xylene carcinogenic
(As a result, we did not have a unit risk factor for these chemicals).
However, exposure to ethyl benzene, toluene, or xylene at high
ambient levels may increase the risks of noncancer health effects.
All unit risk factors presented in this table were developed
by CAG, with the exception of 1,2-dichloropropane.
V-10
-------
Table V-4
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
EXPOSED POPULATION: ORGANICS IN THE AMBIENT AIR
Total
Monitoring Site Exposed Population
Guilford 333,186
Northeast Police Headquarters 745,187
Southwest Police Headquarters 285,517
Holabird Elementary 48,771
Oundalk 14,270
Chesapeake Terrace Elementary 12,880
Sun and Chesapeake 16,848
Ft. McHenry 23,997
U.S. Coast Guard Yard 8,416
Riviera Beach 44,642
Total 1,533,714
Source: Regional Planning Council, Baltimore,
Maryland, Census '80; Population and
Housing Characteristics by Census Tract,
April 1982.
-------
Table V-5
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND CANCER UNIT RISK VALUES:
ORGANICS IN THE AMBIENT AIR
Inhalation (ug/m3)
-1
Pollutant
Benzene
Trichloroethylene
Perchloroethylene
1,2-Dichloroetha ne
Chloroform
Carbon Tetrachloride
1,2-Oichloropropane
Ethyl Benzene
Toluene
Xylene
1985
Analysis
ID'"
io-6
6.9 x
4.1 x
1.7 x
7.0 x
1.0 x
1.5 x
1.8 x 10"5
*
*
Revised
19862
8.0 x
1.3 x
io-6
4.8 x 10-7
10-5
10-5
10-5
1.8 x 10°
2.6 x
2.3 x
1.5 x
Grouping
Based on
EPA Criteria3
A
B2
B2
82
B2
B2
C
Current EPA policy does not regard ethyl benzene, toluene, or xylene
as a carcinogen.
IThe unit risk factors presented in this column were developed in 1985
and used in the preliminary Phase I risk calculations.
2The unit risk factors presented in this column are current as of
5/7/86.
5EPA weight-of-evidence classifications:
A = human carcinogen; B2 = probable carcinogen; C = possible
carcinogen. (See Appendix A for more detail.)
Source: All cancer unit risk factors were developed by EPA's.
Carcinogen Assessment Group, except for 1,2-dichloropropane.
RID calculated the unit risk factor for 1,2-dichloropropane
using the linearized, multi-stage model and data from a 1977
NTP study. The NTP study is currently undergoing EPA review;
thus, the potency value could change.
-------
RID toxicologists and consultants calculated a unit risk factor
for 1,2-dichloroprapane from a potency (qi*) value developed for
EPA1s' Drinking Water Criteria Document on DCP (March 2, 1984).
The potency (qi*) value was based on a 1983 National Toxicology
Program (NTP) study.4 While the NTP study can be used as a
preliminary basis for evaluating risks to human health, we should
note that the Drinking Water Criteria Document on DCP, and thus the
NTP study, are currently undergoing EPA review. As a result, this
unit risk factor is subject to change.
Noncancer Health Effects
Table V-6 details the noncancer health effects and the thresh-
olds above which there is an increased risk of those effects.
These values were developed by RID toxicologists. The disease
categories considered include the following systemic effects:
fetal development effects; liver toxicity; neurobehavioral effects;
kidney toxicity; reproductive effects; and other, e.g., respiratory
difficulties, anemia, and leukopenia.
Preliminary Screening Analysis and
Risk Assessment Results
We calculated and evaluated three measures of risk posed by
exposure to each of the organic compounds considered in our
preliminary investigation: excess aggregate annual incidence of
cancer; lifetime individual cancer risks at the site of maximum
concentration (a surrogate for MEI risks); and the potential for
noncancer health effects. We describe the calculations, used to
derive each measure of risk and the results below.
Aggregate Cancer Incidence
We calculated the annual excess cancer incidence for each
pollutant by first multiplying the ambient concentration at each
monitoring site by the exposed subpopulation and the appropriate
cancer unit risk factor, summing across all sites for each pollu-
tant, and then dividing by 70 (the standard EPA assumption regard-
ing the number of years in a lifetime). Table V-7 presents the
upper-bound estimates of annual cancer incidence for each pollutant
using the unit risk factors from the initial Phase I analysis and
those currently available.
Please note that small differences (a factor of two or
three) between numbers in the following tables do not necessar-
ily warrant the conclusion that one chemical is twice or three-
times the risk of another. The precision in this screening
exercise does not allow the making of such quantitative
V-ll
-------
Table V-6 ,
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
Pollutant
Benzene
Trichloroethylene
Perchloroethylene
1, 2-Dichloroethane
Chloroform
Carbon Tetrachloride
1,2-Oichloropropane
Ethyl Benzene
Toluene
Xylene
*The threshold values
NONCANCER HEALTH EFFECTS AND, PRESUMED
ORGANICS IN THE AMBIENT
RID-Oerivec
Threshold
Health Effect (uq/m^j
Fetal developmental 4. 1
Blood effects *
Liver 26
Neurobehavioral 26
Kidney 3770
Fetal developmental 909
Liver 69.9
Kidney 69.9
Liver 26
Neurobehavioral 26
Gastrointestinal 26
Kidney 26
Fetal developmental 2.4
Liver 69.9
Neurobehavioral 11.7
Fetal developmental 24.2
Liver 2.4
Neurobehavioral 2.4
Kidney 108
Reproductive 430
Liver . 308
Kidney 308
Liver 340
Fetal developmental 476
Liver 1,010
Neurobehavioral 580
^Kidney 1,010
Reproductive . 500
Respiratory 1,010
HUMAN THRESHOLDS:
AIR1
Source
Kuna & Kapp 1981
Snyder et al. 1980
Unverified 1985 ADI
Grandjean et al. 1955
Tucker 1982
Nelson et al. 1979
Coler & Rossmiller 1953
NCI 1977
Kozik 1957
Kozik 1957
USEPA 1982a
Heppel et al. 1946;
Hoffman et al. 1971
Sen wet z et al. 1974
Hey wood et al. 1979; 1986 ADI
Challen et al. 1958;
USEPA 1985
Schwetz et al. 1974
EPA 1980 & 1984; Smyth 1935 & 1936
Moller 1973
EPA 1980
Adams et al. 1952
Basu et al. 1984; NTP 1983
NTP 1983
Verified 1986 -ADI; Wolf et al. 1956
Hudak & Ungvary 1978
USEPA 1983
Hanninen et al. 1976; Seppalainen et al.
1978
USEPA 1983
Matsumoto et al. 1971
Von Oettingen et al. 1942; Bruckner &
Peterson 1981
Fetal developmental 52.8 Ungvary et al. 1980
Liver 215 Bewers et al. 1982; Tatrai et al. 1981
Neurobehavioral 215 USE PA 1984; Savolainen et al. 1979
Respiratory 215 Hipolito 1980; Smyth & Smyth 1928
Cardiovascular 215 Hipolito 1980; Hirsch 1932
Anemia and leukopenia 215 Browning 1965; Hipolito 1980
presented in this table were developed by RID and are current as of 5/7/86.
These threshold effect levels are currently undergoing review
Assessment Office.
by EPA's Environmental Criteria and
-------
Table V-7
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF ANNUAL EXCESS CANCER INCIDENCE:
ORGANICS IN THE AMBIENT AIR
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
Upper-Bound Annual Cases1,2
Pollutant 1985 Revised
(weight of evidence)3 Analysis* 19865
Benzene (A) 1.6 1.8
Trichloroethylene (B2) 0.1 0.02
Perchloroethylene (B2) 0.2 0.1
1,2-Dichloroethane (B2) 0.04 0.1
Chloroform (B2) 0.2 0.4
Carbon Tetrachloride (82) 0.3 0.3
1,2-Oichloropropane (C) . 0.1 0.1
Total 2.5 2.8
XTHE UNIT RISK FACTORS USED IN THIS ANALYSIS.ARE BASED ON
CONSERVATIVE ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND
ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS IN
SEVERAL AREAS OF THE ANALYSIS, SUCH AS EXPOSURE CALCULA-
TIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE CALCU-
LATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF
ACTUAL CANCER RISKS IN BALTIMORE. ACTUAL RISKS MAY BE SIG-
NIFICANTLY LOWER; IN FACT, THEY COULD BE ZERO. THE PROPER
FUNCTION OF THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT
AND EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS
EXAMINED.
2RID'S ESTIMATE OF THE ACTUAL NUMBER OF CANCER CASES IN THE
STUDY AREA IN 1984 IS 8,000 CASES. (SEE 11-8 AND 9.) THIS
NUMBER SHOULD SERVE ONLY AS A POINT OF REFERENCE IN UNDER-
STANDING THE RISK ESTIMATES PROVIDED. IN ADDITION, THE
RISK ESTIMATES SHOULD NOT BE INTERPRETED AS REPRESENTING
THE TOTAL UPPER-BOUND CANCER RISKS FROM ALL POLLUTANTS IN
ANY PARTICULAR MEDIUM. THEY DO NOT TAKE INTO ACCOUNT ALL
POLLUTANTS THAT MAY BE PRESENT IN THE MEDIUM, ALL SOURCES
OF THESE POLLUTANTS, AND ALL EXPOSURE SCENARIOS INVOLVING
PATHWAYS OR EXPOSURES OF SHORT DURATION TO RELATIVELY HIGH
DOSES.
'EPA weight-of-evidence classifications: A = human carcino-
gen; B2 = probable carcinogen; C = possible carcinogen.
(See Appendix A for more detail.)
*The incidence estimates listed in this column were calcu-
lated using cancer unit risk values developed in 1985.
5The incidence estimates listed in this column were calcu-
lated using current (5/86) cancer unit risk values.
-------
distinctions. At best we can state that one chemical qualitatively
poses greater risk than another. Benzene is clearly a greater risk
than all other chemicals examined, but carbon tetrachloride can
not be said to be much worse than chloroform.
This preliminary analysis suggests that exposure to a subset
of organics found in the ambient air may result in an upper bound
estimate of about 3 excess cancer cases a year (the revised unit
risk factors do not dramatically change the findings). Benzene ex-
posures alone account for almost two-thirds of the total estimated
excess cancer incidence. Perchloroethylene, chloroform, and carbon
tetrachloride together contribute less than one case per year, or
most of the remaining third for the subset of pollutants considered.
Upper Bound Lifetime Individual Cancer Risks
We estimated upper bound lifetime individual cancer risks by
multiplying the maximum observed ambient concentration for each
pollutant by the appropriate unit risk factor. Table V-8 details
the calculated upper-bound lifetime cancer risks, as well as the
number of people exposed to thesemaximum concentration levels.
Using the 1985 unit risk factors, upper-bound lifetime cancer
risks range from close to 2 chances in 100,000 (trichloroethylene
and perchloroethylene) to almost 9 chances in 100,000 (benzene).
The revised unit risk factors result in a wider spread of estimates,
ranging from a low of roughly 0.5 chances in a 100,000 (perchloro-
ethylene to a high of about 10 case in 100,000 for both benzene and
chloroform. The most noticeable change is for chloroform, for which
the estimated lifetime individual cancer risk increases by a factor
of slightly more than two.
\
Noncancer Health Effects
We assessed the potential for noncancer health effects by
comparing the measured ambient concentration with the no-ffect
threshold levels of each individual pollutant. In general, there
should be a concern over the possibility of adverse health effects
if the ambient concentrations exceed the threshold. The degree of
concern should be gauged by the number of people potentially exposed
and the extent to which the threshold is exceeded.
A comparison of ambient levels at each of the monitoring sites
to the no-effect threshold levels indicates that there may, be an
increased risk of noncancer health effects for benzene and chloro-
form exposures. In the case of benzene, the ambient levels at each
of the ten monitoring sites exceed the no-effect threshold for fetal
development effects. Ambient chloroform levels at monitoring sites
#6, #7, and #8 also exceeded the no-effect threshold for fetal
development effects.
V-12
-------
Table V-8
BALTIMORE KMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF LIFETIME INDIVIDUAL CANCER RISKS
AT SITE OF MAXIMUM AVERAGE CONCENTRATION: ORGANICS IN THE AMBIENT AIR
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
Upper-Sound Lifetime
Cancer Risk (inhalation)1,2
Pollutant 1985 Revised Exposed
(weight of evidence)3 Analysis4 19865 Subpopulation
Benzene (A) 8.9 x 10~5 1.0 x 10"* 48,771
Trichloroethylene (B2) 1.6 x 10~5 5.1x 10"6 48,771
Perchloroethylene (B2) 1.6x10"' 4.5. x 10"* 14,270
1,2-Oichloroethane (B2) 1.8 x 10"5 6.8 x 10~5 12,880
Chloroform (B2) 4.7 x 10~5 1.1 x 10~4 23,997
Carbon Tetrachloride (B2) 2.1 x 10~5 2.1 x 10~5 23,997
1,2-Dichloropropane (C) 3.6 x 10~5 3.6 x 10~5 16,848
JTHE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE
ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUM) ESTIMATES. BECAUSE OF
LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES MERE CALCU-
LATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL CANCER
RISKS IN BALTIMORE. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER; IN FACT, THEY
COULD BE ZERO. THE PROPER FUNCTION GF THE ESTIMATES IS TO HELP LOCAL
OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS
EXAMINED.
^Individual risks were calculated using the maximum average concentration
measured across all monitoring sites.
3EPA weight-of-evidence classifications: A = human carcinogen;
B2 = probable carcinogen; C = possible carcinogen. (See Appendix A for
more detail.)
*The risk estimates shown in this column were calculated using cancer unit
risk factors available in 1985.
*The risk estimates listed in this column were calculated using current
(5/86) unit risk factors.
-------
We used statistics on the number of births occurihg in the
Baltimore area to identify the subpopulation exposed to the higher
levels of benzene. In 1984, this number was 27,426. Please note,
however, that the total number of pregnancies (successful and un-
successful) that occur at any one time are likely to exceed by a
considerable amount the number of pregnancies that successfully come
to terra. Furthermore, we do not even know the number of pregnancies
that come to term in the areas around the three monitoring sites
(sites number 6,7 and 8) and hence have not estimated the sub-
populations in these areas who are also exposed to higher levels of
chloroform. Thus our estimates of exposed subpopulations are likely
to be underestimated.
We must reiterate that the threshold values that RID calcu-
lated are not official EPA values. In this screening exercise,
these values serve only to identify chemicals that warrant greater
regulatory attention. Much more study is needed to determine how
much the ambient levels for these pollutants actually contribute,
if at all, to adverse health effects in these subpopulations.
Limitations
As with most analyses, there are limitations that must be
carefully considered when interpreting the results. At the beginn-
ing of this chapter, we mentioned the limitations of risk assessment.
In this section, we list the major uncertainties associated with
the scope of our analysis and the exposure data used to estimate
risks from organic pollutants in the ambient air.
First, we must emphasize that this preliminary investigation
considered only ten organic compounds. While the study participants
agreed that analysis of this limited set of pollutants was sufficient
for guiding decisions on Phase II study topics, the results from
such analysis should .not be construed as definitive. Phase II will
explore a wider group of pollutants, and the results from this
effort may be different than those presented here.
Second, the use of Tenax GC as the absorbent may introduce
a bias in our results for certain compounds. Although Tenax
was considered the best sampling medium at the time we conduc-
ted our ambient air monitoring, its reliability has been called
into question in the last ±wo years; EPA is currently exploring an
alternative to Tenax.
Third, the ambient air monitoring data may not be representa-
tive of actual pollutant concentrations in the Baltimore metropolitan
area. We selected sites near point sources because of our interest'
in verifying emission estimates from these sources at some point in
the future. Limited resources prevented us from pursuing a more
V-13
-------
elaborate sampling program. As a result, the ambient data we
collected may not adequately represent actual receptor ambient
levels because we could not site monitors throughout the entire study
area.
Fourth, our usage of average ambient values masks the peaks
in concentrations that individuals are exposed to daily. Thus,
our risk calculations may understate the total body burden from
exposures to organics in the environment. This applies to carcino-
genic as well as noncarcinogenic effects.
Finally, though we calculated the lifetime risks to the MEI
using the maximum observed concentrations, it is unlikely that
these samples were taken at the point of maximum concentration.
Thus', we may have underestimated lifetime risks to the MEI for the
pollutants considered. We simply can not estimate the how these
uncertainties affect our risk estimates.
METALS IN THE AMBIENT AIR AND LEAD IN THE GENERAL ENVIRONMENT
The second topic nominated by the Human Health Subcommittee
for further investigation in Phase II was metals in the ambient air
and lead in the general environment. Metals are released from many
sources, are commonly detected in the ambient air, and can be
associated with acute and chronic adverse health effects.
Unlike the analysis of organics, we relied entirely on existing
information to assess the exposures and risks associated with
metals in the ambient air and did not generate new data. We also
utilized available ambient air quality data to determine ambient
concentrations. We describe the data used in our analysis and our
findings in more detail below.
Because of the ubiquity of lead in the environment, the
risk from exposures to lead could not be examined solely from the
perspective of ambient air. Consequently, other routes of exposure
were also considered for this metal in evaluating its importance as
a risk to human health in the study area.
Pollutants
While many metals are found throughout the environment, the
Technical Advisory Committee focused our information gathering
efforts on four heavy metals: cadmium, chromium, lead, and zinc.
These pollutants were selected because they were considered to be
reasonable indicators of the potential human health risks from all
metals in the ambient air. There were also many known sources of
these pollutants.
It is important to note that chromium can have two valence
states: trivalent and hexavalent. While there is currently
insufficient health data to suggest that trivalent chromium
poses a threat to human health, hexavalent chromium is a proven
V-14
-------
human carcinogen. We attempted to distinguish between these two
valence states/ where possible, but found that information was only
available for total chromium. As a result, we used sensitivity
analysis to explore how differing'assumptions about the levels of
hexavalent chromium as a fraction of the total (e.g., 100 percent
vs. 1 percent of total chromium) would affect our estimates of
risk.
Sources
The metals of concern in this study are released to the air
from both area and point sources. Cadmium is released from fuel
combustion and road vehicles; trace amounts are also emitted from
the burning of used oil. The major point sources emitting cadmium
are electroplaters and large industrial boilers.
While the valence state of chromium from specific sources is
not known, the following generalizations can be made on the basis
of existing information. Trivalent chromium is usually associated
with steel, ferrochromium, refractory, and cement manufacturing,
chromium ore refining, and coal and oil combustion. Hexavalent
chromium may be emitted by such sources as chromium chemical plants,
sewage sludge and municipal refuse incineration, and cooling towers*
Lead is emitted to the ambient air in large part by road
vehicles, although these loadings will decline as the lead in
gasoline is phased out. Other area sources of lead include used
oil combustion, which will also become less significant with the
lead phasequt, and heating. Lead is also released from several
point sources, such as steel mills, municipal waste incineration,
and utility boilers, but these emissions do not appear to be as
significant as those from area sources.
Other sources of lead to which people, particularly children,
are exposed are street, backyard or playground, and household dust.
Lead in the ambient air settles over time in outdoor, and to a
lesser extent, indoor dust. Another important source is lead from
lead-based paint which can be ingested directly in paint chips or
becomes a constituent of household dust.
Finally, zinc is emitted from only a few known sources in
the Baltimore study area, and these releases are small. Heating
and used oil combustion are the principal area sources emitting
zinc. Trace amounts are also released by sewage sludge incineration.
Ambient Levels
We were able to locate ambient air quality data for three of
the four pollutants: lead, chromium (total), and cadmium. The data
were obtained from the 1982 Annual Air Quality Data Report for the
State of Maryland and were judged sufficient for characterizing
average concentrations for the Phase I screening activities.
V-15
-------
The values used for cadmium deserves special attention be-
cause no cadmium was actually detected at any of the monitoring
sites. Instead, the values for cadmium were derived from the
lowest concentrations that could be detected by the monitoring and
analytical instruments and hence are likely to overestimate actual
average cadmium levels at the monitoring sites. The use of these
values allows us to place an upper bound on the possible risk posed
by cadmium. It cannot serve as an even a rough estimate of actual
risk. We also substituted detection limit values for chromium
where actual concentrations were below detection limits.
Table V-9 summarizes the average ambient concentrations found
at 9 monitoring sites; Figure V-2 shows the locations of these
sites throughout the study area. A more detailed discussion of the
metals monitoring program can be found in the State of Maryland
report.
Exposed Population
As in the organics analysis, we used a 5-km. grid system to
partition the study area population. The population in each grid
section was then assigned to a monitoring site that was considered
most representative based on land use and proximity. Table V-10
lists the exposed subpopulation corresponding to each monitoring
site.
Unit Risk Factors
Cancer
Of the four pollutants considered, cancer unit risk factors
were available only for cadmium and hexavalent chromium (see
Table V-ll). EPA does not currently regard trivalent chromium,
lead, or zinc as carcinogenic through inhalation. However, there
is limited evidence that some lead compounds induce tumors in ex-
perimental animals. GAG is reviewing these data, but has not yet
made a determination as to the potential carcinogenicity of lead.
Noncancer
Table V-12 lists the noncancer health effects for cadmium,
hexavalent and trivalent chromium, and zinc. The health effects
data for lead are currently undergoing internal review and are,
V-16
'in'1 mw
-------
Pollutant
Lead
Chromium
Cadmium2
Zinc
A
0.29
0.01
0.00
N.A.
B
0.40
0.03»
0.00*
Table V-9
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
AVERAGE MEASURED AMBIENT CONCENTRATIONS:
METALS IN THE AMBIENT AIR1
(ug/m3)
Site
K M
0.48 0.33
0.03 0.01
0.00 0.00
C
0.49
0.03*
0.00*
N 0 P Q
0.49 0.23 0.23 0.47
0.00 0.02** 0.02 0.02**
0.00 0.00** 0.00 0.00**
N.A. N.A. N.A. N.A.
Sites
A: Northeast Police Headquarters
B: Guilford
C: Southwest Police Headquarters
K: Fire Department
M: Essex
N: Glen Burnie
0: Sellers Point
P. Canton Pier 4
Q: Southeast Police Headquarters
Average
Across Sites
0.38
0.02
0.00
N.A.
concentration estimates were rounded to two significant decimals.
^Average measured ambient concentrations for cadmium were .001 ug/m at sites A, B, C, K, M, and N;
cadmium concentrations at sites 0, P, and Q were .002 ug/m3. These concentrations are in the range of
detection limits.
N.A. = Not available.
» = No data available for this site; estimates based on station K because of proximity and similar
meteorological conditions.
*» = No data available for this site; estimates based on station P because of proximity and similar
meteorological conditions.
Source: Stste of Maryland, 1982 Annual Air Quality Data Report for the State of Maryland.
-------
-f
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
CO*. DMMH
THE BALTIMORE IEMP
BALTIMORE CITY
Bs
K
LEGEND
NEPD
GUILFORD
SWPD
FT. MCHENRY
SUN /CHESAPEAKE
HOLABIRD ELEMENTARY
COAST GUARD
DUNDALK
RIVIERA BEACH
CHESAPEAKE TERRACE
FIRE DEPT.
195
ESSEX
GLEN BURNIE
SOLLERS POINT
CANTON PIER 4
SEPD
FORT HOLABINE
OLD TOWN
BALTIMORE COUNTY
Q
Sparrows Point
Glen Burnle
BWI Airport -IJ +
ANNE ARUNDEL COUNTY
-------
Table V-10
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
EXPOSED POPULATION: METALS IN THE AMBIENT AIR
Total
Monitoring Site Exposed Population
Northeast Police Headquarters 672,036
Guilford 181,166
Southwest Police Headquarters 233,500
Fire Department (near Guilford) 76,024
Essex 92,251
Glen Burnie 159,693
Sellers Point 24,372
Canton Pier 4 27,655
Southeast Police Headquarters 66,384
Total 1,533,081
Source: Regional Planning Council, Baltimore, Maryland,
Census '80: Population and Housing Character-
istics by Census Tract. April 1982.
-------
Table V-ll
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND CANCER UNIT RISK VALUES:
METALS IN THE AMBIENT AIR
Inhalation (ug/m3)
-1
1985
Pollutant Analysis1
Cadmium 2.3xlO~3
Chromium (hexavalent) 1.2xlQ-2
Chromium (trivalent) *
Lead »
Zinc »
Revised
19862
1.8xlO~
1.2x10'
,-2
*
*
Grouping
Based on EPA
Criteria3
Bl
A
Current EPA policy does not regard trivalent chromium, lead
or zinc as carcinogens. As a result, we could not perform
a risk assessment for these compounds.
*The unit risk factors presented in this column Mere
developed in 1985 and used in the preliminary Phase I risk
calculations.
2The unit risk factors presented in this column are current
as of 5/7/86.
3EPA weight-of-evidence classifications:
A = human carcinogen; Bl = probable carcinogen. (See
Appendix A for more detail.)
Source: All cancer unit risk factors Mere developed by
EPA'a Carcinogen Assessment Group.
-------
Table V-12
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
NONCANCER HEALTH EFFECTS AND PRESUMED HUMAN THRESHOLDS:
METALS IN THE AMBIENT AIR1
RID-Derived
Threshold
Pollutant
Cadmium
Chromium (hexavalent)
Chromium (trivalent)
Zinc
Health Effect
Fetal developmental
Liver
Respiratory
Kidney
Reproductive
Fetal developmental
Liver
Ulceration/perforation
of nasal septum
Reproductive
Fetal developmental
Liver
Kidney
Reproductive
21.0
0.42
2.0
0.24
119
17.0
17.0
17.0
17.0
3,500
3,500
3,500
1,540
Source
Schroeder & Kitchener 1971
Friberg 1950
Comm. European Communities 1978
Kjellstrom et al. 1977
Scharpf et al. 1972
Gale 1978
NIOSH 1975
NIOSH 1975
Gale 1978
.Clement; unverified ADI
Clement; unverified ADI
Clement
Schicker & Cox 1968
he threshold values presented in this table were developed by RID and are current as of 5/7/86.
-------
therefore, unavailable for our use at this time. There is,
however, strong scientific evidence that airborne lead can have
adverse health impacts, particularly for children. These health
effects include blood-related problems and neurological dysfunc-
tions, such as potential impairment of intelligence as measured in
IQ tests. . At low levels of exposure, lead can also lead to increased
hypertension in adult males.
We could not estimate total exposure to lead; hence we could
not evaluate the relative importance of ambient lead to total
non-cancer risk from lead. As a surrogate for an estimate of
total exposure, we used the number of hospitalizations for lead
chelation therapy as an indication of the potential public health
threat posed by the presence of this element in the environment.
Lead chelation therapy is used to filter lead from the blood;
however, it cannot help reverse damage that has already occurred,
especially for neurobehavioral effects. Clearly, the use of this
surrogate represents a lower bound of the importance of non-cancer
health effects of lead.
Preliminary Screening Analysis and
Risk Assessment Results
We assessed the impact of metals on human health using the
three of the four measures of risk discussed earlier: annual
cancer incidence; lifetime individual cancer risks at the site of
maximum concentration (a surrogate for MEI risks); and the increased
potential for noncancer health effects. We calculated the excess
aggregate incidence of cancer for each pollutant by first multiplying
the ambient concentration at each monitoring site by the exposed
subpopulation, summing this product across the sites, and then
multiplying this sum by the appropriate cancer unit risk factor.
We then divided by 70 to estimate annual incidence.
We computed upper-bound lifetime individual cancer risks by
multiplying the maximum average ambient concentration for each
metal by the corresponding unit risk factor. Finally, we compared
the ambient air levels with the no-effect thresholds to determine
whether these concentrations presented levels of .regulatory concern
for noncancer health effects. We present our results for each
measure of risk below.
Cancer Incidence
Since cancer unit risk factors were only available for hexava-
lent chromium and cadmium, we could only assess the risks from
exposure to these two metals. Without any knowledge of the actual
ambient levels of hexavalent chromium, we evaluated a range of
possible concentrations (100 percent, 50 percent, 10 percent, 1
percent, and 0% of total chromium levels). Zero percent is possible
for the reasons explained above. in this case, the cancer risk for
cadmium Would be negligible given current information on the carci-
nogenicity of other prevalent species of chromium. Our results
V-17
-------
suggest that exposure to chromium and cadmium could result in an
excess (upper-bound estimate) of a negligible number to an upper-
bound estimate of roughly 4 excess cancer cases per year (see Table
V-13); the risks are almost entirely due to chromium ambient
levels and the fraction of chromium assumed to be hexavalent.
Future initiatives to better quantify the ambient concentrations
of hexavalent chromium, as well as its sources, could reduce
the uncertainty of such risk assessments.
Upper-Bound Lifetime Individual Cancer Risks
Table V-14 presents the upper-bound lifetime individual
cancer risks from hexavalent chromium and cadmium. As mentioned
above, we assumed detection level values for cadmium and chromium
whenever we were unable to detect the metals at a monitoring site.
Thus, the actual monitored concentrations were below analytical
detection levels.Using the detection level value rather than zero
(or someothervalue smaller than the detection level value) for
the monitored concentration serves to overestimate actual ambient
air exposures and hence risk from both metals.
The upper-bound lifetime cancer risks for cadmium, using the current
CAG unit risk factor and assuming the detection level value for its.
concentration in the ambient air, is roughly 4 chances in a million
(5 chances based upon 1985 CAG value). This estimate of risk from
cadmium is presented only to show the extent to which we are able
to detect a significant risk to human health from the metal. Since
the real ambient concentration is less than the value we used, we
can state with reasonable confidence, for the purpose of this~
screening analysis, that cadmium does not pose a significant cancer
risk through ambient exposures.
Also, uncertainties in the ambient levels of hexavalent
chromium forced us to consider a wide range of possible average
concentrations. This was useful because Phase I was a screening
exercise to identify possible problems requiring greater regulatory
attention and not an attempt to estimate absolute risk. The range
of potential upper-bound lifetime cancer risks for chromium spans
two orders of magnitude, from a negligible value (if 0 percent of
total chromium is hexavalent) to roughly 4 chances in 10,000 (if
100 percent of total chromium is hexavalent). The average ambient
level is also overestimated because the value for the detection
level was used whenever the metal was not detected at a monitoring
site (and hence its actual concentrations were below the level of
detection).
V-18
-------
Table V-13
BALTIMORE IEMP PRELIMINARY
RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF ANNUAL
EXCESS CANCER INCIDENCE:
METALS IN THE AMBIENT AIR
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
1 7
Upper-Bound Annual Cases ,'
Pollutant 1985 Revised
(weight of evidence)3 Analysis4 19865
Chromium (A)
Total hexavalent 4.2 4.2
SOS hexavalent 2.1 2.1
10% hexavalent 0.4 0.4
IS hexavalent 0.04 0.04
OS hexavalent 0.00 0.00
Cadmium6 (81) 0.00 to 0.05 0.00 to 0.04
Total 0.00 to 4.25 0.00 to 4.24
*THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED
ON CONSERVATIVE ASSUMPTIONS THAT GENERALLY PRODUCE
UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA
AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK
ESTIMATES WERE CALCULATED AS AIDS TO POLICY DEVELOP-
MENT, NOT AS PREDICTIONS OF ACTUAL CANCER RISKS IN
BALTIMORE. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER;
wrnO7~TH&MMHtgHibZiSi*- THE PROPER FUNCTION OF
THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND
EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS
EXAMINED.
2RID'S ESTIMATE OF THE ACTUAL NUMBER OF CANCER CASES IN
THE STUDY AREA IN 1984 IS 8,000 CASES. (SEE II-8 AND
9.) THIS NUMBER SHOULD SERVE ONLY AS A POINT OF
REFERENCE IN UNDERSTANDING THE RISK ESTIMATES
PROVIDED. IN ADDITION, THE RISK ESTIMATES SHOULD
NOT BE INTERPRETED AS REPRESENTING THE TOTAL UPPER-
BOUND CANCER RISKS FROM ALL POLLUTANTS IN ANY PARTICU-
LAR MEDIUM. THEY DO NOT TAKE INTO ACCOUNT ALL POLLU-
TANTS THAT MAY BE PRESENT IN THE MEDIUM, ALL SOURCES
OF THESE POLLUTANTS, AND ALL EXPOSURE SCENARIOS
INVOLVING PATHWAYS OR EXPOSURES OF SHORT DURATION TO
RELATIVELY HIGH DOSES.
*EPA weight-of-evidence classifications: A = human
carcinogen; 81 = probable carcinogen. (See Appendix A
for more detail.)
*The incidence estimates listed in this column were
calculated using cancer unit risk factors available in
1985.
5The incidence estimates listed in this column were
calculated using current (5/86) unit risk factors.
^Measured cadmium concentrations were below detection
limits. For screening purposes only, we calculated
risks to human health assuming a range in ambient ,
concentration from zero to the detection limit.
-------
Table V-1A
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF LIFETIME INDIVIDUAL CANCER
RISKS AT SITE OF MAXIMUM AVERAGE CONCENTRATION:
METALS IN THE AMBIENT AIR
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
Pollutant
(weight of
evidence)
Chromium7 (A)
--All hexavalent
508 hexavalent
105 hexavalent
IS hexavalent
OK hexavalent.
Cadmium8 (Bl)
1985
Analysis
Upper-Bound Lifetime
Cancer Risk
(inhalation)1,2
Revised
3.6 x 10-4
1.8 x 10~*
3.6 x 10-5
3.6 x 10"6
0.0
0.0 to 4.6 x 10~6
19B63
Exposed
Subpopulation^
3.6 x 10~* 490,690
l.B x 10~r 490,690
3.6 x 10"; 490,690
3.6 x 10~6 490,690
0.0
0.0 to 3.6 x 10~6 118,411
JTHE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVA-
TIVE ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES.
BECAUSE OF LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE
ANALYSIS, SUCH AS EXPOSURE CALCULATIONS AND POLLUTANT SELECTION,
RISK ESTIMATES WERE CALCULATED AS AIDS TO POLICY DEVELOPMENT, NOT
AS PREDICTIONS OF ACTUAL CANCER RISKS IN BALTIMORE. ACTUAL RISKS
MAY BE SIGNIFICANTLY LOVER; IN FACT, THEY COULD BE ZERO. THE
PROPER FUNCTION OF THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT
AND EVALUATE ISSUES AND SET PRIORITIES FOR THE TOPICS EXAMINED.
'Individual risks calculated using the maximum average concentration
measured across all monitoring sites.
3EPA weight-of-evidence classifications: A = human carcinogen;
Bl = probable carcinogen. (See Appendix A for more detail.)
*The risk estimates listed in this column were calculated using
cancer unit risk values available in 1985.
'The risk estimates listed in this column were calculated using
current (5/86) cancer unit risk values.
^Chromium and cadmium concentrations were the same at more than one
location. As a result, the exposed subpopulation is the sum of the
exposed populations at more than one site.
7Maximum concentration is 0.03 ug/ra3.
^Measured cadmium concentrations were below detection limits. For
screening purposes only, we calculated risks to human health assum-
ing a range in ambient concentration from zero to the upper end of
the detection limit (about 0.002 ug/m3).
-------
Noncancer Health Effect
The ambient concentration of cadmium and chromium are well
below the no-effect thresholds for noncancer health effects.
Lead levels are also well below the, national ambient air quality
standard. Because people are exposed to lead in all media, ambient
lead, neverthless, does contribute some, albeit unclear, amount to
total lead exposure. As stated above, we did not attempt to
quantify total lead exposures in the Baltimore study area.
Limitations
As in the analysis of organics, there are several limitations
to our metals risk assessment that must be recognized. First,
because of data constraints, we were only able to consider three
of the four target pollutants. Zinc does not lead to toxic effects,
except at very high levels that we do not expect to find in the
Baltimore area. Second, we must emphasize again that the results
presented in this section are for setting priorities only; they are
not necessarily indicative of the total risks posed by all metals
in the environment.
Third, the metal ambient data, and thus our risk calculations,
may not be representative of actual exposures throughout the study
area. Finally, it is improbable that the monitors were situated
at the sites of maximum ambient concentrations in the Baltimore
metropolitan area; therefore, our estimates of lifetime risk to
the MEI may be understated.
INDOOR AIR POLLUTION
Recent studies have shown that indoor concentrations of many
substances in non-occupational settings particularly VOCs, may be
significantly higher than those found in the ambient air. As a
result, the Technical Advisory Committee expressed interest in
better understanding the risks from indoor air pollution and the
potential sources of these exposures. We gathered and reviewed
all available concentration data from national studies on indoor
air pollution. We did not locate any studies specific to the
Baltimore area; however, some of the results from these studies
are generalizable to the Baltimore area. We used collective ex-
pertise and judgement in estimating the importance of this issue
and whether it should be studied in Phase II. This approach serves
only to provide a rough indication of whether indoor air pollution
could be a problem in Baltimore. It cannot tell us what the actual
magnitude of the risk is in the Baltimore area. Our findings from
this effort are summarized below.
V-19
-------
Pollutants
For the purpose of our analysis, we concentrated on 12
compounds that were detected by recent studies of indoor air
pollution elsewhere in the nation. These substances were:
o Radon
o Asbestos
o Formaldehyde
o Benzene
o Chloroform
o Tobacco smoke
o Products of incomplete combustion (PICs)
o 1,1,1-Trichloroethane (methyl chloroform)
o Perchloroethylene
o Carbon tetrachloride
o Trichloroethylene
o Chlordane
Sources
Table V-15 lists the major sources and health effects' asso-
ciated with indoor air concentrations of the 12 target compounds.
There is also evidence indicating that volatilization from showers
may be a source of chloroform in the home. Chloroform and other
trihalomethanes are commonly found in most drinking water supplies.
Trihalomethanes are a by-product of the water disinfection process.
We discuss the risks from trihalomethanes in drinking water in the
next section.
Preliminary Screening Analysis
Using National Data
Cancer
Our review of existing studies on indoor air pollution indic-
cated to us that indoor air pollution could also be a problem in
the Baltimore area. We used the risk estimates for tobaccco smoke
and radon directly from the referenced reports and refer the reader
to these documents for a more thorough discussion of the risk esti-
mation techniques employed in these analyses.
V-20
-------
Compound
Radon
Asbestos
Formaldehyde
Benzene
Chloroform
Products of incomplete
combustion (NO , CO,
S02, hydrocarbons, hyd.
cyanide)
Tobacco smoke (PIC)
(arsenic, nicotine,
benzene, NOX, CO,
hydrocarbons, reapirable
particles)
1,1,1-Trichloroethane
Perchloroethylene
Carbon Tetrachloride
Trichloroethylene
Chlordane (pesticide)
Table V-15
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
POLLUTANTS, SOURCES, AND HEALTH EFFECTS ASSOCIATED WITH
INDOOR AIR POLLUTION FROM NATIONAL STUDIES
Source
Soil gas, building materials, well
water
Insulation and speckling compounds,
building materials (roofing and floor-
ing products)
Building materials (particleboard,
plywood, paneling), UFFI (urea-
formaldehyde foam insulation), furnish-
ings (furniture, carpets, draperies),
fueled appliances (gas stoves, ovens,
end unvented space heaters), consumer
products (room deodorizers, paper
products), tobacco smoke
Cleaning solvent, lacquer and cement
solvents, paint remover
Solvent, aerosol propellent
Gas stoves, kerosene heaters, unvented
gas heaters
Smokers
Cleaning and degreasing solvent,
furniture polish, drain cleaners
Dry cleaner solvent, cleaning and
degreasing solvent, fabric scouring
solvent
Cleaning and degreasing solvent,
pesticide ingredient
Cleaning and degreasing solvent, fabric
scouring solvent
Terniticide
Health Effects
Lung cancer
Lung cancer, meaoethelioma (rare
cancer), gastrointestinal tract cancer,
asbestosis (fatal scarring of lung),
variety of neoplaatic diseases
Suspect carcinogen, headaches, nausea,
dizziness, tearing and eye irritation,
lower respiratory irritation, pulmonary
edema, possible effect on central
nervous system (CNS)
Carcinogen (particularly leukemia),
blood and bone marrow disorders, CNS
disorders, chromosomal, aberrations
Suspect carcinogen, fetal developmental
effects, liver toxicity, fatigue
Respiratory illness, systemic toxicity
Lung cancer; oral cancer; larynx and
esophagus cancer; lower respiratory
illnesses; eye, nose, and throat irri-
tation; changes in heart rate, systolic
blood pressure, psychomotor functions,
small airway disfunctions
Suspect carcinogen; mutagen; death;
cardiovascular effect; eye, nose, and
throat irritation; nausea; light-
headedness
Suspect carcinogen, death, CNS depres-
sant, liver and kidney effects, cardio-
vascular effects, dizziness, headaches,
fatigue
Suspect carcinogen, death, cardio-
vascular and respiratory effects, liver
and kidney effects, headaches
Suspect carcinogen, death, CNS depres-
sant, decreased psychomotor function,
headaches, fatigue
Suspect carcinogen, CNS disorders, toxi
buildup in body tissues (liver, kidney,
and adipose tissue)
Note: Some of the health effects information is based on experimental animal studies.
-------
For the other compounds, we used professional judgement in evalua-
ting evaluating whether the reported measured indoor concentrations
could pose a health problem should they also be at similar.con-
centrations in homes in the Baltimore area.
All concentration values for these compounds are averages,
with the exception of formaldehyde. The range in formaldehyde
concentrations represents typical levels measured in residences
throughout the country. Again, we must emphasize that the concen-
tration data used in this analysis were based on national studies
and are not specific to the Baltimore area.
It is important to note that EPA's current policy does not regard
1,1,1-trichloroethane (TCA) as a human carcinogen. EPA previously
considered TCA a possible carcinogen, but has suspended that
classification pending a key study currently under review. Because
of the uncertainty surrounding TCA, and the ongoing debate over its
carcinogenicity, we have included TCA as a carcinogen for the pur-
poses of sensitivity analysis only.
Based on limited U.S. data, EPA's office of Radiation
Programs also believes that perhaps a million homes throughout
the country could exceed an upper-bound estimate of 3 to 4 in
100 lifetime risk from exposure to radon. Homes in Pennsylvania,
especially the Reading Prong area, have been observed with radon
concentrations as high as 15 WL (working levels, the unit of
measure for radon decay products), or about a 70 to 90 in 100
lifetime cancer risk.5
While we were unable to locate information on radon levels
specific to Baltimore, radon exposures could be a concern within
'the study area. As- indicated in Table V-15, radium in soil and
rock is a. major source of exposure. Radon flows through the
-soil and enters homes through cracks or openings in foundation
floors or walls. Additional research is needed to better charac-
terize the potential for radon exposures in the Baltimore study
area.
The upper-bound estimate of average lifetime individual
cancer risks from exposure to indoor tobacco smoke (i.e., passive
smoking) could also be as high as in 1 in 1000 lifetime cancer
risk. The risk estimates were based on a study of lung cancer
mortality rates in nonsmoking women with smoking husbands conducted
by the National Cancer Center Research Institute in Tokyo. These
potential exposures are of concern to nearly every individual
throughout the country, including the Baltimore study area.
Nearly everyone has been exposed to tobacco smoke at one time or
another. In addition, roughly 33 percent of the adult population
regularly smokes cigarettes, thus exposing the families, especially
children, of these individuals to tobacco smoke on a regular
basis in the home.6 The data obtained from the Total Exposure
V-21
-------
Assessment Methodology (TEAM) study that EPA conducted in Bayonne
and Elizabeth, New Jersey during 198i7 indicated that indoor air
concentrations of benzene, chloroform, 1,1,1-trichloroethane, and
carbon tetrachloride can get quite high relative to outdoor levels.
While it is impossible to extrapolate directly from these city-
specific data to the Baltimore area, one would expect to find some
indoor levels of volatile organic compounds in homes and buildings
throughout many areas of the country, including Baltimore. These
substances are commonly found in numerous household products, such
as personal care products, cleaning materials, paints, and pesticides.
Although indoor levels of formaldehyde may not be a concern
for all homes throughout the countryfor example, the TEAM study
did not detect statistically significant concentrations of formald-
ehyde in any of the homes sampled in the 1981 New Jersey study
formaldehyde is released by sources that could be found in various
homes in the Baltimore area. Formaldehyde is contained in building
materials (e.g., plywood and particleboard), furnishings (e.g.,
draperies and carpets), and some types of foam insulation containing
formaldehyde resins, especially urea formaldehyde. Unvented gas
combustion and tobacco smoking are other sources of indoor formald-
ehyde .8
Noncancer
The data from these studies suggest that indoor air concentra-
tions may under certain circumstances be high enough to cause
increased concern for fetal developmental effects from exposures to
benzene and chloroform. There is also an increased concern for
liver effects from exposure to indoor air concentrations of TCA.
There is also an increased concern for noncancer health
effects from indoor exposures to tobacco, and formaldehyde. Many
substances in cigarette smoke are irritants, and nonsmokers exposed
to cigarette smoke have experienced such symptoms as conjunctival
irritation, nasal discomfort, coughing, sore throats, and sneezing*
Some studies have also measured changes in heart rates, systolic
blood .pressure, and psychomotor functions in nonsmokers exposed to
tobacco smoke. Finally, most studies have indicated a strong
correlation between reported respiratory problems in children,
especially under two years of age, and parental smoking.^
In the range of typical formaldehyde levels measured in
selected homes throughout the country (.01 to .5 ppm), there
could also be an increased concern for noncancer health
effects. Formaldehyde has been shown in chamber studies with
humans to be an eye irritant at concentrations ranging from 0.1
V-22
-------
to 0.4 ppm. In addition, some studies have suggested that formald-
ehyde levels of approximately 1 ppm can affect the central nervous
system, although the results from these studies are difficult to
interpret.10
Limitations
There are several important limitations that must be considered
when interpreting the results presented in this section. First,
the indoor air risk calculations were based on data from selected
studies not specific to Baltimore; it is unclear whether these data
on indoor concentrations are directly applicable to the Baltimore
area. Second, our usage of average indoor concentration levels may
not be representative of actual exposure levels faced by many
individuals. Finally, there is equivocal evidence supporting the
carcinogenicity of TCA.
TRIHALOMETHANES IN DRINKING WATER
Trihaloraethanes in drinking water is the final issue of interest
identified by the Human Health Subcommittee. This section presents
our findings on the risks to human health associated with trihalom-
ethanes (THMs) in drinking water. It also reports the risks for
other compounds found in finished drinking water, specifically
pesticides and heavy metals.
Pollutants
We felt that important insights on the relative contribution
of trihalomethanes versus other pollutants to human health risks
could be provided by.looking at THMs, pesticides, and metals. On
the basis of water quality data for Baltimore's two drinking water
treatment plants, Ashburton and Montebello, we analyzed the following
pollutants:
o Pesticides - Endrin, Lindane, Methoxychlor,-
Toxaphene, 2,4-D, 2,4,5-TP
o Total trihalomethanes (chloroform)
o Heavy Metals - Arsenic, Barium, Cadmium, Chromium,
Copper, Lead, Mercury, Silver, and Selenium Sources
Sources
Most levels of pesticides and herbicides found in drinking
water are released - from nonpoint sources, particularly agricul-
tural run-off. Most metals are generally from background sour-
ces (i.e., soil), but some may also be released from urban run-
off. In contrast, trihalomethanes are present in finished
drinking water primarily as a. by-product of the water disinfec-
tion process. Disinfection of water is conducted to kill
V-23
-------
microbial pathogens that can, if' unchecked, lead to periodic
oubreaks of diseases such as cholera or typhoid. The disinfec-
tant, generally chlorine, reacts with organic matter in the water
to produce trihalomethanes, particularly chloroform.
Ambient Levels
Table V-16 presents the ambient water data for the Ashburton
and Montebello drinking water treatment plants for the years 1980
through 1983. We obtained these data from the City of Baltimore.
For our assessment of the exposures associated with ingestion of
drinking water, we used an ambient concentration value that repre-
sented the mean of the average values reported for 1981 through
1983. For our analysis of risks, we assumed that all trihalomethane
concentrations were composed of chloroform because past studies
have shown that chloroform accounts for the majority of the triha-
lomethane levels measured.
Table V-16 also indicates the Maximum Contaminant Levels (MCLs)
established by EPA for each of the pollutants listed. As is clear
from the table, the concentrations of all compounds are well within
these standards. The average trihalomethane (THM) level for each
year is slightly more than half of the maximum allowable THM con-
centration. Nevertheless, given our current understanding of the
health risks from exposure to THMs, the present MCL may no longer
be deemed protective of public health.
It is important to recognize that concentrations of chloroform,
in particular, may increase in the distribution system over time
even after the water has left the water treatment plant. This is
because residual chlorine may continue to react with organic matter
to form more chloroform. As a result, our risk calculations based
on average ambient levels do not capture the actual risks that
some individuals may face. Furthermore, since we only have data
on average ambient levels, we can only estimate lifetime risks to
the average exposed individual.
Exposed Population
The exposed population in our assessment of drinking water
risks is the population served by each of the drinking water
treatment plants. The Montebello plant serves a population of
approximately 700,000; Ashburton serves about 900,000 residents.
This includes all of the population in Baltimore City, 89% of
the population in Baltimore County, and a small number of people
in Anne Arundel County.
Unit Risk Factors
As in the previous analyses, we assessed the possible
cancer risks from ingestion of Baltimore water, as well as the
V-24
-------
Table V-16
BALTIMORE .IEMP PRELIMINARY RISK SCREENING RESULTS
WATER QUALITY DATA
ASHBURTON TREATMENT PLANT FINISHED WATER
1980 1981 1982
Pesticides (ug/1)
Endrin
Linda ne
Methoxychlor
Toxaphene
Herbicides (ug/1)
2,4-0
2,4,5-TP
Total trihalomethanes
(TTHM)
Heavy Metals (mg/1)
Arsenic
Barium
Cadmium
Chromium
Copper
Lead
Mercury
Sil ver
Selenium
IMCL = Maximum contaminant
^For all compounds, except
limits.
MCL1
0.2
4.0
100
5.0
100
10
100
0.05
1.0
Q.01
0.05
0.05
0.05
0.002
0.05
0.01
level
total
Average Range Average
<0.04 . <0.04
<0.01 <0.01
<1.0 <1.0
<1.0 <1.0
<0.5 <0.5
<0.03 <0.03
52
<0.01
0.02
<0.001
<0.001
<0.01
<0.01
< 0.001
<0.001
< 0.005
(EPA Interim Final Drinking
trihalomethanes and barium,
Range Average Range
<0.04
<0.01
<1.0
<1.0
<0.6
<0.03
37-68 55 35-72
<0.01
0.02
<0.001
< 0.001
<0.01
<0.01
<0.001
< 0.001
<0.005
1983
Average
56
0.001
0.03
0.002
<0.001
<0.01
<0.01
<0.001
< 0.001
0.002
Water Standards; source: EPA Primary
pollutant concentrations are reported
Auc
Range 198:
<0.
<0.
<1.
<1.
<0.
<0.
34-79 54
<0.
0.
<0.
<0.
<0.
<0.
<0.
<0.
<0.
Water Rules)
at detection
-------
potential for noncancer health effects from drinking water exposures.
As discussed above, we estimated cancer and noncancer risks for
trihalomethanes assuming that total concentration were chloroform.
Cancer
Table V-17 lists the upper-bound cancer unit risk values (for
ingestion) for the pollutants measured in the Baltimore drinking
water. In Phase I, only chloroform was evaluated. In this section,
however, we present the risks associated with other compounds
present in drinking water to provide perspective. The current unit
risk factors used for these substances, as well as the revised
potency value for chloroform, were all developed by CAG.
Noncancer
Table V-18 lists the noncancer health effects for the measured
pollutants in drinking water, as well as the no-effect thresholds
for ingestion. We were able to obtain these data for 9 of the 16
compounds. We were unable to assess risks from lead because the
dose-response data for this element are currently undergoing internal
review. What is clear, however, is that low levels of exposure to
lead can lead to adverse reproductive and neurological effects in
the unborn and children and hypertension in adults.H
We were, however, able to gather qualitative information
regarding situations in which lead levels could be above EPA's
Maximum Contaminant Levels (MCL), which are established on the
basis of non-cancer effects.
Preliminary Screening Analysis and
Risk Assessment Results
Using the data described above, we estimated the possible
cancer and noncancer risks associated with ingestion of Baltimore
drinking water. We calculated the upper-bound annual cancer incid-
ence for each drinking water treatment plant separately. We perform-
ed these computations by first multiplying the average concentration
for each pollutant by the appropriate unit risk factor and the
population served by the plant. We then divided by 70 to derive
our estimate of annual cancer incidence.
Because of the availability of only average ambient data, we
could only estimate upper-bound lifetime cancer risks to the average
exposed individual. We estimated these risks by simply multiplying
the average ambient concentration by the appropriate unit risk
factor. For noncancer health effects, we compared the average
ambient concentrations with the no-effect thresholds. Our specific
findings for each measure of risk follow.
V-25
-------
Table V-17
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND CANCER UNIT RISK VALUES:
POLLUTANTS IN DRINKING WATER
Pollutant
Chloroform (total
trihalomethanes)
Barium
Ingest ion
»««
1985
Analysis
Revised
19862
1.0 x 10-6 2.3 x 10~6
Grouping
Based on
EPA Criteria3
B2
"Current EPA policy does not regard barium as carcinogenic through
ingest ion. As a result, we could not perform a risk assessment for
this compound.
unit risk factor presented in this column was developed by CAG in
1984 and used in the preliminary Phase I risk calculations.
2The unit risk factor presented in this column is current as of 5/7/86.
^EPA weight-of-evidence classifications:
A s human carcinogen; B2 = probable carcinogen; C = possible
carcinogen. (See Appendix A for more detail.)
-------
Table V-18
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
NONCANCER HEALTH EFFECTS AND. PRESUMED HUMAN THRESHOLDS:
POLLUTANTS IN DRINKING MATER1
Pollutant
Arsenic
Cadmium
Chloroform (total
trihalomethanes )
Chromium +3
Chromium +6
Copper
Lindane
Mercury
Selenium
Toxaphene
Endrin
Methoxychlor
2,4-0
2,4,5-TP
Bariun
Lead2
Cil war
J±l VCF
N.A. = Mot available.
Health Effect
Fetal Developmental
Liver
Neuro behavioral
Vascular Disorder
Hyperpigment and
Keratoses
Chronic Poisoning
Reproductive
Fetal Developmental
Liver
Kidney
Reproductive
Fetal Developmental
Liver
Neurobehavioral
Kidney
Fetal
Liver
Kidney
Fetal Developmental
Reproductive
Dietary Problem
Liver
Neurobehavioral
General Toxicity
Fetal Developmental
Fetal Developmental
Neurobehavioral
Fetal Developmental
Liver
Neurobehavioral
Selenosis
Kidney
Reproductive
Liver
Reproductive
__
IT he threshold values presented in this table
2The dose-response 'data
available for our use.
for lead are currently
RID-Derived -
Threshold
(ug/1) - Source
133 Hood et al. 1977
133 Clement-Iron & Steel Study
133 Clement-Iron & Steel Study
133 MHO 1981
Perry et al. 1948; NIOSH 1975
133
1,070 Zaldivar 1977
133 Hood et al. 1977
874 Schroeder & Kitchener 1971
17.5 Friberg 1950
17.5 ADI 1985; Kjellatrom et al. 1977
4,090 Scharpf et al. 1972
795 Thompson et al. 1974
699 Heywood et al. 1979; ADI 1986
117 Challen et al. 1958; USE PA 1985
225 Heywood et al. 1979
35,000 Clement-Iron & Steel Study, Unverified
ADI
35,000 Clement-Iron & Steel Study, Unverified
ADI
35,000 Clement-Iron & Steel Study
170 Gale 1978
170 Gale 1978
149,000 Venugopal et al. 1978
11.0 ADI 1986
11.0 Doc. 4 EPA
11.0 Doc. 4 EPA
11.0 Earl et al. 1973
11.0 EPA 1980
11.0 ADI 1985; EPA 1980
105 Schroeder et al. 1971
105 NAS 1976; EPA 1980
105 Smith et al. 1936; Harr et al. 1967
105 Venugopal & Lucky 1979; ADI 1986
105 Venugopal & Lucky 1979
105 Schroeder & Kitchener 1971; NAS 1976
43.7 NAS 1977
131,000 Deichmann 1972
N.A.
N »A»
N.A.
N.A.
M A ^f^t^fn*
N.A.
N*
n« .- _ ^^
were developed by RID and are current as of 5/7/86.
undergoing internal review and are, therefore, not
-------
Cancer Incidence
Table V-19 presents the upper-bound estimates of annual
excess cancer incidence from exposure to drinking water treated
by the Ashburton and Montebello plants. Using the 1985 unit
risk factors, the chloroform ambient levels accounted for an
upper-bound estimate of less than one excess annual cancer case
from ingestion of either Ashburton or Montebello water.
Using the revised 1986 unit risk factors for chloroform
resulted in a revised upper-bound estimate of annual cancer
cases of roughly 2 excess cancer cases per year for Ashburton,
and about 1 excess annual cancer cases for Montebello.
Chloroform accounts for all of the predicted excess annual
cancer cases from ingestion of the drinking water from Ashburton
and Montebello. No other monitored pollutants of carcinogenic
concern were detected. We have assumed that the actual concentrations
for these were neglible for the purpose of estimating the upper-bound
cancer risk from drinking water. However, were these pollutants
pr-esent in the drinking water at levels equivalent to their respective
levels of detection, they would, in total, account for an additional
excess 0.5 cancer cases at Ashburton and 0.4 cancer cases for
Montebello beyond those cited above (using the latest, and higher,
GAG cancer potency values).
Please note, however, that only a small number of carcinogenic
compounds were monitored in the drinking water. Consequently, we
cannot rule out the possibility that we have overlooked other
chemicals posing possible cancer risk. The statements we make can
only relate to compounds for which we have actual monitoring data.
Lifetime Cancer Risks to the Average Exposed Individual
Our estimates of the upper-bound lifetime cancer risks to the
AEI are detailed in Table V-20. The Phase I calculations show an
upper-bound lifetime cancer risk to the AEI from chloroform in
drinking water of roughly 5 chances in 100,000 at each of the
drinking water treatment plants (Ashburton and Montebello). The
revised unit risk factor (1986) for chloroform increased these risk
estimates to approximately 10 chance in 100,000 from ingestion of
drinking water at each of the two drinking water treatment facilities.
Neglible risk is attributed to the other chemicals monitored in the
drinking water. As stated above, our statements regarding risk
apply only to those compounds for which we have monitoring' data.
We also must emphasize that chloroform ambient levels at each of
these plants are well within the MCL.
Because this is a screening exercise designed to identify
issues of potential concern, we have also estimated the risk from
these monitored chemicals should they be present at their respective
levels of detection. The upper-bound lifetime cancer
risks to the AEI for the remaining compounds ranges from a low of
about 3 chances in 10 million for lindane to a high of more than 3
chances in 10,000 for toxaphene at each treatment plant. Cumulative
V-26
-------
Table V-19
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF ANNUAL EXCESS CANCER INCIDENCE:
POLLUTANTS IN BALTIMORE DRINKING-WATER1,2
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
Ashburton3 Montebello3
Upper-Bound Annual Upper-Bound Annual
Cancer Cases Cancer Cases
Pollutant Average Average
(weight of Concentration 1985 Revised Concentration 1985 Revised
evidence)4 (ug/1)5 Analysis6 19B67 (ug/1)5 Analysis6 1986 '
ChloroformS (B2) 54.3 0.7 1.6 49.3 0.5 1.1
Total 0.7 1.6 0.5 1.1
*THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE ASSUMPTIONS THAT GENERALLY PRODUCE
UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS
EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK ESTIMATES WERE CALCULATED AS AIDS TO POLICY DEVELOPMENT,
NOT AS PREDICTIONS OF ACTUAL CANCER RISKS IN BALTIMORE. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER; IN FACT,
THEY COULD BE ZERO. THE PROPER FUNCTION OF THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE
ISSUES AND SET PRIORITIES FOR THE TOPICS EXAMINED.
2RID'S ESTIMATE OF THE ACTUAL NUMBER OF CANCER CASES IN THE STUDY AREA IN 1984 IS 8,000 CASES. (SEE II-8 AND
9.) THIS NUMBER SHOULD SERVE ONLY AS A POINT OF REFERENCE IN UNDERSTANDING THE RISK ESTIMATES PROVIDED. IN
ADDITION, THE RISK ESTIMATES SHOULD NOT BE INTERPRETED AS REPRESENTING THE TOTAL UPPER-BOUND CANCER RISKS FROM
ALL POLLUTANTS IN ANY PARTICULAR MEDIUM. THEY DO NOT TAKE INTO ACCOUNT ALL POLLUTANTS THAT MAY BE PRESENT IN
THE MEDIUM, ALL SOURCES OF THESE POLLUTANTS, AND ALL EXPOSURE SCENARIOS INVOLVING PATHWAYS OR EXPOSURES OF
SHORT DURATION TO RELATIVELY HIGH DOSES.
^Ashburton serves a population of 900,000; Montebello serves a population of 700,000.
*EPA weight-of-evidence classification: B2 = probable carcinogen. (See Appendix A for more detail.)
^Measure of .pollutant concentration in finished water over 1981-1983, City of Baltimore Drinking Water Quality
Data (Versar, 1984).
^Chloroform risk calculation based on the unit risk factor available in 1985.
'Chloroform risk calculation based on the most current unit risk factor (5/7/86).
^Analysis assumes total trihalomethane concentration is chloroform.
-------
Table V-20
BALTIMORE EMP PRELIMINARY RISK SCREENING RESULTS
UPPER-BOUND ESTIMATES OF LIFETIME CANCER RISKS TO
THE AVERAGE EXPOSED INDIVIDUAL:
POLLUTANTS IN BALTIMORE DRINKING MATER1
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
Ashburton Montebello
Upper-Bound Risks Upper-Bound Risks
to the AEI2 to the AEI2
Pollutant 1985 Revised
(weight of evidence)3 Analysis4 j 19865
Chloroform6 (B2) 5.A x 10"5 1.2 x 10~4
4.9 x 10~5 1.1 x 10~*
XTHE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE ASSUMPTIONS
THAT GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND
METHODS IN SEVERAL AREAS OF THE ANALYSIS, SUCH AS EXPOSURE CALCULATIONS AND
POLLUTANT SELECTION, RISK ESTIMATES WERE CALCULATED AS AIDS TO POLICY DEVELOP-
MENT, NOT AS PREDICTIONS OF ACTUAL CANCER RISKS IN BALTIMORE. ACTUAL RISKS MAY
BE SIGNIFICANTLY LOWER; IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF THE
ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND SET PRIORI-
TIES FOR THE TOPICS EXAMIhCD.
^City of Baltimore Drinking Water Quality Data (Versar, 1984) provided the average
pollutant concentration in finished water.
5EPA weight-of-evidence classification: B2 = probable carcinogen. (See
Appendix A for more detail.)
\)nit risk factor utilized in risk calculation Mas developed in 1985.'
^Unit risk factor utilized in risk calculation is current as of 5/7/86.
^Analysis assumes total trihalomethane concentration is chloroform.
-------
upper-bound lifetime risks to the AEI for these other compounds are
close to 4 chances in 10>000 from - ingest ion of Ashburton and
Montebello water separately.
Noncancer Effects
The average ambient concentrations of the pollutants considered
in drinking water at the distribution plants are well below the
no-effect thresholds for noncancer health effects. However, a
more definitive analysis of noncancer effects could be performed
if we used data showing the range of pollutant concentrations
throughout the drinking water distribution system. We did not have
these data for this analysis.
Nevertheless, a qualitative assessment of the potential for
contamination of drinking water from lead solder and flux in the
distribution system indicated that, in areas of corrosive ground
water, lead may occur at levels in excess of the 50 ug/1 MCL estab-
lished by the EPA.12 Studies by Carroll and Montgomery Counties
showed that new plumbing, either alone or in combination with
corrosive water, may also cause lead levels well in excess of the
drinking water standard.I3 In Carroll County, 78 out of 350 (22%)
of random samples of tap water taken in the early morning showed
levels above 50 ug/1. Five percent of the samples were above
the drinking water standard when water was in normal daily use.
Levels ranged up to 500 ug/1.1^
Limitations
There are three major limitations to our analysis of the
risks from ingestion of drinking water. The first area of uncertain-
ty results from our usage of average ambient data. As emphasized
in our preceding discussions, average values do not permit a thorough
assessment of the range of potential concentrations to which indiv-
iduals are exposed. They also do not permit an analysis of the
risks to the most exposed individual or the increased risks of
noncancer health effects. For example, average drinking water
values measured at the treatment plant may not be representative of
ambient levels of chloroform in people's tap water because chloroform
can continue to be formed in the distribution system.
The second area of uncertainty in our analysis is the assumption
that total trihalomethane levels are entirely due to chloroform.
While chloroform is often found in drinking water, other trihalom-
ethanes, such as dichlorobromomethane, chlorodibromomethane, and
bromoforra can also be present in finished drinking water. If the
potency of these compound is less than that for chloroform we have
overstated the risks from trihalomethanes in drinking water. On
V-27
-------
the other hand, if the toxicity of these substances is greater than
that for chloroform, we may haye underestimated the total risks
posed by drinking water.
The final limitation relates to the relatively small number
of substances for which we have monitoring data. Our statements
regarding risk can apply only to those chemicals. Nevertheless,
our inability to detect the substances, other than the trihalometh-
anes, that are the most likely chemicals to contaminate drinking
water (should it become contaminated) supports us in our belief
that we have not overlooked the most important risks.
RISK COMPARISONS
Having presented separately our findings for the four issues
of interest, it will be useful to summarize and compare the risk
data by pollutant and route of exposure for three of these specific
to Baltimore (the estimated indoor air risks were based on data
from other cities throughout the country and are not representative
of the Baltimore area). We have compiled summary tables for the
three measures of risk considered: aggregate cancer incidence;
lifetime individual cancer risks; and the potential for noncancer
effects. Such comparisons provided guidance in developing targeted
research agendas for Phase II.
Baltimore Risk Comparison
Cancer Incidence
Table V-21 presents our upper-bound estimates of annual cancer
incidence for the Phase II priority issues using the most current
unit risk factors. It is difficult to compare the cancer risks
from exposures to ambient air and drinking water because of the
uncertainty surrounding the chromium risk estimates. If we assume
that there is no hexavalent chromium, the excess cancer incidence
from air and drinking water exposures is somewhat similar (an
upper-bound estimate of roughly 3 excess annual cases in air versus
an upper-bound estimate of almost 3 excess annual cancer cases in
drinking water). However, if we assume that total chromium levels
are composed entirely of hexavalent chromium, the cancer risks
associated with air toxics is more thanx twice the cancer risks
from ingestion of contaminated drinking water.
From the standpoint of pollutant contribution, there are
several interesting observations. Of the pollutants studies in
air, the pollutants of greatest concern are chromium and benzene.
Second, of the substances examined in Baltimore area (surface)
drinking water, only chloroform (THMs) posed potential risks.
Third, while no single pollutant dominates both routes of exposure,
chloroform appears to be relatively significant for air and drinking
V-28
-------
Table V-21
BALTIMORE IEMP PRELIMINARY RISK SCREENING RESULTS1,2
COMPARISON OF UPPER-BOUND EXCESS" ANNUAL CANCER
INCIDENCE ACROSS ISSUES AND POLLUTANTS IN BALTIMORE
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
(1986 analysis)
Drinking
Compound (weight of evidence)^ Air Water
Volatile Organic
Compounds
Benzene (A) 1.80
Trichloroethylene (B2) 0.02
Perchloroethylene (B2) 0.10
l,2-Dichloroethane(B2) 0.10
Chloroform (B2) 0.40 2.7
Carbon Tetrachloride (B2) 0.30
1,2-Oichloropropane (C) 0.10 ' -
2.82 2.7
Chromium (hexavalent)4 (A) 0.00 to 4.2
Cadmium5 (Bl) 0.00 to 0.04
Subtotal 0.00 to 4.24
TOTAL6 2.8 to 7.1 2.7
XTHE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON
CONSERVATIVE ASSUMPTIONS THAT GENERALLY PRODUCE UPPER-BOUND
ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS IN
SEVERAL AREAS OF THE ANALYSIS, SUCH AS EXPOSURE CALCULATIONS
AND POLLUTANT SELECTION, RISK ESTIMATES WERE CALCULATED AS
AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL
.CANCER RISKS IN BALTIMORE. ACTUAL RISKS MAY BE SIGNIFICANTS
LOWER; IN FACT, THEY COULD BE ZERO. THE PROPER FUNCTION OF
THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE
ISSUES AND SET PRIORITIES FOR THE TOPICS EXAMINED.
2RID'S ESTIMATE OF THE ACTUAL NUMBER OF CANCER CASES IN THE
STUDY AREA IN 1984 IS 8,000 CASES. (SEE II-8 AND 9.) THIS
NUMBER SHOULD SERVE ONLY AS A POINT OF REFERENCE IN UNDER-
STANDING THE RISK ESTIMATES PROVIDED. IN ADDITION, THE RISK
ESTIMATES SHOULD NOT BE INTERPRETED AS REPRESENTING THE TOTA
UPPER-BOUND CANCER RISKS FROM ALL POLLUTANTS IN ANY PARTICU-
LAR MEDIUM. THEY DO NOT TAKE INTO ACCOUNT ALL POLLUTANTS
THAT MAY BE PRESENT IN THE MEDIUM, ALL SOURCES OF THESE
POLLUTANTS, AND ALL EXPOSURE SCENARIOS INVOLVING PATHWAYS OR
EXPOSURES OF SHORT DURATION TO RELATIVELY HIGH DOSES.
%PA weight-of-evidence classifications: A = human carcino-
gen; Bl, 82 = probable carcinogen; C = possible carcinogen.
(See Appendix A for more detail.)
^Chromium incidence calculations indicate a range of possible
ambient levels of hexavalent chromium from 0 percent to
100 percent.
'Cadmium incidence calculations indicate a range of possible
ambient levels from 0.0 ug/ra3 to detection limits (between
.001 and .002 ug/ra3).
^Numbers have been rounded to one significant decimal.
-------
water. It accounts for all of the estimated excess annual cancer
incidence from those pollutants analyzed in drinking water. In the
ambient air, it accounts for roughly 14 percent of the estimated
incidence for the subset of organics studied.
Lifetime Cancer Risks
Our summary of the highest upper-bound lifetime cancer risks
for the priority issues is shown in Table V-22. The increased
upper-bound lifetime cancer risks to the average exposed individual
for drinking water (chloroform) are estimated to be over 100 chances
(110 for Montebello and 120 for Ashburton) in a million. The
exposed population is roughly 1.6 million. In air, the upper-bound
lifetime cancer risks for the maximally exposed individual for
benzene and chloroform are both around 100 chances in a million.
(100 for benzene and 110 for chloroform; whoever, the exposed
subpopulation is relatively small compared with drinking water
(roughly 1.6 million in drinking water compared to less than 50,000
for any one chemical in the ambient air). The increased upper-bound
lifetime cancer risks for chromium in the ambient air range from
negligible (0 percent hexavalent chromium) to 360 chances in a
million (100 percent hexavalent chromium). The exposed subpopulation
for metals is larger than for organics, approximately 0.5 million.
Noncancer
Finally, Table V-23 summarizes our findings on the increased
potential for noncancer health effects from organics through the
different routes of exposure considered in Baltimore. The results
from our analysis suggest that noncancer health effects are of
concern for exposures to organics in the ambient air. The ambient
levels of the metals evaluated, cadmium and chromium, were below
the no-effect thresholds. Similarly, the pollutant concentrations
in drinking water monitored at the distributions points were below
the estimated no-effect thresholds for ingestion.
Of the pollutants evaluated, only two substances appear to
present an increased risk of noncancer effects at their measured
levels: benzene and chloroform. Benzene and chloroform concen-
trations present an increased risk of fetal developmental effects
from ambient air. We estimated the exposed subpopulation by gathe-
ring birth rate statistics in the study area. In 1984, there were
27,426 births; the total number of fetuses that were at risk is
unknown. These same pollutants were not found to be a problem in
drinking water.
Lead, because of its ubiquity in the urban environment, must
be discussed separately from the other metals. Lead is present
in the ambient air, drinking water, street and household dust,
and food. The extent to which any single route contributes to
total exposure in the Baltimore area has yet to be evaluated,
but it is clear that all routes must be considered in assessing
V-29
in I1
-------
Table V-22
BALTIMORE IEMP PRELIMINARY RISK AND SCREENING RESULTS
COMPARISON OF UPPER-BOUND LIFETIME CANCER RISKS IN BALTIMORE1,2
PHASE I RESULTS INTENDED FOR DEVELOPMENT OF THE
PHASE II RESEARCH AGENDA
(1986 Analysis)
Ambient Air
Drinking Water
Compound (weight of evidence)
Volatile Organic Compounds
Benzene (A)
Trichloroethylene (B2)
Perchloroethylene (82)
1,2-Dichloroethane (B2)
Chloroform (B2)
Carbon Tetrachloride (B2)
1,2-Dichloropropane (C)
Metals
Chromium (hexavalent)' (A)
Cadmium6 (Bl)
Increased
Lifetime Risk
(chances
in a million)
100
5
5
68
110
21
36
0 to 360
0 to 4
Exposed
Population
48,771
48,771
14,270
12,880
23,997
23,997
16,848
490,690
118,000
Increased
Lifetime Risk
(chances
in a million)
Exposed
Population
110 (Montebello)
120 (Ashburton)
700,000
900,000
1THE UNIT RISK FACTORS USED IN THIS ANALYSIS ARE BASED ON CONSERVATIVE ASSUMPTIONS THAT
GENERALLY PRODUCE UPPER-BOUND ESTIMATES. BECAUSE OF LIMITATIONS IN DATA AND METHODS IN
SEVERAL AREAS OF THE ANALYSIS, SUCH AS EXPOSURE CALCULATIONS AND POLLUTANT SELECTION, RISK
ESTIMATES WERE CALCULATED AS AIDS TO POLICY DEVELOPMENT, NOT AS PREDICTIONS OF ACTUAL CANCER
RISKS IN BALTIMORE. ACTUAL RISKS MAY BE SIGNIFICANTLY LOWER; IN FACT, THEY COULD BE ZERO.
THE PROPER FUNCTION OF THE ESTIMATES IS TO HELP LOCAL OFFICIALS SELECT AND EVALUATE ISSUES AND
SET PRIORITIES FOR THE TOPICS EXAMINED.
2Riska in ambient air are presented as upper-bound individual lifetime cancer risks at the site
of maximum average concentration (not a true MEI risk calculation); risks in drinking water
are presented as cancer risks to the average exposed individual. Risks from indoor air are
also presented primarily as cancer risks to the average exposed individual, but are not .based
on data specific to Baltimore.
3EPA weight-of-evidence classifications: A = human carcinogen; Bl, B2 = probable carcinogen;
C s possible carcinogen. (See Appendix A for more detail.)
*The exposed population is in the grid at the monitoring site of maximum concentration.
^Chromium risk calculations indicate a range of possible ambient levels of hexavalent chromium,
0 percent to 100 percent.
Cadmium risk calculations indicate a range of possible ambient levels from 0.0 ug/m to the
upper end of the detection limit (about .002 ug/m5).
-------
Table V-23
BALTIMORE IE HP PRELIMINARY RISK SCREENING RESULTS
POLLUTANTS OF POSSIBLE REGULATORY CONCERN FOR NONCANCER HEALTH EFFECTS IN BALTIMORE
Population Expoaed
Potential Above No-Effect
Pollutant Exposure Pathway Health Effecta1 Threshold
Benzene Ambient air Fetal developmental *
Chloroform Ambient air Fetal developmental "*
*In 1984, there were 27,426 births. The total number of fetuses that did not come to terra
is unknown.
iflECAUSE OF SIGNIFICANT UNCERTAINTIES IN THE UNDERLYING DATA AND ASSUMPTIONS, THESE
ESTIMATES OF THE POPULATIONS POTENTIALLY AT RISK OF DISEASE ARE ONLY ROUGH APPROXIMATIONS.
THEY ARE BASED ON CONSERVATIVE ESTIMATES OF EXPOSURE AND CHEMICAL TOXICITY. UNLIKE CANCER
RISK ESTIMATES IN THIS REPORT, THESE ARE ESTIMATES OF POPULATIONS EXPOSED AT LEVELS THAT
MAY POSE HEALTH RISK} THEY ARE NOT ESTIMATES OF POSSIBLE NUMBERS OF CASES OR PROBABILITY
OF DISEASE.
-------
the importance of a particular medium to exposure. In isolated
circumstances, lead in drinking water in residential plumbing
containing lead solder and flux has reached levels where this
pathway of exposure alone poses significant risk to human health.
And there are documented cases of lead poisoning in children exposed
to lead in household dust165 hospitalizations alone in 1985.
SUMMARY OF RESULTS
Within the limitations discussed throughout this chapter, our
analyses have yielded the following results. The results have been
useful in focusing Phase II research to help State and local autho-
rities in the Baltimore area in setting priorities and in conducting
other activities associated with the management^ of environmental
risks.
Exposure Results
o Of the organic compounds evaluated in the ambient air, the
highest measured concentrations were for benzene, toluene,
and ethyl benzene.
o Results from the Baltimore IEMP air toxics monitoring pro-
gram indicated that there can be significant variation in
the concentrations of some organics and metals throughout
the Baltimore area. This is an important factor to consider
during Phase II when we will explore control options.
o Indoor air pollutant levels could be higher than ambient
air concentrations for the subset of pollutants evaluated
in the Baltimore IEMP. In some instances, indoor air con-
centrations may be ten times higher.
o Data on lead in drinking water suggest that levels may be
be higher than the drinking water standard for houses with
the new plumbing containing lead solder and flux and corro-
sive tap water. Between the time that our Phase I analysis
was completed and this report was prepared, the local plumb-
ing boards voted to prohibit lead (95%) solder and the
Maryland Legislature acted to ban lead solder in new plumbing
and to educate the public on the risks.
Results for Cancer Risks
o Exposure to a subset of organics found in the ambient air
in Baltimore could result in an upper bound estimate 6T
roughly 3 excess cancer cases a year. Benzene exposures
alone account for about two-thirds of the total excess
cancer incidence; perchloroethylene, chloroform and carbon
tetrachloride together contribute nearly one-third of the
total. Exposures to 1,2 dichloropropane, 1,2-dichloroethane,
and trichloroethylene account for the remaining amount of
the total excess cancer incidence estimated for the compounds
studied.
V-30
-------
trichloroethylene account ' for the remaining amount of the
total excess cancer incidence estimated for the compounds
studied.
o The estimated upper-bound risks from exposure to chromium
and cadmium in the ambient air range from roughly O.I ex-
cess cancer cases a year to over 4 excess cancer cases a
year. This range representsdifferingassumptionsabout
regarding the fraction of total chromium that is hexavalent
(1 percent to 100 percent). Because no cadmium was actually
detected in the ambient air, our risk estimates were based
upon the lowest concentration at which it can be detected.
Even under these very conservative assumptions (protective
of public health), it contributes very little to total risk
(0.06 upper-bound cancer cases annually).
o The estimated upper-bound excess cancer incidence from in-
gestion of drinking water could be almost 3 cases per year
All of the estimated cancer risks are attributable to
trihalomethanes (chloroform and related chemicals). The
levels of trihalomethanes are below the EPA drinking-water
standard for the substances.
o Preliminary risk calculations for indoor air using national
data suggest that exposures to indoor air pollutation may
be a relatively significant potential source of risk in the
Baltimore study area. The estimated risks for certain
chemicals could be ten-fold higher than those calculated
from ambient exposures.
Results for Noncancer Risks
o The preliminary Phase I findings suggest that most of the
Baltimore population will be exposed to chemicals at con-
.centrations high enough to justify regulatory concern
Benzene concentrations in the ambient air may increase the
risk of fetal developmental effects throughout the study
area. (In 1984, there were 27,426 births in the Baltimore
area; the total number of fetuses at risk is unknown. )
Similarly, ambient air concentrations of chloroform could
pose an elevated risk of fetal developmental effects, but
at only a select number of sites in the Baltimore area.
o There have been over 150 hospitalizations per year in
Baltimore of children for lead chelation therapy. Nearly
all of the cases of high levels of lead in the children's
blood were the result of their exposure to lead in dust in
older homes where lead paint had been used. To prevent
future cases of lead-poisoning from exposure to lead from
lead paint, the State of Maryland
V-31
-------
has issued a regulation that lead in contaminated homes
must be abated.
Items for Future EPA Consideration
o We have a fairly good understanding of the potential risks
for several pollutants, but not the sources of these pollu-
tants. Phase II data collection activities need to be
focused on better characterizing the sources of toxic
pollutants in the Baltimore area. This effort should
gather data sufficient to quantify pollutant loadings by
source type.
o Many of the cancer unit risk factors should be investigated
further. The cancer unit risk factors for 1,1,1-
trichloroethane, formaldehyde, and arsenic (indestion),
in particular, have been called into question.
o The significance of the risk from chromium depends upon its
valence state. Increased effort to resolve the current
uncertainty would be useful.
V-32
-------
1. While the monitoring program also sampled for methylene
chloride and 1,1,1-trichloroethane, the measured data were
generally found to be unquantifiable.because of
contamination during transport and handling.
2. Black & Veatch, Industrial Waste Control Program. Draft
. Final Report. City of Baltimore, Maryland Bureau of Water
and Wastewater, April 1984.
3. A more detailed discussion of the monitoring program can be
found in the following report: Versar, Inc., Quality
Assurance Plan for Monitoring and Analytical Activities to
evaluate Selected Air Pollutants in Metropolitan Baltimore.
Maryland. November 8, 1983.
4. National Toxicology Program, Technical Report on the
Carcinocrenicitv Bioassav of 1.2-Dichloropropane (CAS No."
78-87-5) in F 344/N Rats and B6C3F1 Mice fGavaae Study). May
1983, Draft (NIH Publication No. 83-2519).
5. Richard Guimond, "Indoor Radon: Briefing for Lee M.
Thomas," March 12, 1985, Office of Radiation Programs.
6. John D. Spengler and Ken Sexton, "Indoor Air Pollution: A
Public Health Perspective," Science. Vol. 221, No. 4605
(July 1, 1983), p. 11.
/
7. Memo from L. Wallace to B. Steigerwald, 10/30/84.
.Attachment B: Elizabeth-Bayonne, NJ personal exposure;
pertaining to Total Exposure Assessment Methodology (TEAM)
Study in New Jersey, 1981.
8. Spengler & Sexton, p.12; and J. Repace, Tobacco Smoke, the
Double Standard, a report from the Center for Philosophy and
Public Policy, University of Maryland, College .Park, Winter
1984.
9. Spengler and Sexton, pp. 11-12.
10. Spengler and Sexton, p. 13.
11. USEPA, Costs and Benefits of Reducing Lead in Gasoline;
Final Regulatory Impact Analysis. Chapters IV and V, Feb. 1985.
-------
12. Personal Communication between Hope Pillsbury, US EPA and
Katherine Farrell, M.D., Maryland OEP, January 1986.
13. John Love11, Richard Isaac, and Ruth Singer, M.D., Carroll
County Health Department, Westminster, Maryland, Control of
Lead and Copper in Private Water Supplies, Oct. 18, 1978.
14. John Love11 et al.
-------
CHAPTER VI
ANALYSIS OF SOURCES WITH POTENTIAL ADVERSE
IMPACTS ON GROUND-WATER
-------
CHAPTER VI. ANALYSIS OF SOURCES WITH POTENTIAL ADVERSE
IMPACTS ON GROUND-WATER
The TAG established a Subcommittee on Ground-Water Resources
and charged it with developing a method for setting priorities
among ground-water issues for consideration in Phase II analyses.
The TAG had developed a list of thirty-two issues for study in
Phase I (see figure IV-2 for list). The subcommittee was speci-
fically asked to identify from that list no more than three
issues that posed the greatest potential impact on ground-water
resources.
This chapter describes the analysis that the Subcommittee
conducted. After providing an overview of the method, it de-
scribes the background of the members and summarizes data that
were'collected. The next three sections describe the scoring of
ground-water resource impact, the results of the scoring, and the
limitations of the analysis. Finally it presents additional anal-:
yses conducted by the Multi-Media Metals Workgroup which the TAG
used to help evaluate the recommendations of the Ground-Water
Resources Subcommittee.
The Subcommittee made an initial important decision: to
focus on the overall impact on the availability (without treat-
ment) of the ground-water as a resource, rather than to focus on
the expected human health risk or ecological impacts from conta-
minated ground-water. Of course, consideration of ground-water
resource impact takes into account potential or actual exposure,
and therefore risk to humans. Had human health risk been the
only criterion however, the impacts from contaminated ground-
water would have been neglected in two types of situations. The
first is where ground-water is not currently used for drinking,
but could be in the future; the second is situations where the
health risks are abated by either treating the water or finding
alternative sources.
The Subcommittee also identified economic impact as an im-
portant aspect of ground-water degradation. Contaminated ground-
water loses some of its usefulness to humans. When ground-water
that is being used as drinking water becomes contaminated, it can
be treated to restore its usefulness or an alternate supply devel-
oped. Either alternative can be quite expensive. Even when
groundw.ater that becomes contaminated is not being used, the
contamination impairs its usefulness for future needs.
In summary, the Subcommittee on Ground-Water" Resources
decided that it needed measures other than health and ecological
risk with which to evaluate the importance of ground-water conta-
mination issues. Furthermore, it decided to focus specifically
on' potential impact to the ground-water resources.
VI-1
-------
OBJECTIVES
The TAG asked the Ground-Water Subcommittee to determine
which of the 32 issues were relevant to ground-water resource
impact, and then rank them according to their potential for
causing degraduation of ground-water resources. The subcommit-
tee first devised a "common currency" for comparing the severity
of impacts to the ground-water resource from a diverse group of
sources and pollutants. It was not necessary that the method be
absolutely precise in ranking problems; after all, the purpose
of Phase I of the Baltimore study was to identify a set of high
priority issues for further study in Phase II. The subcommittee
recognized that it was necessary, though, that the method be
flexible enough to allow an un-biased evaluation of the diverse
sources for which there were very different amounts of data.
OVERVIEW OF THE METHODOLOGY
The Ground-Water Subcommittee, composed primarily of local
environmental management professionals, developed an index to
evaluate and score different sources of ground-water contamina-
tion according to the impact the source or pollutant could have
on ground-water resources. These scores provided a relative
ranking for the sources. The subcommittee decided to rely pri-
marily on the judgments of experts and to use an index because
of the lack of technical data necessary to do more sophisticated
analysis. The subcommittee also«felt that the budget would not
allow collection of substantial amounts of these kinds of data.
Had more resources been available, we could^ have performed
ground-water modeling and/or monitoring to help us set priorities
among the sources of contamination.
o Data needed for modeling of quantitative effects (like that
done to assess impacts on human health) were not available.
For many important types of sources that could contaminate
ground-water [e.g., underground storage tanks' (USTs)]» even
the simplest sort of information on number and location of
facilities did not exist. For most types of sources there
were no data on failure/release/emission rates. And at the
time of Phase I, there was no model available that could
integrate source, 'release, and hydrogeological data to
yield an estimate of effects.
o Ground-water monitoring data were also inadequate to com-
pare problems from different source types. Data existed
for only a few pollutants. Monitoring reports were fairly
plentiful for some source types (landfills) while only
addressing the very worst cases for other source types
(UTS). New ground-water monitoring efforts would have been
prohibitively costly. Because contaminants can travel at
VI-2
-------
widely different rates in ground-water this lack of cer-
tainty would have necessitated- ax vast number of monitoring
wells to characterize the contamination plumes from diffe-
rent sorts of point sources.
The Subcommittee thus developed an index to be applied to each
source type to determine which kinds of activities are likely to
lead to the greatest adverse impact on ground-water resources.
The index had two basic components capturing different aspects of
the effect of a source on ground-water--pollution impact and eco-
nomic impact.
The Subcommittee designed the pollution impact component as
a very broad measure of the volume, severity, and general impor-
tance of ground-water contamination from different source types.
The pollution impact score reflects both current impact from the
sources and potential future dmpact. It gives higher scores based
on degree of current use of 'the ground-water and also gives high-
er scores to sources that contaminate ground-water not currently
used but possibly needed in the future. This decision reflected
the subcommittee's value judgment that both the present and
future use of ground-water is of concern.
The elements comprising the pollution impact score were:
o number of individual units represented by the source type
(e.g.v/ number of USTs),
o release volume over time from" each unit,
o concentration of contaminants in the material released,
o persistence of the released contaminants in the environment,
o existence of known contamination incidents,
o potential for future incidents,
o location of units in vulnerable hydrogeologic environments,
o size of population using- ground-water near the units,
o amount of industrial ground-water use near the units.
The economic impact component of the index was designed to reflect
a different aspect of the impact on ground-water resources. This
portion of the index focused exclusively on the economic costs
that the contamination from that source was likely to cause. In
contrast to the pollution impact score, the economic impact score,
rather than weighting all ground-water equally regardless of use,
depends heavily on how the water is currently used, the potential
for contamination by source type, and how costly it would be to
maintain that use, if contamination occurs. The elements included
in the economic impact score are:
o relative cost of programs to prevent contamination from
the source type; '
VI-3
-------
o relative cost of alternative clean water supplies for
those ground-water users likely to be affected by contami-
nation from the source type;
o relative cost of water treatment for those ground-water
users likely to be affected by contamination from the
source type.
For each source type, the Subcommittee scored all of these
elements individually and used a weighting.scheme to develop the
pollution score and the economic impact score. (See "Performing
the scoring of ground-water resource impact".) The two scores
were then added together to generate a final score. As a first
cut, pollution and economic impact were weighted equally. Addi-
tional sensitivity analysis was performed to see if different
weighting schemes would affect the results. (See "Results of the
ground-water resource impact rankings.") The higher the final
score, the greater the potential impact on ground-water resources
from the source category. A more detailed description of the
scoring system is included in the Addendum to this chapter.
BACKGROUND OF MEMBERS OF THE GROUND-WATER SUBCOMMITTEE
Because the application of the ground-water scoring method
relied so heavily on the judgment of the ground-water subcommit-
tee members, we include a brief outline of the background and
experience that they brought to this task. Emery Cleaves, the
Subcommittee Chairman, is the Principal Geologist of the Maryland
Geological Survey. As such, he has many years of experience with
geological issues in Maryland, and special knowledge of local
ground-water contamination incidents.
The two counties were represented by officials who knew
their counties' ground-water issues intimately. Colin Thacker
is the Director of Northern Environmental Services in the Balti-
more County Health Department, while N. Singh Dhillon is the
Director of Environmental Health for the Anne Arundel County
Department of Health.
Representing EPA was Hope Pillsbury who over the last
several years had been directing a major EPA project to assess on
a national basis the comparative risks to ground-water from diffe-
rent source types. She also had a Master's degree in environmen-
tal studies and a B.S. in geology.
DATA COLLECTION
The information used for the index and scoring procedure was
a combination of data collected for the project, and, where data
were not available, the judgment and experience of the Subcommit-
VI-4
-------
tee members. EPA consultants to the Baltimore project prepared
background memoranda on various ground-water issues to summarize
existing data before the scoring process began. Important con-
clusions from the consultants' memoranda follow.
Hydrogeology. Ihe Baltimore study area includes two
geological regions. The Piedmont area is located in the north-
west portion of the study area and the Coastal Plain is located
in the southeast portion. (See figure VI-1). As the map shows*
Baltimore County is largely Piedmont geology, and Anne Arundel
County is completely Coastal Plain geology. The Piedmont terrain
consists of crystalline rocks, including marble, schist, gneiss
and gabbro. The rocks have low porosity and permeability, with
water moving mainly along joints in the rock. Ground-water re-
charge occurs locally, and discharge occurs within the same
local watershed. Consequently any ground-water pollution may be
expected to occur in restricted local areas.
The terrain of the Coastal Plain consists of unconsolidated
sand, silt, clay, and some gravel. The sediments occur in con-
tinuous and extensive layers and lenses with aquifers of perme-
able sands and silty sands interdigitated with impermeable silty
clays and clays. The aquifers dip to the east and south-east
and, unlike the Piedmont, do not mimic surface watersheds. The
Coastal Plain aquifers have significantly greater porosity and
permeability, and are more vulnerable to contamination. ^'^ On
the other hand, it is easier to get water out of the aquifer and
thus easier to clean up than the Piedment.3
Water use. The bulk of the population in the study area is
in Baltimore City where there are few wells and the public system
relies on surface supplies. About 36% of the households lie in
Baltimore County. Slightly over 8% of Baltimore County residents
obtain drinking water from private wells, and the remaining peo-
ple obtain drinking water from public systems (which rely primar-
ily on surface water). Anne Arundel County includes 19% of the
study area population. Over 99% of these residents rely on
ground-water. Most industrial ground-water use occurs in the
industrial areas of Baltimore City.
Landfills. -There are seven active non-hazardous waste land-
fills, one active hazardous waste landfill, and at least sixteen
inactive landfills in the1 study area. Landfills contain and re-
lease a variety of pollutants. Monitoring has detected three
toxic pollutants in ground-water at landfills in the study area.
Chromium has been detected at the hazardous waste landfill at
Hawkins Point.* Benzene and trichlorethylene have been detec-
ted at the inactive Solley Road landfill. One portion of this
landfill accepted controlled hazardous substances in the past.
No pollutants have been detected in wells downgradient of the
active non-hazardous waste landfills. No data were found on
VI-5
-------
/ ( ISK;
y \ CANONIC
-------
downgradient water quality at inactive sites, other than at
Solley Road.^
There are no known sites in the study area where there, is
human exposure to pollutants released in landfill leachate. How-
ever, leachate release is occurring at all sites that are unlined
and is very likely to occur in the future at sites that are lined.
(In humid climates, virtually all man-made liner systems eventual-
ly will fail.) Even so, there is little potential for exposure
at most of the sites for one or more of the following reasons:
o there are no downgradient drinking water wells;
o there is an aquitard or hydrogeologic divide that protects
downgradient wells from contamination^
o all active sites conduct ground-water monitoring programs
to detect contamination should it occur.
Septic tanks. About 12% of the households in Baltimore
County and 32% of them in Anne Arundel County dispose of domestic
wastes in septic tanks; many of these households also use private
wells to supply drinking water. Septic tanks are not widely used
in Baltimore City.
In Baltimore County both the Coastal Plain and Piedmont
Regions have experienced localized ground-water contamination
from septic tanks. The greatest risk appears to be from the im-
proper disposal of chlorinated solvents, particularly in the
Piedmont because of the fractured flow. In a few isolated cases
ground-water concentrations of trichloroethylene from septic
tanks have reached as high as 47 ug/1, but these were the result
of using underground disposal facilities such as septic tanks
and drywells as septic systems as disposal systems for waste sol-
vents, both at military installations and commercial facilities.^
In Anne Arundel County, the far eastern regions along
Chesapeake Bay are at greatest risk from septic tank use. Al-
though septic tanks exist throughout the county, most of the
areas that are particularly vulnerable to ground-water contamina-
tion because of poor hydrogeology and/or high population density
are now sewered. However, in some highly vulnerable areas locat-
ed south of Annapolis along the western shore of the Bay, septic
tanks are still in use in high densities, and ground-water conta-
mination has resulted. In the Mayo Peninsula area, nitrate con-
centrations more than two times the drinking water standard of
10 mg/1 have been found.7
A 1978 survey of the Mayo Peninsula area in Anne Arundel
County found that about ten percent of the drinking water wells
in the peninsula were contaminated with fecal coliform bacteria
(i.e. bacteria were present).** The Phase I study, however, did
not address microbial pathogens because they are an acute rather
than a chronic risk.
VI-6
-------
Surface impoundments. The only available survey of surface
impoundments was published in 1980.^ At that time, there were 41
facilities in the study area using a total of 89 surface impound-
ments to handle wastes. Of the 41 facilities:
o 2 were agricultural; *
o 4 were municipal; . <
o 35 were industrial. Of these/ 11 were hazardous waste
facilities/
o 15% of the impoundments were lined.
Of the 41 facilities, 37 were located over the sedimentary rock
formations of the Coastal Plain region; only 4 .were found overly-
ing the fractured flow systems of the Piedmont. There was little
data on contamination from these facilities. Only one contamina-
tion incident had been reported but only two of the facilities
had any ground-water monitoring systems in place.^
In theory one might expect a moderate likelihood of contami-
nation from such facilities. Only about 15% of the impoundments
were lined, and unlined surface impoundments will recharge ground-
water at a very high rate relative to other sources. However,
the Subcommittee expected that the concentration of contaminants
in discharges from surface impoundments is probably low relative
to other sources. .
''. ~ ~-
None of the 41 surface impoundment sites were within one
mile (1600 meters) of a downgradient drinking water well. Prima-
rily because of the low likelihood of surface impoundments affec-
ting drinking water supplies, and the probable rapid discharge to
surface water of any contaminants leaking- from them, it appears
surface impoundments would not be a major threat to ground-water
resources in the study area is minimal. Ifl2 Surface impound-
ments could, however, pollute the Baltimore Harbor and its tribu-
taries.
Chromium in the area around Baltimore Harbor. Some of the
issues under consideration in Phase I related to particular pol-
lutants -vrather than to sources. Chromium was considered a possi-
ble prevalent ground-water pollutant in the study area. Likely
sources of chromium in the area's ground-water include:
o Chrome ore tailings used as fill, material" around the
Harbor between 1930 and 1975. Because the concentration
of chromium in fresh tailings is quite high and extensive
leaching is likely, elevated concentrations in the Harbor
surface water and nearby ground-waters would have been
found during this period of time. The ultimate environ-
mental fate of chromium in the Harbor is not known. Moni-
toring in the Harbor, however, showed chromium concentra-
tions in sediments to be high, averaging 435 mg/kg.^
VI-7
-------
o Landfills at which the tailings have been disposed of since
1975. A portion of the Hawkins Point Landfill appears to
have discharged chromium to the subsurface, but it is now
retrofitted with a leachate collection system. A nearby
hazardous waste landfill has been the most recent reposi-
tory for chrome ore tailings, and it appears not to release
any leachate into the subsurface.
o Natural chromium deposits are found in fairly high concent-
rations in the Baltimore study area.
o Industrial facilities which have produced or used chromium
in large amounts may have chromium waste piles on-site.
We found no data on chromium concentrations in ambient ground-
water in the study area. -We suspected that chromium concentra-
tions might be high in ground-water adjoining the Harbor area.
But ground-water in this area is already seriously degraded, not
only due to industrial and wetland fill activities, but also due
to saltwater intrusion from the Harbor.^^,11 jn sum, the Subcom-
mittee found no persuasive case for chromium constituting a major
ground-water problem.
Examination of permits and the literature provided only some
of the information needed to evaluate the likely impact that each
of the source types was causing to ground-water resources. Addi-
tional information needed to score the source types on the
resource impact index came from the judgment and experience of
the Subcommittee members.
PERFORMING THE SCORING OF GROUND-WATER RESOURCE IMPACT
The ground-water Subcommittee used the rating system
described in the Overview Section to score the issues in terms of
impact on ground-water resources. (See Addendum-to this chapter
for a 'description of the rating system.) Each Subcommittee member
performed the ratings separately, and then each individual's
ratings on the different sources were combined to form a group
rating for each source type. These rankings were then discussed
by the group. The following section describes how the scoring
was performed, in particular in relation to the two issues chosen
for further study: underground storage tanks (UST) and metals.
The sources represented by the metals issue, although not 'speci-
fied in writing, included chromium in the harbor area, landfills
and surface impoundments.
The first four elements of the pollution impact score all
related to source characteristics. These are: number of sources,
release volume, concentration of contaminants, and persistence of
VI-8
-------
contaminants. USTs and multi-media metals tended to rank high
for these elements.
The second two elements of - the pollution impact score
related to the present and future rate of contamination incidents.
Generally underground storage tanks were rated high. Multi-media
metals were rated moderate to high by some members of the Subcom-
mittee..
- The final three elements of the pollution impact score are
related to potential extent of impact. The Subcommittee compared
potential extent of impact for source types, based on their proxi-
mity to more vulnerable hydrogeologic settings and proximity to
major population or industries using ground-water.
For instance, farming ranked medium to high for hydrogeolo-
gic vulnerability, because much farming occurs in areas which
could be hydrogeologically vulnerable. Sanitary sewer overflows,
however, ranked low to medium for hydrogeological vulnerability
because they tend to occur in the city. Little or no ground-
water in the city is presently used for drinking, and it will
probably not be used for drinking in the foreseeable future.
USTs and metals tended to rank fairly high for the criterion
of proximity to major populations or industries which might be
affected by any ground-water pollution. This is because they are
likely to be located in areas where ground-water is used by major
centers of population and industry. This is in contrast to muni-
cipal landfills, which in general were rated lower, because they
tend to be located in areas which are not densely populated.
The next score was the economic impact score. Subcommittee
members used their best professional judgment in estimating (1)
the cost to prevent, or reduce contamination, and (2) the cost
of response to contamination, either by providing an alternate
supply or treating the existing supply.
In examining the cost of programs to prevent contamination
from the source type, prevention of contamination from feed lots
was generally ranked low. Controls are fairly inexpensive, and
more importantly, there are not a lot of feedlots in the study
area. Prevention of contamination from underground storage tanks
was generally ranked high because the Subcommittee believed con-
trols would be very expensive, particularly since there are many
underground storage tanks.
The cost of alternate supply was generally ranked !Low
because it was believed that contamination incidents for many
sources were more likely to affect just a few private wells
rather than a number of private wells or municipal well fields.
VI-9
-------
Among the exceptions were .underground storage tanks./and septic?
tanks, which, because of the magnitude of possible leaks and/or
the widespread distribution of those sources, could cause wide-
spread impacts to water supplies. Therefore the potential cost
was thought to be fairly high.
The cost of treatment was generally related to the frequency
and magnitude of expected contamination incidents. Farming,
because of the extent of possible contamination, and landfills,
because of the possible extent, magnitude, and toxic nature of
potential contamination, ^tended to be ranked higher than some of
the other sources. USTs and metals generally ranked high.
RESULTS OF THE RANKINGS
'"''
Each member's score for each source was combined in various
ways to produce an aggregate ranking of issues:
1. The first scheme, weighting pollutant impact and economic
impact equally, was chosen because it was the simplest.
Each person's pollution impact (PI) and economic impact
(El) scores were added for each issue, and the scores
were then added for the four Subcommittee members.
2. The second approach weighted pollutiovn impact twice as
strongly as economic impact. Some of the Subcommittee
members argued that the ranking should reflect more con-
cern for the ground-water resource itself, and less con-
cern for the economic consequences of contaminating it.
The (2PI + El) scores that each individual gave to an
issue were added for each individual to arrive at a
total score for each issue.
3. The first two approaches could be strongly influenced by
one or two members of the subcommittee, if they have a
tendency to give most of the sources higher ratings. The
choices of these people therefore would have a greater
chance of being rated highest. Therefore, a third
approach was added, which involved counting the number
of times each issue was ranked in someone's ,3. top five
, issues, whether for pollution or economic impact reasons.
This, we felt, helped mitigate the bias that could deve-
lop if some members of the Subcommittee gave sources con-
sistently high or low ratings.
V -. ' *
Other weighting schemes are possible; for instance, a system
where the economic impact is rated twice as heavily as the pollu-
tion impact. We felt,.however that these three schemes provided
sufficient variety for sensitivity analysis, especially because
the scheme (and its application) relied heavily on best profes-
sional judgment.
VI-10
-------
Out of the 32 issues, each subcommittee member scored only
those which he or she judged to have ground-water resource im-
pacts. If more than 1 person rated a certain issue, then the
issue was included in the ranking. The ranking of each issue was
determined by averaging the scores of the individual members.
The results of this ranking are presented in Table VI-2.
Table VI-2
Baltimore IEMP
Adverse Impacts on Ground-Water Resources
Priority-Setting Results
P and E # of times
Issue equal weights (2P + E)' in top fi\
Underground storage tanks 28.4 41.6 4
Multi-media metals 27.1 40.0 3
Benzene 26.0 39.7 3
Pesticides/herbicides 25.4 37.8 2
Pollution from farming 25.0 35.5 1
Landfills 24.5 36.0 1
Septic tanks 22.5 33.3 1
Chromium in Harbor 22.3 32.4 1
Surface Impoundments 19.9 30.3 2
Acid rain 19.7 27.4 1
Sanitary sewers 19.4 28.3 1
Road salting 18.1 28.7 0
Feedlots 10.0 14.8 0
The Subcommittee discussed extensively how to translate these
scores into a recommendation to the TAG on the three highest pri-
ority issues. There was an unwillingness to follow the index
ranking blindly, for several reasons. First, the aggregate
scores for some issues masked substantial differences of opinion
across the scores by different individuals. Pesticides/herbicides
and septic tanks fell into this category; the members who knew of
or suspected pollution problems from these sources through their
work experience scored these highest.
A second reason that discussion was required was that two of
the issues dealt with pollutants rather than sources, and thus
did not fit neatly into the scoring scheme. These were scored as
if they were sources. For instance, multi-media metals encompas-
sed the primary source types releasing metals, (landfills and sur-
face impoundments) and benzene encompassed primarily underground
storage tanks.".
Finally, there was disagreement as to which sources polluted
ground-water at all. Some members, therefore, scored more issues
VI-11
-------
than others. Those listed in Table VI-2 are the issues which
were rated by 50% or more of the participants.
The Subcommittee eventually decided to recommend only two
issues for study in Phase II underground storage tanks and
multi-media metals. The general rationale for this was that the
two issues appeared at the top of each ranking scheme, and there
was agreement among all Subcommittee members that they were among
the most important issues in each geographic region of the study
area. Thus, the scoring system served its purpose as a mechani-
sim for expressing technical views in a systemmatic way. In the
end, however, the scoring system served as an adjunct to.profes-
sional judgement, not as a replacement for it.
USTs were thought to be particularly important. Most of the
subcommittee members gave them high ratings for these reasons:
o the large number of underground storage tanks,
o the existence of known contamination incidents,
o the potential for future incidents,
o the size of the population using ground-water near the
units, and
o expected hiqh values for the three types of cost that
could be incurred: prevention of contamination, treat-
ment of contamination, and provision of an alternate
water supply.
In addition, the existence of Maryland's regulations on
underground storage tanks increased the importance of studying
USTs. Maryland's requirements to test tanks by January of 1987
meant that the rate of finding new leaking tanks would probably
increase, making further action and knowledge on this issue even
more important than before.
Metals were of particular concern because of their persis-
tence once they have been released into the environment and the
large volume of metals in landfills, fill in the harbor, and sur-
face impoundments.
Benzene problems are closely related to USTs, and chromium
in the Harbor could be included in multi-media metals. Combining
benzene and USTs, and chromium and metals further emphasized the
significance of the USTs and metals issues in comparison with the
other ground-water issues.
r.
Although agriculture-related issues (pesticides/herbicides
and pollution from farming) received moderately high scores, the
Subcommittee did not feel confident in recommending them as
issues for Phase II study. Pollution problems associated with
agriculture are not common to much of the^area. Also,, there was
virtually no documented occurrence of pesticides/herbicides in
ground-water in the region. This could be due to the fact that
VI-12,
-------
little monitoring for pesticides has been performed. The Subcom-^
mittee felt that USTS and metals were clearly more important,
given the 'demonstrated potential for ground-water contamination
and its possible extent. Also, the Subcommittee was unwilling to
recommend any issue which was not associated with at least one
tangible contamination event.
LIMITATIONS OF THE ANALYSIS
1. The most critical assumption underlying the work" of the
Ground-Water Subcommittee was that degradation of ground-water
resources is an important concern apart from and in addition to
human health and ecological impacts. The Subcommittee assumed
ground-water contamination can have high economic costs when:
o the contamination affects an underground drinking water
supply, necessitating expense, either to treat the water
or to obtain a new supply; or
o the contamination affects ground-water -that is suitable as
a drinking water supply but is not currently used. In
this case, the contamination has diminished the value of
the ground-water as a resource for meeting future needs.
An example of this occurred in the Phoenix area of Baltimore
County. Solvents were discarded into several septic tanks. The
solvents then spread into the ground-water, polluting several
private wells in the area. The contamination affected the current
drinking water wells in the area and the desireability of siting
future wells there. The remedy that was recommended is to expand
the city drinking water system into this previously rural area,
though the expansion will come at great expense.
This concern for protecting the ground-water resource for
present or future use, rather than merely protecting human health
was a new area for IEMP. We found no standard approaches or mod-
els for dealing with it since EPA has traditionally focused more
on the -health risks from ground-water. The Subcommittee, there-
fore, had to develop its own methodology in a short period of
time. .Their response was to select a simple, straightforward
scoring scheme. , t '
-} '
2. The ground-water resource impact scoring methodology was
designed as a reasonable approach in the face of missing informa-
tion. Data were extremely scarce on ground-water quality and on
the extent of contamination in the study area. Data was slightly
better for potential sources of ground-water contamination
their number, location, and release rates. A moderate but not
intensive effort was made to find and orga'nize whatever relevant
data existed. No new environmental monitoring or sampling was
performed. Where data were missing, judgment, experience and
VI-13
-------
informed guesswork were applied ,to fill the gaps. These judg-
ments might well be inaccurate. Although the scoring method pro-
duced specific numerical results, these results lacked precision.
The Subcommittee did not rely on any single set of results, but
instead ranked the issues in various plausible ways and examined
the results to determine which source types were consistently
near the top of the rankings.
This approach probably would not have uncovered a new or un-
expectedly important ground-water issue. It relied too much on
judgment and experience to yield a result that differed substan-
tially from the Subcommittee members' preconceptions. This is
not to denigrate the scoring system. It served as a systematic
framework for organizing information so that the Subcommittee
could set priorities in an explicit, logical manner. To uncover
any unexpected ground-water problem, though, would have required
much more new data such as new monitoring or modeling data.
3. Each of the 12 factors (9 under pollution impact and 3
under economic impact) were weighted equally. It is possible
that alternative weightings could be more accurate. However
sensitivity analyses on the weightings produced the same results.
4. The scoring system included elements reflecting each of
the multiple factors that combine to yield ground-water contami-
nation, and resource impact: sources, releases, fate and trans-
port, hydrogeology, exposure and potential responses. The scor-
ing system combined these factors in additive fashion; each ele-
ment was scored and added to get a total score for the source
type.
One might argue that the factors should be combined instead
in a multiplicative fashion. If one of the factors that combine
to yield resource degradation is absent or particularly benign
for a source type, then that source type will cause minimal
resource impact. For example, a source type (landfills) can
cause only minimal resource impact if it affects ground-water
only in locations where the ground-water very rapidly dischargees
to the bay and thus there is no existing or potential human use
of that ground-water. A multiplicative combination of factors
(with population use of ground-water getting scored zero or close
to it) would give landfills an appropriately low score. An addi-
tive approach, on the other hand, would not allow the water use
score to dominate.
The result of using a multiplicative rather than an additive
approach in the scoring scheme would have been to distinguish USTs
and metals even further from the other issues. USTs and metals
have few factors on which they should receive particularly low
scores, whereas many of the other issues have more of those
factors.
VI-14
-------
5. There were substantial differences among individuals in
the scores they assigned on some issues. This could imply either
a lack of reliability in the scoring approach, ground rules for
scoring that were open to misinterpretation, or simply legitimate
differences of opinion among experts. The first two possible
explanations represent criticisms of the scoring process,* the
third one does not. In general, the larger differences in scores
tended to arise from differences in the individuals' experiences,
with those more ^highly involved in toxic/industrial contamination
issues (metals, surface impoundments, landfills, USTs, benzene,
chromium) rating them higher and those involved with rural/
agricultural issues (farming, septic tanks, pesticides) rating
those issues higher than the other Subcommittee members.
This is a particular example of the difficulty of using an
ordinal approach, (i.e. ranking), for the analysis of a technical
problem. Formally, individuals' rankings cannot be reliably com-
bined because individuals' preferences are not comparable. In
the subcommittee's judgment, however, the procedure was legiti-
mate as it served only as a focus for organizing discussion.
6. The list of issues for consideration by the Subcommittee
omitted two that may have been sources of ground-water resource
impact as important as those that were considered: pathogens from
septic systems and salt water intrusion into ground-water in the
Harbor area. However, a previous TAG decision excluded these
issues from the scope of the?IEMP Baltimore project.
In the final analysis, the TAG -felt that despite the obvious
limitations of the method, with its simplifications and frequent
reliance on judgment rather that data', it fulfilled the intended
purpose. It pointed out the issues for further study in Phase II
that appeared to have the greatest impact on ground-water
resources, and were most appropriate for further study by IEMP.
The Phase I work of the Ground-Water Subcommittee was thus
complete, after the TAG accepted the recommendation of USTs and
metals for further study. The Subcommittee proceeded to develop
a workplan for USTs. The process of developing a workplan which
included metals in ground-water was done by a different subgroup
of the TAC-> specifically the multi-media metals workgroup. The
portion of their work that applies'to ground-water is described
below. -- '
ADDITIONAL ANALYSIS OF METALS IN GROUND-WATER
Toward the conclusion of Phase I of the Baltimore Project,
the Ground-Water Subcommittee recommended USTs and metals for
further study. At the same time, other"Subcommittees also recom-
mended metals for further studv. Therefore, the proposed metals
studies were combined and the issue was re-named "multi-media
VI-15
-------
metals." This included metals in drinking water, in air, the
harbor, and ground-water. This issue then, became very broad.
The TAG decided it was too broad for Phase II study and they
wanted to define it better and reduce its scope. They therefore
set up a new workgroup, the multi-media metals workgroup, to
accomplish this task over the next three months. This section
briefly describes the multi-metal media groups overall approach,
and then provides their results which relate to ground-water.
More detail on the other aspects of the work performed by the
metals workgroup can be found in Chapter VIII.
The multi-media metals workgroup decided to make a second
review of metals in all media. The approach to studying metals
in surface water, air, and drinking water changed only minimally,
and the results are included in the chapters on human health risk
and ecological analysis. Additional information was found on
lead, which is included in the human health risk chapter and in
the chapter concluding the priority-setting process. More infor-
mation was available on methods to analyze ground-water than had
been available previously, so the multi-media metals workgroup
reviewed this information as described below.
The ground-water aspect of this review consisted of consult-
ing a study that had recently been performed by the EPA's Office
of Policy Analysis.!2 This national study estimated likely
ground-water resource impact from a variety of source types in a
wide variety of environmental settings. The relevant sources it
addressed included hazardous waste landfills, municipal land-
fills, and five types of surface impoundments (municipal waste-
water treatment; iron and steel; metal fabricating; inorganic
chemicals; and organic chemicals).
The basic analytic tool in the model is a ground-water model
known as the Liner-Location Risk and Cost Analysis Model (LLM).
The LLM was developed by EPA's Office of Solid Wa'ste for estimat-
ing the fate, transport and risks from hazardous constituents
from landfills and surface impoundments. The LLM contains nine
generic hydrogeologic flow fields that represent different aqtti-
fer configurations (single and multiple layer, confined and un-
confined aquifers) and1 ground-water velocities (ranging from 1
m/yr to 10,000 m/yr). These nine flow fields were used in con-
cert with four unsaturated zone thicknesses and four net recharge
rates to create 144 possible environmental settings.
Projected releases from each source were modeled in each of
the 72 environmental settings. It was assumed that there was a
drinking water well 600 meters downgradient of each source, and
impacts were calculated on the basis of the expected volume of a
plume of contaminated ground-water passing by this well, over
time. Ground-water was defined to be impactd or contaminated if
it was polluted at concentrations greater than EPA's drinking
VI-16
-------
water standards. (It^would also be defined as polluted if it was
above taste and odor thresholds, but the metals modeled have no
known taste or odor thresholds.) Impacts were calculated over a
400-year horizon.
The model did not address all the metals that are likely to
leach .from these sources. (The only metals included for these
sources are those thought to be most toxic, persistent, and
mobile. These are: arsenic, cadmium, chromium VI, and mercury).
In addition, a number of types of surface impoundments which
exist in Baltimore are excluded in the study. Specifice sources
in Baltimore also differ from the "typical" national sources in
model in size, leachate content and concentration, and other
parameters.
However, the model indicated that each of the source types,
as well as the closed hazardous waste landfill and the chrome
tailings fill, showed a likelihood of causing resource impact.
The pollutants responsible for this resource impact included
arsenic and cadmium for municipal landfills, and chromium for
the other sources. The multi-media metals workgroup then inter-
preted- the modeled results in light of the actual exposure situa-
tion in Baltimore. Most of the surface impoundments, the closed
hazardous waste landfill, the chrome ore tailings fills and some
of the municipal landfills were all located in industrial areas
where the ground-water was already degraded. Therefore any
ground-water resource impact which might occur would make an
already polluted situation somewhat worse, rather than contamin-
ating ground-water that is clean, but unused, and miqht be needed
for future uses.
Potential human health risk was also considered for these
sources. In the Baltimore study area, however, all of the sur-
face impoundments were in industrial sections of Baltimore City,
where ground-water was already substantially contaminated and
unlikely to be used for drinking in the foreseeable future.
In addition, 5 of the 24 landfills are located in areas
where the ground-water is used for drinking. All open landfills
and some closed landfills have ground-water monitoring wells sur-
rounding them. This monitoring has shown no evidence of contami-
nation by metals at levels of concern. This does not mean, of
course, that there could never be a problem in the future. .With
the very slow ground-water movement in the Coastal Plain area,
contamination could merely be delayed.
Using this information, the multi-media metals workgroup
found no actual human health risk. They believed however, that
municipal landfills could potentially cause ground-water resource
impact. If not carefully monitored, they could cause human
health risks. Because monitoring is already occurring, and
VI-17
-------
because of the high cost of any additional monitoring, the Work-
group decided not to recommend that IEMP fund a monitoring pro-
gram for municipal landfills.
The Workgroup did, however, recommend that the counties
ensure careful monitoring of both open and closed municipal land-
fills in the areas where ground-water is presently used for drink-
ing, or might be used in the future. This is particularly impor-
tant for metals, because they can cause substantial risks at
levels not detectable by humans through taste or odor. This is
in contrast to many organics, which can be tasted or smelled at
very low levels. People presumably stop drinking the water when
it tastes or smells bad. Such avoidance substantially reduces
an individual's exposure to, and therefore risk from, contami-
nated ground-water.
VI-18
-------
Addendum" to Chapter VI
Ground-Water Resource Impact Scoring Methodology
In developing the Phase I screen, the Subcommittee decided
to provide two scores for each issue Pollution Impact and Eco-
nomic Impact. Both affect decisions on ground-water assessment
and management.
Pollution Impact (P) and Economic Impact (E) were treated as
separate criteria. The Subcommitte recommended scoring each of
them separately so that the flavor of the one could be compared
to the other, rather than being lost in a single numerical score.
The two scores could later be combined in various alternative
ways to provide a single score for relative ranking of the issues.
Each of these criteria was split into units: Pollution
Impact into three units, and Economic Impact into two units. In
turn, units were further subdivided except for E.I. Each subunit
(such as P.I.a.) was scored, with 5=high, 3=intermediate, and 1=
low. These scores were averaged for each unit, as shown on the
final page. In turn the units were averaged to determine a score
for the criterion, as shown on the last page. Ihe final score
was reported as: Pollution Impact/Economic Impact. (4/5 or 1/3
for example)
Definitions: In this process the Subcommittee used these
guidelines for scoring each unit:
- P.I.a. number of facilities comorising the source type
1= 1 to 10
3= 11 to 100
5= 100 +
b. release volume over time for one facility
1= small; less than the capacity of the facility
5= large, continuous
c. concentration of contaminants
1= dilute
5= very concentrated (e.g; - gasoline)
d. persistence of contaminant
1= degrades within one month
5= degradation exceeds one year
P.2.a. present extent of contamination
1= few examples of contamination
5= many examples of contamination
b. future extent of contamination
1= safe or lower rate of contamination incidents
expected over next 10-20 years.
VI-19
-------
Addendum (page 2 of 4)
5= much higher rate of contamination expected over
next 10-20 years ;
P.3 Potential extent of impact refers to whether or not source
types are more densely distributed in areas where they are
more likely to cause more impact.
P.3.a. Hydrogeologic setting. A high score here would imply a
source type is more densely distributed in a hydrogeologic
setting which is particularly vulnerable. The Subcommit-
tee believed Coastal Plain area (located SE of the fall
line) is more vulnerable because there are few barriers to
transport; contaminants continue to move for years over
significant distances. In the Piedmont area, (NW of the
fall line) however, transport occurs for fairly short
distances, usually no further than the edge of the surface
water drainage area (i.e. local watershed), where the con-
taminated ground-water may be discharged to surface water
streams. Therefore, sources densely distributed in Coas-
tal Plain hydroqeology would be ranked 5, sources densely
distributed in Piedmont hydrogeology would be ranked 1,
and sources equally distributed among both hydrogeologic
types would be ranked 3.
P.3.b Population served by affected wells. Using parallel reason-
ing to the situation above, sources located preferentially
where a large in areas of Anne Arundel County served by
municipal ground-water) would be scored 5; and sources
located preferentially in areas where there is no ground-
water use (perhaps because ground-water in the area is
already badly contaminated) would be ranked 1.
P.3.c. Industries served by affected wells. Sources located
where a large amount of industry using well water may be
affected would be scored 5; and sources located preferent-
ially where water is presently too contaminated for indus-
trial use would be ranked 1. (This category is separate
from 3b because industrial users can often tolerate higher
concentrations of contaminants in water they use.)
E.I. Prevention - To be more accurate, this is really "reduced
likelihood of contamination" because accidents will con-
tinue to occur. This measures the likely costs of all
reasonable prevention programs such as tank permitting
programs, the goal of which is to prevent future pollution.
E.2.a. Response - One form of response is alternative supply, or
in other words, the cost of providing an alternative water
VI-20
-------
Addendum .4page 3 of 4)
supply. Examples of alternative supply are bottled water,
and extending city water supplies (from reservoirs). This
will vary primarily in proportion to population affected.
1= a few homes
5= alternative supply for a large municipal well field
*
b. Response - A second form of response is treatment, or in
other words the cost of providing treatment. This will
vary primarily in proportion to the extent of impact. It
will also vary with the type of pollutant.
1= economically treatable at individual household level
5= large scale purging of aquifer over months or years
VI-21
-------
Addendum (page 4 of 4)
MATHEMATICAL APPROACH FOR GROUND-WATER RESOURCE SCORING
Pollution Impact (1+2+3) / 3 =
1. Source characteristics (a+b+c+d) / 4 =
a. number ( )
b. release volume over time ( )
c. concentration of contaminants at ( )
point of release
d. persistence of contaminants ( )
2. Likelihood of contamination (a+b) / 2 =
a. present ( )
b. future
3. Potential extent of impact (a+b+c) / 3 =
a. hydrogeologic setting ( )
b. population served by affected wells ( )
c. industry served by affected wells ( )
Economic Impact (1+2) / 2 =
1. Prevention (cost to reduce likelihood ( )
of contamination)
2. Response to contamination (a+b) / 2 =
a. alternate supply ( )
b. treatment ( )
VI-22
11111 HI
-------
Footnotes
1. EPA, Regulatory Integration Division, "Baltimore IEMP Issue
Summary Paper: Septic Tanks", February 1, 1985.
2. Personal communication between Dr. Emery T. Cleaves, Maryland
Geologic Survey and Hope Pillsbury, EPA, November 1986.
3. Personal communication between Jim Dieter, Baltimore County
and Ed Weber, Maryland Department of Natural Resources, with
Hope Pillsbury, EPA, November 1986.
4. Personal communication between Reid Rosnick, Maryland Depart-
ment of Health and Paul Lewandowski, Sobotka and Co., Inc.,
quoted in EPA, Regulatory Integration Division, Memorandum
from Sobotka & Co., Inc.,: "Baltimore IEMP Issue Summary
Paper: Landfills Working Draft", February 1, 1985.
5. Personal communication between Bob Beyer, Maryland Department
of Health and Paul Lewandowski, Sobotka, Inc., quoted in EPA,
Regulatory Integration Division, Memorandum from Sobotka &
Co., Inc., "Baltimore IEMP Issue Summary Paper: Landfills
Working Draft", February 1, 1985.
6. Personal Communication between Brooks Stafford, Baltimore
County, and Hope Pillsbury, EPA, March 1987.
7. Environmental Source and Engineering Inc.7 "Assessment of
Contamination - Phoenix Military Reservation Final Report"
July 15, 1983.
8. EPA, Addendum to Working Draft for Baltimore Technical commit-
tee. Multi-Media metals. undated.
9. Anne Arundel County Health Department, Mayo Peninsula Water
and Sewage Survey, 1978, quoted in reference #1, above.
10. EPA, Regulatory Integration Division, "Surface Impoundments
in the Baltimore Study Area", April 24, 1985.
11. McGlincy, D.A., J.T. Chaconas, andB.D. Huang. 1980. The Mary-
land Surface Impoundment Assessment Program: A Final Report.
Maryland Department of Natural Resources, quoted in EPA, "Sur-
face Impoundments in the Baltimore Study Area", April 24,
1985.
12. EPA, Regulatory Integration Division, "ground-water Resource
Degradation Ranking for Chromium in the Area Around Baltimore
Harbor", April 25, 1985.
13. EPA, Regulatory Integration Division, "Chromium in and Around
Baltimore Harbor", April 24, 1985.
14. EPA, Regulatory Integration Division, "Draft Comparative
Impact Analysis of Sources of Ground-Water Contamination
Phase II Draft Report", March 27, 1986.
VI-23
-------
VII. ANALYSIS OF ECOLOGICAL IMPACT
-------
VII. ANALYSIS OF ECOLOGICAL IMPACT
Several of the initial thirty-two issues the TAG identified
as important to analyze in this study dealt with the ecological
condition of Baltimore Harbor and its tributaries. To establish
study priorities among pollutants and sources of pollution that
enter these bodies of water, the TAG recognized that it needed a
priority-setting method that covered ecological effects.
No analytic device for this purpose was readily apparent at
the outset of Phase I of the IEMP. Therefore, the TAG decided to
have project staff spend their initial efforts evaluating exist-
ing studies on the Harbor and its tributaries and to collect some
data on levels of various harbor pollutants in a monitoring study.
From these efforts, the TAG recognized it could take existing
data and develop an ecological "index" rating for various harbor
pollutants to assess the relative significance they may have on
the ecological health of these water bodies. The pollutants with
the highest index numbers would represent the chemicals most im-
portant to study in Phase II.
This chapter presents, first, a section which describes the
difficulty of defining problems one might examine in the harbor
a.nd the approaches to studying them. The second section provides
background on biological and pollution-related aspects of the
harbor. The third section describes a harbor monitoring program
which EPA undertook in 1984, to obtain an improved data base.
The fourth section presents how the method for setting priorities
for ecological impact "was chosen and applied. The fifth section
summarizes results and limitations of the Phase I harbor studies.
The Baltimore Harbor was included in the IEMP study for a
number of reasons. First, there was State and local concern for
the ecological health of the harbor, the viability of the harbor
as a resource for fishing and recreation, and its aesthetic value.
Second, there was concern regarding the adverse effects that pol-
lution in the harbor may have on the Chesapeake Bay. An addition-
al reason to study the harbor was to further the goals of the
Chesapeake Bay Initiative, a major cross-government effort to
preserve,and improve the Chesapeake Bay as a natural resource.
Finally, the harbor provided the first opportunity for IEMP to
address ecological impacts in an IEMP study. EPA saw the need to
develop and apply priority-setting methods to ecologic impacts,
and in particular, ecologic impacts on an aquatic environment.
The scope of the study did not include the Chesapeake Bay.
The Chesapeake Bay is too large geographically, and in addition
was too complex a topic for the resources of the IEMP. Therefore
the study area was restricted to Baltimore Harbor. Data limita-
tions led to the decision to restrict the examination of ecolo-
gical effects to those pertaining to the aquatic environment.
VII-1
-------
DEFINING AN APPROACH TO THE HARBOR ISSUES
There are as many goals for the harbor as there are potential
uses. Ecological health of the harbor can be defined in terms of
any or all of these goals. The principal question posed by the
TAG was whether it could identify a cost-effective way of attain-
ing any one of these goals. The harbor subcommittee believed it
lacked the information necessary to address this major question.
Therefore, it approached this question in terms of several subsi-
diary, but fundamental questions, which follow:
0 Is the harbor a source or sink for pollutants? In
other words, does the harbor serve as a source of
pollutants to the Chesapeake Bay or act as a sink for
pollutants discharged into the harbor or entering
from the Chesapeake Bay?
0 What is the relative contribution which pollution in
the sediment and pollution from direct dischargers
(point and non-point) make to the ecological health
of the harbor?
0 What is the relative importance of specific pollutants,
and pollutant classes, to ecological health of the
harbor? Are the so-called "conventional" pollutants
(i.e. dissolved oxygen and nutrients) or the "toxic"
pollutants (metals and organics, such as pesticides)
more limiting to harbor life? Within each of these
groups, which specific pollutants cause the most harm?
0 What is the relative importance of habitat (including
food chain effects, physical factors such as tempera-
ture, salinity, suspended sediments due to dredging,
and existence of appropriate spawning areas) versus
pollution, in determining the ecological health of the
harbor?
The harbor subcommittee was aware that in order to determine
which, if any of these issues they could address, they would need
to determine what information was currently available for the
five basic elements of iany study of pollution: pollutants,
sources, environmental fate and transport, ambient data, and
impacts or damages to the biota of concern., The following pre-
sents background information on these elements for the'harbor.
Over the course of Phase I the harbor subcommittee was also
called the harbor workgroup. Thus, the term subcommittee and
workgroup are used inter-changeably throughout this report.
VI1-2
-------
HARBOR OVERVIEW
-- " . **
Pollution Impacts on Biota
A study performed by the Center of Environmental and Estuarine
Studies (GEES) in 19751 compared the ecological health of Baltimore
Harbor to that of the Chester River, a relatively unpolluted tri-
butary with similar latitude and salinity levels as.the Baltimore
Harbor. The intent of the comparison was to determine whether
Baltimore Harbor's ecology has been adversely affected by pollu-
tion.
For fish, the GEES study found that the Baltimore Harbor
water column supports a wide variety of ,life and serves as a nur-
sery and feeding ground for many species. In fact, the inner
harbor supported various fish populations.
The researchers observed that striped bass and hogchoker
were not spawning in Baltimore Harbor. In addition, the juve-
niles of several bottom feeders, the hogchoker,'winter flounder,
and the Atlantic croaker, were absent from the Harbor. Possible
reasons for this include pollution and habitat; however the study
was not able to accurately identify the reasons for failure of
certain species' life-stages to be present of to spawn in the
harbor and its tributaries.
Ecological principles dictate that the ecological health of
benthic organisms can be affected by stresses such as pollu-
tion2^3. Unlike fish that swim from environment to environment,
benthic organisms are primarily sedentary bottom-dwellers.
Because they feed and breathe close to the highly polluted bottom
sediments, they are very susceptible.to exposure to pollution in
the sediments.
Two possible indicators of environmental health of benthic
organisms are benthic biomass and the number of benthic species.1
GEES researchers compared these parameters in Baltimore Harbor
and the Chester River. The Baltimore Harbor had a lower dry
weight of benthos and about one third fewer benthic species than
did the Chester River. These findings suggested to them that the
sediments of Baltimore Harbor provide a less healthy environment
than the sediments of the Chester River.1 ,
The researchers classified the Baltimore Harbor into three
zones based on sediment pollution: 1)"a semi-healthy zone from
the* harbor mouth to Fort Carol!; 2) a semi-polluted zone from
Fort,,Caroll to Fort McHenry; 3) and a polluted zone ;within the
inner ftarbor and the tributaries. (See Figure VII-1 for map).
The GEES study provides information on the biological state
of the Baltimore Harbor during the 1970's. Environmental condi-
tions may have changed since then due to pollution abatement and
VII-3
-------
natural environmental processes. Although the ambient water
could support diverse fish life during the study period, the GEES
study suggests that the Baltimore Harbor and its tributaries
could not support the entire life cycles of- certain fish, nor
does it support the diversity or biomass of benthic species that
would be possible if it was less polluted.
Sources of Pollution in Baltimore Harbor
By convention EPA divides sources of water pollution into
two categories: point sources and non-point sources. The point
sources are industrial facilities and sewage treatment plants dis-
charging effluents. The non-point sources consist of activities
leading to urban and residential runoff that enters from streams
and storm drains which empty into the harbor. Two additional
sources that may pollute the Harbor are accumulated pollution in
the sediments, which is described in the section on ambient data,
and air deposition of pollutants directly into the harbor, which
we deferred to Phase II because of resource considerations.
Point Sources
Data was available for the first three categories of sources
of pollution to the harbor. We consulted NPDES (National Pollu-
tion Discharge Elimination System) permits for information on the
"major" facilities polluting the harbor. The 27 "major" facili-
ties include the Patapseo and Back River sewage treatment plants.
'-." ' * -
A total of 27 "major" dischargers produce a total of over
2 billion gallons/day of .effluent. The 135 "minor" sources of
direct discharges together are estimated to provide 21 million
gallons/day of effluent to the harbor.41 in contrast, non-point
sources are estimated to provide an average of 107 million
gallons per day of effluent to the harbor5»6.
EPA's definitions of major and minor dischargers is based on
knowledge of the usual toxicity and concentrations of pollutants
discharged from facilities, based on their SIC (standard indus-
trial classification) codes. Therefore unless a facility's dis-
charge is grossly more t6xic or greater in volume than expected
given its SIC code, it is reasonable to assume that the "majors" (
provide substantially greater loads of toxics to the harbor than'
do the " mi nors " . ^
Non-point Sources
*> We examined three types of non-point source runoff catego-
ries: urban stormwater runoff, agricultural stormwater runoff,
and forest stormwater runoff. To get expected contribution per
acre of pollutants from urban stormwater runoff, we used monitor-
ing data from the Jones Fall Urban Runoff Program.8 This was an
intensive study of one rivershed, the Jones Falls, in Baltimore
' VII -4
-------
City and County. We extrapolated data from this urban watershed
to the other urban watersheds in the Baltimore region. To deter-
mine expected contribution per acre of metals and pesticides from
agricultural and forested lands in the study area, we used
studies in Lancaster, Pennsylvania and the Rhode River area in
Maryland. We determined the acreage in each watershed which is
urban, agricultural, and forested, using 1975 data on land use
from Baltimore Regional Planning Council and the NASA Landsat
Program. Then, using the data from the monitoring studies men-
tioned above, we estimated average annual loadings by pollutant
from non-point sources for the entire study area.5»6 These are
very rough estimates because the specific types of agriculture
and forests studied are not identical to those found in the
study area.
The non-point source discharge estimates do not include ade-
quate data on two potentially important categories: runoff from
industrial lots (including stockpiles, work areas, etc.) and
sanitary sewer overflow (which occurs during storms). Of these
two, defining and quantifying pollution loadings from industrial
areas appears to be the most important to pursue since these dis-
charges can cause high chronic toxicity.
A third possibly major source of non-point source pollutants
is run-off carried down the Susquehanna River into the Chesapeake
Bay and the Harbor mouth. We were not able to estimate this
pollutant load to the harbor.
Table VII-1 compares the mass of different metals that are
discharged by point sources versus non-point sources. For these
metals, point source discharges are at least 10 times and often
more than 100 times greater than the non-point discharges.
Figure VII-1 provides the total mass loadings from major direct
discharges, by pollutant. We also distinguish between the mass
contributed by Bethlehem Steel, versus all other major sources
combined, for each pollutant. For all metals shown, Bethlehem
Steel is the single largest individual source. This is due in
part to this facility being much bigger than any other facility.
Information on Fate and Transport
Information on fate and transport of pollutants in the har-
bor is sparse. Maryland's Office of Environmental Programs has
developed a hydrodynamic model of the harbor. The model, however,
does not describe all the complexities of flow in the harbor.
The harbor has been observed showing an unusual three-layer flow
near the mouth, at certain times. Water in the upper and lower
portions (vertically) of the harbor moves into the harbor, and
water in the middle section moves out of the harbor. In addition,
the model does not address water quality, or provide information
on the fate of pollutants in the harbor. It would be very expen-
sive to augment the model to provide such information.9
VII-5
-------
Table VI I-1
BALTIMORE IBMP
DISCHARGES TO BALTIMORE' HARBOR OF METALS
(1000 kg/year)
MAJOR POINT SOURCES1
ESTIMATED NON-POINT SOURCES2
Zinc
Cadmium
Mercury
Lead
Chromium
Copper
Nickel
Arsenic
Beryllium
Antimony
411.9
42.4
.915
457.2
182.6
158.43
707.17
159.11
3.2
17.0
18.3
.4
N/A
14.0
2.7
5.1
3.9
1.1
0.1
0.4
TOTAL
2139.9
46.0
1. Source of data is EPA discharge monitoring reports, collected
for the National Pollution Discharge Elimination System. Data
is for 1983-1985, except 1982 data for Bethlehem Steel.
Only "major" dischargers were included. Metals were included
if they were discharged by at least two "major" facilities.
2. Source of data is the Baltimore Regional Planning Council's
"Jones Falls Watershed Urban Storm Water Runoff Project, "
Interim Report. Baltimore, MD., 1982. For method used to
estimate direct discharges of metals from non-point sources,
see p. VI1-4;
N/A indicates no data available.
HI IB HIPI' in 'lli'iiffllllp
ri|iin in fi
-------
Figure VII-1
Baltimore IEMP
Mass Loadings to Baltimore Harbor
(372)
All Other Major
Point Sources
°°°°°°°°°°o°°°oo°°°°°°°o°°°°°°°°°°°°°°
(24.9)
(134.3)
100 200 300 400 500 600
Cumulative Mass Loadings to Baltimore
Harbor fcko/yr)
700
Source of data is EPA discharge monitoring reports, collected for the
National Pollution Discharge Elimination System for "Major" point sources.
Data is the yearly average for 1983-1985, except for Bethlehem Steel,
which is 1982 data.
-------
Assessment of Ambient Data
A review of available information on ambient water concentra-
tions and sediment concentrations of toxics in Baltimore Harbor
revealed very little data on ambient water "concentrations. The
EPA STORET data base contained toxics data for several of the
tributaries to Baltimore Harbor, but .nothing for the estuary it-
self. Only one study, a 1977 report produced for EPA by Trident
Engineering, had data on toxics in ambient water in the harbor.
Several studies, however, showed Baltimore Harbor sediments to be
enriched with both toxic organic compounds and metals.
A geographic range of pollution was found for metals in the
sediments. A study performed by Villa and Johnson in 1984 moni-
tored sediments at over 150 locations, evenly spaced throughout
Baltimore Harbor. They found that levels of metals are highest
near the mouths of the tributaries to the harbor, and lower in
the middle areas of the harbor. The location and spatial arrange-
ment of this area's range in pollution was similar to that found
for biota in the sediments. (See p. VI1-3.)
Information on Specific Pollutants a;nd Biota
Much of the information on specific pollutants was reviewed
in the context of biota, sources or ambient data. However, one
study pertaining to aluminum and its possible effect on spawning
of fish, deserves special mention, and is included-.here.
This issue is one that has been researched extensively by
the Smithsonian Environmental Research Center in Anne Arundel
County. Analyses indicate that acid rain as well as the use of
nitrogen fertilizers can acidify rainwater and ground water,
causing metals, particularly aluminum, to leach from soils into
surface and ground water. This acidic water may then impact fish
and amphibian ^populations. The low pH and, aluminum are thought
to be toxic, perhaps synergistically, to, fish and amphibians
including larval stages.
The results of the Center's studies suggest a correlation
between increasing levels of acidity in area streams and decreas-
ing larval survival in fish species such as striped' bass and
yellow perch. In some streams :in the Rhode River Watershed the
total extinction of these species, and others, has .been recorded.
This situation is especially acute in areas with pboirly buffered
soils (i.e. low alkalinity), such as those in the Atlantic Coast-
al Plain r-egion, which encompasses the entire eastern portion of
the study area. 0 '
DEFINING AN APPROACH TO STUDYING THE HARBOR
lEMP's early consideration of how to set priorities for the
VI1-6
-------
harbor led to the concern that we needed more data. lEMP's
initial assessment showed a need for data on pollutant concentra-
tions in biota in the harbor, environmental fate and transport
of pollutants in the harbor, and ambient water quality in the
harbor. .
Because studies were ongoing by the State of Maryland on
pollutant concentrations in Harbor biota, we decided to address
the latter two questions. The IEMP chose a monitoring rather
than a modeling approach because a modeling approach would have
been prohibitively complicated and expensive.
;
HARBOR MONITORING
Objectives of the Monitoring Plan
The IEMP designed a monitoring plan to characterize condi-
tions near the mouth of the Harbor and to provide a better under-
standing of the inter-relationship between the Harbor and Bay. A
second goal was to improve the characterization of ambient water
quality in the harbor. As it turned out, this latter goal was
achieved but little light was shed on the first goal. The speci-
fic objectives of this monitoring and analytical plan were as
follows:
o to obtain a data base to identify water quality para-
meters which may be limiting to aquatic organisms in the
Baltimore Harbor.
o to determine if there is net pollutant inflow to the har-
bor from the Chesapeake Bay, or net pollutant outflow
from the harbor to the Bay. This would be determined by
examining any differences in pollutant concentrations
that are present between the inflowing and outflowing
Baltimore Harbor waters during two 12-hour periods.
o to learn more about the role the harbor hydrodynamics
play in harbor water quality variations, by gathering
more information on when and where the tide flows in two-
to-three layers and' in one layer.
Due to funding constraints, the monitoring was extremely
limited and should be viewed as providing some "snapshots" in
time of levels of certain pollutants in ambient water at certain
sites in the harbor, in a variety of tidal and meteorologic con-
ditions. The monitoring is not a detailed, statistically valid
assessment of pollutant concentrations. In addition, because of
the limited number of samples collected, only the first objective
above was partially met; the results of the monitoring cannot
address the other two objectives.
VI I-7
-------
Pollutants Selected for Analysis
Prior to making a final selection of parameters, IEMP
comprehensively reviewed and analyzed existing data. In addition,
we performed a preliminary monitoring program on July 9, 1984 to
test detection limits for pollutants. Table VII-2 presents the
water quality parameters EPA chose for the Baltimore Harbor water
sampling study based upon a review of previous monitoring studies
and EPA Discharge Monitoring Reports^*11. The pollutants found
were primarily metals, phenols, cyanide, and purgeable organics.
IEMP analyzed for both the total and dissolved metals. Most
ambient water quality criteria for protection of aquatic life are
based on total metals. Total metals are also required to relate
source loadings to ambient water quality data (point source dis-
chargers are only required to monitor total metals). The portion
of total metals which are dissolved is important in the aquatic
ecosystem because they are the most readily bioaccumulated and
their concentrations in the water column must be known before
assessing potential impacts to the biota.
IEMP monitored pesticides because they may be carried into
Baltimore Harbor by the Susquehanna River water and by non-point
source runoff. IEMP monitored metals and some polynuclear aroma-
tic hydrocarbons because they could be associated with eutrophi-
cation, low dissolved oxygen, and fish kills. In addition they
may be harmful to phytoplankton, the primary source of the aqua-
tic food chain.
Finally, physical parameters of dissolved oxygen, water tem-
perature, salinity, conductivity, and current direction were
monitored to provide some indication of vertical differences in
the, water column and to help identify the different flow layers.
Sampling Plan
A preliminary monitoring effort was conducted on July 9,
1984 to determine detection limits (detection sensitivity) of the
planned monitoring and analysis techniques and to determine which
parameters to analyze. Based on these results a more comprehen-
sive monitoring program was designed and discussed with members
of the scientific and regulatory community.^ After incorpora-
ting the contributions of these experts, a final monitoring plan
was developed.*3.
The monitoring program was conducted on September 12, 1984
and September 26, 1984. Although the majority of planned samples
were collected, some were not due to rough seas and high winds.
Both of the planned sampling dates were preceded by unusually
long periods of dry weather. The period following the second sam-
pling day, however, saw several days of rain. Because rainfall
VI1-8
-------
Table VII-;-
Baltimore IEMP
IEMP MONITORING PROGRAM
PARAMETERS
General Water Chemistry
Total Dissolved Solids Total Suspended Solids
Nitrate Orthophosphate
Cyanide, Total
Metals (Dissolved/Total)
Beryllium Cadmium
Chromium, Total Calcium
Copper Lead
Magnesium Manganese
Mercury Nickel
Sodium Zinc
Organics
Total Organic Carbon Polynuclear Aromatic Hydrocarbo
Phenols , Total Recoverable Purgeable Organics (Volatile
Pesticides (13 specific pesticides) Organic Compounds)
Other Parameters
pH Salinity
Conductance Flow direction
Temperature Dissolved Oxygen
-------
increases the volume of non-point source runoff, the IEMP decided
to compare ambient water quality both before and after rainfall.
A third day of sampling was therefore conducted to observe any
measureable effects of nonpoint sources. This "wet weather"
sampling was carried out on October 5, 1984.
Sampling Station Locations
The Baltimore Harbor monitoring plan included 13 monitoring
sites; 11 in the harbor, and two in the bay, outside the harbor.
Figure VI1-2 shows the location of the various monitoring sta-
tions in the Harbor. The numeric stations were sampled at var-
ious depths and were intended to be used to determine net inflow
or outflow of pollutants at the Harbor/Bay interface. The alpha
stations were sampled at mid-depth and were intended to be used
to determine spatial water quality variability within the Harbor.
Quality Assurance/Quality Control
Before the monitoring began, a formal Quality Assurance (QA)
Plan was prepared by EPA Headquarters and approved by EPA's
Region III. The plan outlined the scope of the monitoring pro-
gram, the objectives of the program and the reasons for selecting
particular monitoring sites in the harbor. The QA plan also de-
tailed the exact procedure to be used in the collection, handling
and analyzing of samples. The plan established quality control
procedures to be followed throughout the program including:
o Field blanks for VOA's, PAH's, and metals,
o Sample container cleaning procedures.
o Sample custody procedures,
o Standard equipment combination and maintenance, and
o Data validation measures.
Results of IEMP Monitoring
Table VI1-3 shows the maximum observed metals concentration
during the dry weather sampling (September 12 and 26, 1984) the
wet weather sampling (October 5, ,1984} mean values of the sampl-
ing, published sampling data (Trident 1977) including mean values,
and EPA chronic and acute effects criteria levels for saltwater
organisms. (USEPA 1980). No pesticides were found.1*
The initial analysis of this data, and other available data
suggested the following to us:
o Water column concentrations for metals with corresponding
EPA ambient water quality criteria did not exceed the acute
VI1-9
-------
Figure vii~ 2
Colgate Creek
Lazaretto Pt.
NAUTICAL MIES
\_X Trident Sampfng Sites
IEMD Sampling Sites
-------
Table VII-3
BALTIMORE IEMP
MONITORING DATA AND EPA CRITERIA USED IN ANALYSIS
CF BALTIMORE HARBOR
Parameter
(Units)
IEMP Data, 1984-1-
7/9
9/12
9/26
10/5
("Wet)
Weather")
Mean"*
(Metals
only)
Trident Data-*
Range 1977 Mean4
(Metals
only)
EPA
Saltwater Criteria-3
Chronic
Acute
Beryllium (ug/1)
Cadmium (ug/1)
Chromium (ug/1)
Copper (ug/1)
Lead (ug/1)
Mercury (ug/1)
Zinc (ug/1)
Nickel (ug/1)
PAHs (ug/1)
VQAs (ug/1)
Pesticides (ug/1)
<0.25
<0. 5-1.1
<4.0-4
<10.0
3.5-5.4
<0.2-0.24
<4.0
<4. 0-6.0
*
*
*
<1.0
<0.5-1.5
<4.0-12
<2.0-9
<5_,0-9.6
<0.2-0.5
<3.0-19
<15.0
<2.0-32
<10.0
-------
24 hour exposure concentration though some exceeded the
corresponding EPA chronic criteria concentration. The
metals for which exceedances of EPA's chronic salt water
criteria occurred in at least one data set, are mercury/
lead, copper, nickel, cadmium and zinc. Therefore, ambi-
ent metals concentrations in Baltimore Harbor are likely
to have adverse sub lethal effects on aquatic organisms
but are not likely to directly cause fish kills.
o Interstitial waters tend to exhibit metals concentrations
ranging from two to five times greater than the water
column, and interstitial waters consistently exceed acute
ambient criteria for the six metals listed above. While
there are no criteria for interstitial waters, the rela-
tively high concentrations will tend to adversely impact
benthic organisms and impede the reproduction of bottom
spawning organisms in the harbor.
o Metals concentrations in the sediments tend to vary
substantially, with the highest concentrations usually
found in the tributaries and harbor head.
DETERMINING METHODOLOGIES FOR SETTING PRIORITIES WITHIN THE HARBOR
It was very difficult for the harbor workgroup to determine
goals for the harbor. It was also very difficult for them to
determine the best way to set priorities within the harbor.
Should this be done by pollutant, by source, by exposure, or by
magnitude or type of effect on biota, or some combination of
these? The entire budget for the study could have been spent on
improving methods for setting priorities, or on obtaining more
data, to enable us to use more sophisticated priority-setting
schemes.
We already had made a relative ranking of point and non-
point sources. We did not* however, know which of the pollutants
in ambient water (monitored by IEMP and other institutions) were
most .likely to cause adverse effects on biota. There were sever-
al priority-setting methods available or under development for a
pollutant-based approach. The harbor workgroup, therefore, inves-
tigated four methods, each of which would address individual pol-
lutants or classes of pollutants. The first three were thought
to be relatively inexpensive, as they relied on existing data to
set priorities. 14,15 The fourth, laboratory and field bioassay
and bioaccumulation studies, was investigated to determine if
this generally more expensive approach to setting control priori-
ties might be the only scientifically acceptable one.16,17 in
addition, the workgroup wanted a review of available literature
that would pertain to the study area. Each method of priority
setting is mentioned below. The harbor workgroup eventually
chose indexing as the method it would use in Phase I. They
decided to defer any use of the other methods until Phase II.
VII-10
-------
INDEXING: We developed this method to directly compare
existing pollutant concentrations in ambient water to generally
applicable reference values, specifically EPA's chronic water
quality criteria for saltwater and freshwater,, and EPA draft sedi-
ment threshold contamination concentrations.
ECOLOGICAL SCORING: A common means of scoring likely harm-
ful ecologic effects from a given pollutant on a given species,
is to measure its LC^Q (concentration of the pollutant at which
10% mortality of the species occurs within a specified amount of
time). LCso's are also commonly developed. Ecological scoring
uses the same data used in developing the lethal concentrations
to generate an entire dose-response curve, thus enabling the
analyst to estimate the mortality rate for a species from a cer-
tain pollutant at any concentration.
REFERENCE ENVIRONMENTS: This is another approach under
development at EPA; it compares ambient conditions in different
"reference" areas (of similar physiography) in order to determine
which of a wide variety of pollutants or pollutant classes are
most responsible for environmental deterioration in a given study
area.
BIOACCUMULATION AND BIOASSAY TESTS: This entailed, a review
of the technical approaches to bioassay and bioaccumulation tests
as well as a literature review of all studies known to have been
conducted in recent years in the Baltimore Harbor and Chesapeake
Bay. Bioaccumulation tests measure the degree to which certain
pollutants bioaccumulate in biota such as fish. Bioassay tests
measure the toxicity of the entire mixture of pollutants the fish
or other organism is exposed to. It is, thus, a measure of toxi-
city. Additional work, however must be performed to isolate the
individual pollutants of greatest concern. 1*>, 17
The indexing method appeared to be simple, easy to replicate,
and would provide a good first^cut ranking of pollutants. In
addition, it was based on EPA-approved water quality criteria or
draft criteria. The harbor subcommittee, therefore, chose this
method as the priority-setting tool for Phase I. Results of the
indexing follow.
INDEXING
The indexing concept is well established in environmental
work air indices, for instance, are often used to rate air
quality in general categories such as good, fair, poor, or un-
healthful. In this case the harbor subcommittee investigated the
use of indexing to select pollutants for intensive study in Phase
II.
VII-11
-------
Indexing can be used on measurements of the concentrations
of contamination in the water column, sediments, or effluents.
Its advantage is that it is straight-forward and specific. It
compares the concentration of a certain pollutant to a pollutant
reference value, or standard. We used externally supplied refe-
rence values: EPA1s saltwater chronic criteria, EPA freshwater
chronic criteria, and draft EPA threshold contamination concen-
trations. EPA1s criteria are reference numbers which are devel-
oped through a scientific review process, thus enhancing the
validity of the scores.
The basic methodology of indexing is to derive index values
as a simple ratio of observed values to reference values:
INDEX NUMBER = OBSERVED VALUE
REFERENCE VALUE
Theoretical and Practical Considerations
Indexing is an attractive way of sorting data, screening for
potential impacts, and setting priorities among pollutants. It
enables comparison of individual pollutants, based on potential
ecological impact. A second positive aspect is that it .uses
standard EPA approved reference values. A third advantage is
that it is simple, inexpensive, and is based on readily-available
information.
Indexing has a number of limitations, which are listed below.
o Detection limit problems, changes of conditions over time,
lack of data for particular pollutants, spatial variations
over a study area, definition of values of concern (mini-
mum, mean, maximum), and similar problems can detract from
the effectiveness of the approach.
o Criteria are available for most of the common metals, but
not for the wide range of possible organic contaminants
(pesticides, other industrial organics).
o EPA Water Quality Criteria have been developed to repre-
sent typical national conditions, not specific local con-
ditions. They are based on setting levels at which a
particular sensitive species can survive -- if that
species is not present in a particular locale, the crite-
rion must be adjusted to a level that protects whatever
species is most sensitive in that area.
o Since the EPA Water Quality Criteria were first introduced
they have been modified fairly extensively. Changes in
values of the criteria can easily change the policy results
of an indexing effort.
VII-12
-------
o Competition among species and food chain effects are not
well represented in the criteria's derivation.
o The Baltimore Harbor environment is estuarine. The sali-
nity varies widely over time and in different areas of the
harbor. EPA, however has only developed salt water and
fresh water quality criteria-. We judged salt water qual-
ity criteria to be more appropriate because the majority
of the species were either salt-water or anadromous
species (Anadromous means the species spends most of its
adult life in salt-water but spawns in fresh or brackish
water).
o Consideration of physical factors like water temperature,
availability of spawning areas, adequacy of canopy cover,
turbidity and similar habitat factors is not possible
within indexing. Indexing also does not address "conven-
tional" pollutants such as nutrients, dissolved oxygen,
and salinity.
o Interactions among multiple toxic pollutants cannot be
assessed. Even if all contamination levels are within
permissible limits, the cumulative effects may still pro-
duce adverse reactions.
o Even if a species can theoretically survive in a particu-
lar environment, it may avoid that environment in prefe-
rence to less contaminated ones. There may therefore be
no guarantee that action based on index values will result
in any increase in local populations of desired species.
RESULTS OF INDEXING EXERCISE
Indexing was used in several ways to set priorities. We
used indexing to rank metals in ambient water in the harbor and
the tributaries and the sediments. We also used it to compare
ambient water in the harbor with that in the Chesapeake Bay.
Metals in Ambient Water
The first set of results pertains to toxic metals in ambient
water in the harbor. We used the two data sets which were avail-
able for toxics concentrations in ambient water. We used them
separately rather than averaging them together because they were
devised in different ways. For example, they were collected
several years apart, using different monitoring, sites. Table
VI1-3 provides the range of values, by pollutant, for each data
set. One data set came from a study by Dr. Helz of the Universi-
ty of Maryland, which was summarized in a report prepared by Tri-
dent, Inc., for EPA in 1977 (the Trident data set). The IEMP data
were collected in a monitoring study performed in 1984 (the IEMP
data set).
VI1-13
-------
Figure VI1-3
Baltimore IE HP
Baseline Results of Indexing
for Metals in Ambient Water
in Baltimore Harbor
8.0
NICKEL
COPPER
CHROMIUM
ZWC
CADMIUM
DJ07
0.13
0.08
£,
J04
fi TRDEHTDATA^
.9
J04
-4-
H
0123456789
Indexing Score = Ambient Vater Conc./EPA Ctronfc Saltwater Criteria (1989)
1 When sample concentrations were found to be less than the detection
limit, they would range from zero up to the detection limit. This table
provides the baseline case, using an assumption that undetected values
are equal to half the detection limit. This assumption holds for both
sets of data (IEMP and Trident), (see p. VII-14 for explanation).
2Mean values for individual pollutants at 13 sites, from IEMP
monitoring of Baltimore Harbor in 1984.
3Mean values from U.S. EPA, "Evaluation of the Problems Posed by In-Place
Pollutants in Baltimore Harbor and Recommendation of Corrective
Action", Trident Engineering Associates, Annapolis, MD, 1977.
-------
These monitoring data were used in indexing, as shown in
Figure VI1-3. However, for the IEMP data, there was some uncer-
tainty as to the best way to interpret non-detected values,
especially since many of the detected values were not far above
the detection limits. Various methods of accounting for samples
below the detection limit were tried. These included assuming
all samples below detection were: the detection limit, half the
detection limit, or zero.
The harbor work group felt that the second approach above,
was the best to use for indexing, as long as it was understood by
the group that the actual situation could be more or less conser-
vative then the chosen option. Figure VII-3 shows the results,
assuming that whenever samples were not detected by IEMP, they
were averaged in at half the detection limit. For two metals,
mercury, and nickel, both data sets (IEMP and Trident) resulted
in scores greater than one, (i.e. ambient concentrations were
greater than the EPA's chronic ambient water quality criteria).
For lead and copper, only the Trident data set resulted in
concern. For zinc, cadmium, and chromium, neither data set
resulted in concern.
Sensitivity Analysis Based on Assumptions for Detection Limits
The next set of results consisted of sensitivity analysis
based on differing assumptions for detection limits, using the
IEMP data. The two alternatives to the preferred approach were
described on the proceeding page. Results are provided in Figure
VII-4. Using the most conservative (pessimistic) assumption
(when a metal was not detected, it was at the detection limit),
four metals, mercury, nickel, copper, and lead, were of concern.
The more optimistic assumption (regarding all samples below detec-
tion limits as zero) provides a lower bound to the expected
impact. Using that assumption, the IEMP data shows only mercury
to be of concern.
Metals in the Tributaries
The third set of results for indexing was based on levels of
metals in the tributaries to the harbor compared to EPA fresh-
water quality criteria. The source of this data on metals in the
tributaries, is the Jones Falls National Urban Runoff Program
Data.8 Monitoring data (and criteria were available for five
metals. The results are provided on Figure VI1-5. Again, mer-
cury is of greatest concern, although all the metals monitored
were of concern. (Two of these metals, chromium and zinc, were
not of concern in ambient water in the harbor. A third, lead,
was of concern in ambient water in the harbor according to one
data set but not the other, as shown in VII-3.)
VII-14
-------
Figure VII-4
Baltimore IEMP
/
Sensitivity Analysis Based on Varying Assumptions
Regarding Detection Limits Used in Indexing for Metals in
Ambient Water in Baltimore Harbor
1
2345678
Indexing Score = Ambient Concentration/
EPA Saltwater Chronic Criteria (1985)
1
Based on average ambient water quality, from IEMP monitoring
of Baltimore Harbor in 1984
dl = dl/2. This is the baseline option. Where monitoring
values were less than detection limits, we used half the
detection limit
dl = dl. This is the most pessimistic option. Where moni-
tored values were less than detection limit, we used the
detection limit.
dl = 0. This is the most optimistic option. Where monitored
values were less than the detection limit, we used zero.
-------
Figure VII-5
Baltimore IEMP
Results of Indexing for
Metals in Tributaries to Baltimore Harbor
Mercury
Lead
16.7
Copper
Chromium (VI)
2468 10 12 14 16
Indexing Score = Estimated Ambient Water Coney
EPA Chronic Freshwater Criteria (1985)
18
Source of monitoring data is the Baltimore Regional Planning
Council, "Jones Falls Watershed Urban Storm Water Runoff
Project," Interim Report. Baltimore, MD, 1982.
-------
VI1-14
Metals in the Sediments
The fourth set of results for indexing, provided in Figure
VII-6, was based on levels of metals in the sediments in the har-
bor and the bay. The source of the data for sediments was a
study which G.R. Helz published in 1976,, and which is referenced
in "National Perspective on Sediment Quality", EPA, 1985. This
report also provides the "threshold contamination concentrations"
which we used as criteria for comparison with ambient concentr-
ations of metals in the sediment.
As Figure VII-6 shows, the indexing score for chromium was
higher than the scores for other metals in the harbor. It is
also interesting to note that the score for chromium in the har-
bor was higher than that for chromium in the bay. This makes
sense because there has been a great amount of chromium waste
deposited in and around the harbor area over the last several
decades.
Comparison of Metals in the Harbor with Metals in the Bay
The fifth set of indexing results compared levels of pollu-
tants in the eleven IEMP monitoring sites in the harbor with the
two IEMP monitoring sites in the bay (Figure VII-7). This data
set showed copper, nickel, and mercury to be above water quality
criteria at these two sites in the bay. (See Figure VII-2 for
their locations.) This data set was too small to be conclusive,
and it was not evaluated statistically. Nevertheless., it was
interesting that not all of the pollutants were higher in the
harbor; copper was higher in the bay. It might be worth inves-
tigating this further to determine if this is a consistent
occurrence and if so, what its causes are.
Summary of Indexing Results
Table VII-4 provides a summary of the indexing results pro-
vided in the last five figures. The members of the harbor work-
group used this information to reach their final conclusions.
They were concerned about the level of mercury, nickel, copper,
chromium, lead and zinc.
They also concluded that indexing did not result in any
individual metals ranking far enough above the others, for aqua-
tic ecological impact, to justify singling out one or a few metals
from the above metals for Phase II study. Therefore, given the
similar results for the different metals and the uncertainties in
the data, they felt that additional work should be performed in
Phase II on the six metals mentioned above. They decided to use
eco-scoring to perform this analysis in Phase II because it was
the best developed of the methodologies and had the most support-
ing data.
VII-15
-------
Figure VII-6
Baltimore IEMP
Results of Indexing for Metals in Baltimore Harbor Sediments
Chromium
Copper
Lead
Nickel
Chesapeake Bay
Baltimore Harbor
0 2 4
10 12 14 16 18 20
Indexing Score - Ambient Sediment ConcVDrafl EPA Sediment Threshold Cone. Value
Source of sediment data is a study by Dr. Helz'of the University of
Maryland, which was summarized in a report prepared by Trident
Engineering Associates, Inc. for EPA in 1977.
-------
Figure VI1-7
Baltimore IEMP
Comparison of Ambient Water Quality Indexing Values
for IEMP Harbor Sites with IEMP Bay Sites
Mercury
Chesapeake Bay
Baltimore Harbor
123 4 5
Indexing Score = Ambient Water Concentration/
EPA Chronic Saltwater Criteria (1985)
Source of data is the IEMP monitoring performed at thirteen sites in and
around the Baltimore Harbor in 1984.
We used the eleven IEMP sites in the harbor and compared these to two
IEMP sites in the bay. (See Figure VII-1 for map of monitoring sites.
-------
Table VII-4
Baltimore IEMP
Comparison of Index Values For Harbor Priority-Setting,
Using Different Types of Indexing Techniques *
Ambient Water Quality
Harbor
Trident IfiMD
' Data1 Data2
Ambient Water
Quality-Tributaries
Ambient Sediment Ambient Sediment
Quality-Baltimore Quality-Chesapeake
Harbor Bay
Zinc
Nickel + +
Mercury 0 O
Lead + °
Copper O +
Chromium ° °
Cadmium ° °
>
N/A
O
O
0
+
N/A
+ N/A
+ +
N/A N/A.
0 N/A
O N/A
0 +
N/A N/A
O Index value greater than 2.
Criteria)
+ Index value greater than 1.
0 Index values less than 1.
N/A Not Available.
(Ambient values are more than twice the level of EPA
(Ambient values are greater than EPA criteria)
(Ambient values are less than EPA criteria).
1. See Figures VI1-6 to VII-10 for a presentation of the indexing scores and the
sources of data used to generate this table.
2. Assumed undetected monitoring values were egual to half the detection limit.
figure VII-7 and p. 14 for further explanation.
See
-------
Limitations of Indexing
Recognizing the limitations of ^indexing is important. A
number of limits to indexing were outlined earlier in general
terms. One limit to the usefulness of indexing for the IEMP
study was that IEMP did not monitor arsenic, cyanides, and selen-
ium. Also, the valence states of metals, particularly chromium
and mercury, can significantly affect the resulting toxicity and
bioavailability for those metals, and this method does not diffe-
rentiate between valence states. This list of six metals, there-
fore, is merely a list for possible concern. More work will be
done in Phase II to further evaluate the metals on the list.
VI1-16
-------
FOOTNOTES
1. Koo, T.S.Y., Pfitzenmeyer H..T.', Dovel, .W.L., Wiley, M.L.,
Lippson, R.L., Miller, R.E. A Biological Study of Baltimore
Harbor, College Park, Maryland, University of Maryland Center
for Environmental and Estuarine Studies, Special Report No. 6,
1975
2. Krebs, Charles J., Ecology: The Experimental Analysis of
Distribution and Abundance, Harper and Row, 1972. """^
3. Odum, Eugene P., Fundamentals of Ecology, W. B. Saunders and
Company, 1971.
4. EPA Discharge Monitoring Reports collected for the National
Pollution Discharge Elimination System, for 1983-1985, except
1982 for Bethlehem Steel.
5. Memo to EPA from Versar, Inc., Baltimore Non-Point Sources,
November 17, 1983.
6. Memo to EPA from Versar, Inc., Baltimore Non-Point Sources,
January 13, 1984
7. Personal Communication between Hope Pillsbury of EPA and
John Veil of Maryland Office of Environmental Programs,
October 1986.
8. Baltimore Regional Planning Council, Jones Falls Watershed
Urban Storm Water Runoff Project, Interim Report, Baltimore,
Maryland, 1982.
9. Memo to EPA from Versar, Inc., Modeling Toxics in Baltimore
Harbor, November 13, 1983, and subsequent discussions of this
memo by IBMP.
10. The source of the proceeding paragraphs on aluminum is personal
communication between Catherine Crane of EPA and Dr. David
Correll, Assistant Director of the Smithsonian Research Center
in Edgewater, Maryland, in Fall, 1985.
11. EPA, Office of Water Planning and Standards, Evaluation of
the Problem Posed by In-Place Pollutants in Baltimore Harbor
and Recommendation of Corrective Action, Trident Engineering
Associates, Inc., Annapolis, Maryland, September 1977.
12. Discussions were held between IEMP staff and Dr. Boicourt of
the University of Maryland, Dr. Biggs of the University of
Delaware, and Mr. DeMoss of EPA, in 1984.
-------
13. Memo to EPA from Versar, Inc., Final Monitoring and Analyti-
cal Plan for Baltimore Harbor/Chesapeake Bay - Phase I "
Activities, August 3, 1984'
14. Memo to EPA from Versar, Inc., Baltimore Harbor Data Report,
November 23, 1984
15. Briefing packages prepared for IEMP by Versar Inc., on
Indexing, Ecoscoring, and Reference Environment Analysis,
dated November and December, 1985.
16. Memo to IEMP from Versar, Inc. Draft Overview of the Use of
Bioassay and Bioaccumulation Studies for Water Quality
Evaluations,1985.
17. Memo to IEMP from Versar, Inc. Review of Biological Studies
Conducted in Baltimore Harbor and Chesapeake Bay, 1985.
18. Bolton, H. Suzanne et.al., National Perspective on Sediment
Quality, U.S. EPA, May 10, 1985.
-------
CHAPTER VIII
COMPLETION OF THE PRIORITY-SETTING PROCESS
-------
VIII. COMPLETION OF THE PRIORITY-SETTING PROCESS
This chapter discusses how the Management and Technical Advis-
ory Committees pared the 10 issues,, to five topics for Phase II
study. It also describes the reevaluation of metals-related issues.
After the subcommittees had made their recommendations on
issues for further study, the full TAG decided to undertake three
further/ interrelated efforts before proceeding to Phase Two.
First, it wanted to reduce the number of issues in order to avoid
overextending the analytic resources of the study. Therefore
it developed a set of secondary criteria to evaluate the ten issues
for policy and pragmatic reasons. Second, to help the TAG members
be more informed about the issues, the subcommittee members with
the assistance of EPA staff developed workplans for each of the
study topics. This provided the information necessary to apply the
secondary criteria, determine the budgets, and then of course to do
the work.
The third effort involved the umbrella topic of toxic metals
in the environment. Upon closer scrutiny, the TAG realized that
the issue was too broad; it contained too many issues that needed
to be ranked in importance. Thus, the TAG decided to reevaluate
these numerous issues against the primary criteria to arrive at a
smaller set of issues dealing with toxic metals in the environment.
These would then also be evaluated against the secondary criteria.
We discuss this second and separate priority-setting process for
metals later in this chapter.
REDUCING THE NUMBER OF ISSUES FOR STUDY
As discussed in Chapter IV, the TAG chairman asked the sub-
committee chairs to recommend more issues than the project would
have the resources to study. Once these were identified, the full
TAG would narrow the number through the application of secondary
criteria by consensus. Discussion would focus on the social and
economic importance of issues and on trade-offs between human health
risk and ecological impact. This latter type of trade-off was never
made explicitly because the list was pared down sufficiently using
other criteria.
The secondary criteria were:
1) Probability of making a significant contribution to local
environmental management through the Phase II study of the
issues.
2) Lack of duplication with existing analyses or control programs.
3) Technical and political feasibility of implementing controls
for each issue studied,
VIII-1
-------
4) Feasibility of performing the analysis within the time and fiscal
constraints.
The TAG chose this strategy because of its conviction that it
was inappropriate, and probably infeasible, to develop a a quantita-
tive, scientifically-based approach to compare priorities across
the reference categories of human health, ecological, and ground-
water. The TAG judged that its chances of success would be greater
if it concentrated its debate on tradeoffs in the context of specific
issues rather than on abstract arguments about whether human health
is more important than ecological protection or the integrity of
groundwater.
The TAG was also aware that the decision to create three
subcommittees, each responsible for setting priorities among the
issues in a reference category, would virtually guarantee that at
least one health, ecological, and ground-water issue would emerge
as Phase II study topics. Furthermore, since each member of the
TAG had a personal and institutional interest in one of the study
areas (health, ecology, or ground-water), the creation of the
subcommittees formalized existing interests. Nevertheless, the
implicit understanding that at least one issue in each category
would "survive" into Phase II was an important element in the pro-
cess of building a consensus.
REDUCING THE NUMBER OF ISSUES FROM TEN TO SIX
In practice, further reduction of the total number of issues
proved relatively easy because of the considerable overlap among
the study topics recommended. As Figure VIII-1 shows, all three of
the subcommittees recommended conducting further work on metals
in the second phase of the project. The remaining ecological
problems, the bioaccumulation of toxics, and sediments in the
Harbor as a source of water pollution, were combined into a
single study topic encompassing Harbor pollution. Finally, two
of the health issues recommended for further work, benzene and
the other toxics in air, were collapsed into a single study
topic.
Because of these overlaps, the original ten issues collapsed
into the six topics listed in Figure VIII-1. This number was near
the target implicity set by the Management Committee and the TAG
when they first conceived the priority-setting exercise (see Chapter
IV). Thus the full TAG did not need to eliminate any study topics
from those recommended by the subcommittees.
VIII-2
-------
Figure VIII-1
Flow Chart of How the TAC and the Workgroups
Pared down the List of Issues from Ten to Five
10 Issues
Human Health Risk Subcommittee
Benzene
Toxic Air Pollutior
Trihalomethanes in
Drinking Water
Metals in Air-
Indoor Air Pollu!
Ecological Impact Subcommittee
Metals-
Sediments
Bioaccumulation
Toxics in
Aquatic Organises
Combine and
Redefine to
6 Issues
Air Pollution
Trihalomethanes in
Indoor Air Pollution
Metals in the
Environment
Harbor
Ground-Water Resourced/Carnage Subcommittee
Metals in Ground'
Water
Underground Storage
Tanks
Underground Storage
Tanks
Apply Secondary
Criteria
Tonic Air Pollution
Indoor Air Pollution
Metals in the
Environment
Harbor
Underground Storage
Tanks
Redo
Primary /Secondary
Criteria for Metals
Resulting in Final
List of Issues
Tonic Air Pollution
Indoor Air Pollution
Underground Storage
Tanks
1\Continued in Phase II for comparison, but not for further study
-------
APPLYING THE SECONDARY CRITERIA
After reducing the number of study topics to six, the full
TAG formalized and applied a set of secondary criteria to each of
the remaining issues. With the explicit elaboration of the secondary
criteria, considerations of policy and feasibility entered the
priority-setting process for the first time in a formal way.
The use of the secondary criteria was not, of course, the first
time that policy and feasibility considerations had played a role
in shaping the study. In developing boundaries for the study area
at the outset of the project and in deciding to focus on certain
classes of pollutants, the project's managers were making implicit
policy decisions. For instance, worries about the analytic feasibi-
lity of the study led to the decision to focus on a relatively
small geographic area. Similarly, considerations of the probability
of making a significant contribution to local environmental manage-
ment were largely responsible for the decision to exclude convention-
al air pollution from the IEMP. At this point in the process of
setting priorities, however, practical considerations began to
replace scientifically-based analysis in determining what to study in
Phase II.
Considerations of policy and resourceis were also important in
the discussions of the TAG and of its subcommittees during the
initial screen. Throughout the first phase of the IEMP, both the
TAG and the Management Committee were careful to insure that the
project did not spend its time and effort researching issues to
which it could not reasonably make a significant contribution. In
particular, the committees wanted to concentrate on topics that
could be analyzed within the resource constraints of the IEMP and
that could be addressed by the state and local governments partici-
pating in the study. The committees also wanted to avoid duplicat-
ing the efforts of other analytic studies.
Many issues, such as pathogens in septic tanks, never made
the initial screening list simply because neither EPA nor State
and local officials believed that the IEMP study of them could
contribute meaningfully to their understanding. Some issues such
as pesticides, were rejected as study candidates during the sub-
committees' meetings because existing data (such as .on exposures)
were so poor that the IEMP had no reasonable prospect of analyzing
the problem in a meaningful way. For the most part, however, the
TAG deferred consideration of secondary criteria until completion
of the subcommittee work.
VII1-3
-------
These "secondary criteria" of pragmatic considerations of
time and resources were formalized about the time that the three
methodological subcommittees concluded their deliberations, and
emerged with a list of. ten issues. The criteria were not applied
until the TAG reduced the number of issues to six, as previously
described.
Most of the secondary criteria are straightforward and self-
explanatory. The criterion of "political feasibility of control,"
however, deserves some special attention. Primary management
responsibility for the IEMP rested with state and local govern-
ments, who did not wish to spend time analyzing issues they had
no reasonable prospect of controlling. Some environmental prob-
lems, such as acid rain, are simply unamenable to state or local
control because they entail interstate transport of pollutants.
As it turned out, no potential topics were excluded from the
study on the basis of this particular criterion.
Use of the first criterion, probability of making a signifi-
cant contribution to local environmental management, included
possible duplication of existing or planned programs and the
degree to which the problem could be characterized as national
rather than site-specific. Problems which were national in scope
and which were not expected to be different in Baltimore than in
the nation as a whole were considered poor candidates for further
work. The opportunity to participate in a study conducted by
EPA's Office of Research and Development made indoor air an
important exception. One of the six topics then under consider-
tion for study in Phase II, trihalomethanes (THMs) in drinking
water, was deemed by the TAG to perform poorly in terms of these
criteria. As explained in Chapter IV, THMs are present in virtu-
ally all chlorinated surface water used for drinking water in
this country, or about half of the total drinking water. They
are present because they are a by-product of chlorination required
to kill pathogens in the water.
Baltimore's water supply is well within the federal limit for
THMs and the federal limit is currently under reconsideration by
the EPA. The committee felt that the exposures, risks, and
controls for THMs are well understood for purposes of risk manage-
ment and that there was a low probability that a Phase II study
of this issue would make a significant contribution to local
environmental management. Therefore the TAG and MC decided that
THMs would not be a Phase II study topic. It was not, however,
dropped completely from Phase II consideration. The Management
Committee felt that the relatively well-characterized health
risks from THMs in drinking water could later be compared to the
risks associated with the other Phase II issues.
VII1-4
-------
Use of the second and third criteria entailed consideration
of other projects being conducted by federal, state, and local
governments, as well as a rough appraisal of the extent to which
conceivable controls for a particular problem would be feasible
at the state or local level. No study topics were eliminated on
the basis of either criterion because the studies could be struct-
ured to enhance, rather than duplicate existing State or local
efforts.
Use of the fourth criterion, analytic feasibility, was the
most difficult. For obvious reasons, the managers of the project
wanted to concentrate their attention and funds on problems that
could be feasibly analyzed within budget limitations. Understand-
ing what analysis of each issue was possible within those funding
constraints, and what the project could hope to achieve on each
possible topic, was critical. It was also virtually impossible
to do casually, without a reasonably thorough investigation of
available data, possible monitoring and analytic strategies,
costs of various analytic schemes, and so forth. For this reason,
the TAG decided to form workgroups to develop work plans for each
of the five study topics so it could directly compare the options.
DEVELOPING THE PHASE II WORKPLANS
The TAG drew up an initial set of Work proposals for each
issue which provided the objectives and purpose of the study,
alternative analytical procedures, and the expected product of
the analysis. These were then presented to the MC to help decide
the merits of each study topic. The MC, however, still did not
feel they had enough information to decide which of the five
issues should be studied in Phase II (excluding, for the time-
being, THMs). They requested that the TAC provide estimates
of funding needs for each each element of the work plan and
recommend an option for analysis. Upon revision of the workplans,
the TAC met in full to approve the workplans and pass on its
recommendations to the Management Committee.
The TAC created five workgroups to explore each of the five
topics. It decided that these work groups would include represen-
tation from institutions outside government. Individuals from
industry, trade associations, and public interest groups served
on the workgroups, although they did not become members of the
TAC as a whole.
The TAC chairman assured the MC that each workgroup member
would be informed that no workplan was guaranteed of being
performed, either in whole or in part. Nevertheless, he felt the
added expertise available from the additional work group members
was worth the risk. The MC then decided that the TAC should
proceed with formation of workgroups and made suggestions as
to appropriate organizations to contact. Upon completion of the
workplans, the TAC would meet in full to approve the workplans
and pass on its recommendations to the Management Committee.
VIII-5
-------
The chairs of two workgroups felt that their work plans
should be better formulated before they knew which additional
members of industry or public interest groups would be most
knowledgeable and helpful to the project. Therefore, they only
added some additional governmental members at this stage of
development of the work plan. Others members would be added as
as needed.
For all the work plans, investigating existing regulations,
programs, and plans was essential to ensure that the proposed
work supported existing programs and would help complement activi-
ties or provide resources for studies that would otherwise
not be possible.
During this time, some, additional technical work was done on
most of the issues to revise or supplement Phase I analysis. The
issue of metals required the greatest technical work.
SPECIAL ATTENTION GIVEN TO METALS IN THE ENVIRONMENT
The TAG recognized that the topic, "metals in the environ-
ment," encompassed too many topics whose study could easily
exceed the anticipated resources for phase II. Narrowing its
scope meant disaggregating the issue into its components and
ranking these, in turn, by the primary and secondary criteria.
The TAG formed a special metals workgroup charged specifically
with conducting this task.
In the initial run-through of the priority-setting process,
described in Chapter IV, the issue had been defined in terms
of possible sources of metals released or potentially released
into the environment. In this second round of screening, the
the workgroup decided upon defining metals in the environment
in terms of their concentrations in various receptor environments.
This second run-through also differed from the first in that
new information was available to the workgroup members. They
now had indexing scores for metals in the tributaries to Baltimore
Harbor. (See the discussion of indexing scores in the discussion
of the Harbor in Chapter VII.). Furthermore, unlike the first
round of priority-setting, EPA staff now had available to them a
model that could provide rough estimates of the human health
risks to individuals which could arise from potential degradation
of ground-water resources.
The metals workgroup began with fifteen metals, which are
listed in Table VIII-1. [A sixteenth was considered if one takes
into account the separate evaluation of the two forms of chromium
tri- and hexavalent chromiumthat are commonly found in the
VIII-6
-------
environment.] The workgroup members chose these metals on the
basis of their judgment regarding their likelihood of being a
pollution problem. They were drawn from an initial list of 129
priority pollutants (only a fraction of which were metals) developed
by EPA's Water Office.
TABLE VIII-1
Metals Considered In the Phase I Screen by
the .Baltimore IEMP Metals Work Group
Aluminum Antimony
Arsenic Barium
Beryllium Cadmium
Chromium Thallium
Lead Copper
Nickel Mercury
Silver Selenium
Zinc
After a perusal of the available data, the workgroup discovered
that only the data on nine metals would allow a limited evaluation
against the primary criteria. For the other metals, the little
information that existed suggested that they were unlikely to have
significant impact in any likely scenario for the Baltimore area.
Also, there was no significant source of these metals in the study
area. Given its limited time and resources, the workgroup was
constrained to examine only those metals for which there was some
minimally adequate data indicative of a possible threat to the
environment. Thus the metals that remained under consideration
were aluminum, arsenic, chromium, lead, zinc, cadmium, mercury,
copper, and nickel.
The primary criteria for screening these metals remained
human health risk, ecological impact, and potential for damaging
ground-water resources.
The workgroup estimated the human health risk posed by these
nine metals through ingestion of water and fish, and inhalation
of dust. Because data on exposures for each metal differed greatly
in kind and quality, risk estimates are not directly comparable.
Again, the workgroup hadi often to rely on best judgement in deve-
loping an assessment of the total risk of each through all path-
ways of exposure. Furthermore, it had to decide the relative
weight of actual incidence of disease attributable to a metal and
postulated risk of disease that occurs sometime in the future.
Because no new data became available for this round of
priority-setting for metals, the estimation and relative ranking
VIII-7
-------
of risk through inhalation and ingestion of metals did not
change. These estimates of risk are presented in Chapter V:
Human Health Risks. As before, chromium appeared to be the metal
of primary concern in the air. Only lead presented significant
health risks through ingestion.
The workgroup was particularly concerned by the immediacy
of the public health problem posed by lead. There is strong
evidence that the health of people in the study area can be
adversely affected by lead in water and that children are now
suffering the consequences of lead in household dust. Although
there is no evidence that elevated lead levels exist in the
water supply, two conditions can contribute high lead to the
water faucet or fountain: (1) new or replumbed buildings with
copper pipe and lead solder or (2) older buildings with lead pipes.
Levels of lead in tap water, especially in the first few minutes
of flow have been found to exceed existing drinking water standards.
In addition, the work group had data on actual cases of lead
poisoning in children. Each year roughly 150 hospitalizations of
children per year occur from ingestion of lead in household where
the residences had been painted with lead-based paint. In 1985,
290 Baltimore City children were diagnosed as suffering from lead
poisoning. [A more complete discussion of the problem of lead
poisoning among children in Baltimore City is presented in Chapter
IV.] These numbers could underestimate the actual number of lead
poisonings in' the area because of under-reporting. On the other
hand, there could also be double counting of lead poisoning cases
occurring to the same child.
The workgroup used the ecological indexing system, explained
in Chapter VII, to identify metals which could adversely affect
the aquatic ecology of the Harbor. Figure VI1-6 showed that the
highest scores were for mercury and nickel in the water of the
harbor. Mercury, zinc, copper, chromium, and aluminum were high
in the tributaries. High concentrations of chromium were found
in the sediments.
There are few data on metal concentrations in ground-water.
Details are also very scant for what ground-water resources are
at risk of contamination from metalsespecially sources that
are not now serving as sources of drinking water but could be
used in the future.
In the absence of monitoring data, the workgroup used best
professional judgement and an indexing scheme (which we describe
in Chapter VI) to identify metals and their sources that could
pose significant risks to ground-water resources. From the
screening analysis, the work group identified eight significant
metal-related scenarios. They also determined the most practical
approach for studying each. The issues and their approaches are
listed below:
o lead intake by children from paint, dirt, and air
VIII-8
-------
(testing and evaluation of lead abatement techniques in
the study area);
o human health risk from lead'intake from lead-containing
solder in drinking water pipes, in areas with acidic
water and/or new piping;
o health risks from metals such as cadmium, chromium, and
arsenic through ingestion of locally harvested finfish
and shellfish (only if OEP monitoring data suggests
significant health risks);
o human health risks from ambient air exposure to chromium,
cadmium, arsenic and nickel;
o potential risks to human health, and predicted impacts
to the ground-water resource, from arsenic and other
metals leaching into ground-water from municipal
landfills;
o ground-water resource damage from chromium leaching into
ground-water from the harbor area;
o impacts on aquatic life in the harbor from exposure to
lead, nickel, copper, mercury in the ambient water, and
these metals and chromium in the sediments; and
o impacts on aquatic life in the tributaries to estuarine
waters from leaching of aluminum and other metals from
soils.
To determine the extent of additional work the metals
workgroup should undertake for the above issues, the workgroup
developed and applied the following secondary criteria for judging
their feasibility: (1) whether the issue is being addressed by
other work groups in the IEMP, (2) whether the issue is being
considered by State or local governments, (3) whether the issue
is technically or politically feasible for study in the IEMP, (4)
whether the cost of studying the issue is too great, and (5)
whether State or local governments will be able to effect controls
for any problems the IEMP discovers.
After applying the criteria, the workgroup found that
metals in the ambient air and impacts on aquatic life were to be
handled by other work g,roups. The metals workgroup would only
serve in a coordinating role. Lead in drinking-water was likely
to be addressed by the State. Monitoring for arsenic and other
metals in ground-water or aluminum in tributaries was beyond the
resources of the IEMP. Similarly, the IEMP would not be able to
do anything practical to resolve possible problems from metals
leaching into ground-water from the harbor because these ground-
waters are already severely degraded. Thus, the workgroup decided
to recommend only lead intake by children and risks associated with
harbor fish for Phase II. The latter was dropped when Maryland's
VIII-9
-------
Office of Environmental Programs completed its analysis.1
PRIORITY SETTING THROUGH BUDGET ALLOCATION
The final stage of the priority-setting exercise was to
reconcile the workplans proposed by the five workgroups with the
limitations of the IEMP budget. The workgroups proposed work for
Phase Two that would have cost about two and one-half times
as much as the project had to spend. Deciding where to cut back
the workgroups' requests, a process which occupied the TAG for
three months, became one of the central priority-setting devices
for the study. In deciding how much to cut each budget, the TAG
as a whole was in effect setting priorities among the problems
to be analyzed.
Each workgroup prepared work plans for three budget levels,
asking in most cases for the highest level of funding from the
TAG as a whole. Although decisions were again made by consensus,
the meetings at which the resources were allocated were somewhat
more contentious than most of the TAC's other meetings. The TAG
tried during these meetings to use other forms of priority-setting
as well. Several TAG members felt that the overall budget was
simply inadequate to undertake analyses of all five issues and
wanted to eliminate one or more entirely. The TAG could not,
however, agree on which might be eliminated. Compromises were
made on funding levels for projects and through extending work
periods.
Finally, the budget recommendations and the study alternatives
were presented to the MC for their review and approval. It discuss-
ed the pros and cons of each study at length,.particularly focus-
ing on the need to avoid duplication with existing programs. It
also discussed the possibility of omitting one study from the
list. It took two meetings for the MC to decide to accept the
TAC's recommendations on the issues and the funding allocations
and to approve the work plans. These work plans are presented in
Chapter IX. They comprise the blueprint for the second phase of
the Baltimore IEMP.
VIII-10
-------
1. OEP found levels of metals in fish were below FDA limits.
VIII-11
-------
CHAPTER IX
PHASE II WORK PLANS
-------
IX. PHASE II WORK PLANS
In Chapter IV and the last chapter, we described how the MC
and the TAG set priorities among the 32 issues that the TAG had
identified to be of possible concern in the Baltimore study area.
In this chapter, we present the work plans of the five issues
that the MC chose to be the focus of more detailed analysis and
assessment in Phase II of the Baltimore IEMP.
The work plans are not intended to be detailed, technical
blueprints for how the selected issues are to be studied in Phase
II. In general, they provide only the general scope of the
projects and summarize the goals and the analytical approaches.
More detailed plans were developed separately.
SUMMARY OF WORKPLANS FOR PHASE II
Issue #1: Air Toxics
Ambient (outside air) monitoring data for various pollutants
suggest that air toxics may pose a threat to public health. Air
toxics may also transfer to other media, such as surface water
(e.g., the transfer of metals from the air to the Harbor).
Because of their health risk and the appropriateness of their
examination using IEMP methodology, TAG has chosen the issue of
air toxics for study in Phase II.
The Technical Advisory Group has identified the goals of the
workplan for air toxics to be:
o Estimate ambient air concentrations of selected air toxics
which result from emissions from both point (industrial)
and non-point (area) sources. A particular emphasis is
upon identifying possible "hotspot" exposure, that is,
areas where emission plumes may overlap resulting in higher
than expected concentrations of certain pollutants.
o Analyze associated risks to human health and the environ-
ment. This will entail identifying and quantifying (where
possible) risks to human health or the environment from
ambient concentrations resulting from air emissions and
identifying the sources that are significant contributors
of risk.
o Analyze and develop of control strategies to reduce the
adverse health effects from emission of selected air
toxics.
IX-1
-------
There are three major tasks associated with these goals:
(1) estimate ambient levels of selected pollutants. For
this purpose, we will use widely used ambient air dis-
persion models and data on' industrial emissions from
the Maryland Office of Environmental Programs (OEP)
Science and Health Advisory Group's Hazardous Materiials
Survey. We will also estimate the contribution to
ambient levels of certain pollutants from area sources,
such as transportation vehicles and dry cleaners. The
pollutants will include volatile organic compounds and
metals, selected on the basis of their chronic toxicity
and their likelihood of being present in the ambient air
of the Baltimore study area.
(2) develop quantitative estimations of risk from ambient
exposures to the selected toxics and to use these
estimates to identify the pollutants and their sources
for which control strategies will be developed.
(3) analyze and develop control strategies for point and area
sources, taking into account the potential for pollutant
transfer across media. The analysis will include the
estimation of costs and of the reductions in risk of
alternative control strategies.
A related task is overseeing, along with the indoor air work
group, the Total Exposure Assessment Methodology (TEAM) study
that will be conducted in the Baltimore study area by EPA's
Office of Research and Development. Oversight will involve re-
viewing the study design and objectives, coordinating with the
MC and State and local agencies with regard to publicity and
selection of geographic sites within the Baltimore study area for
monitoring, and recommending pollutants for inclusion in the
study. Ambient monitoring data generated as part of the TEAM
study will serve to corroborate the air toxics work group's air
dispersion modelling estimates. Conversely, the air toxics work
group's air dispersion modelling results will be used in inter-
preting the results from the TEAM study.
The goals of the TEAM study are:
0 Apply "modified" TEAM methodology to Baltimore to esti-
mate exposure of Baltimore area residents of specific
geographic areas to selected volatile organic compounds;
0 Compare modeled concentrations with measured ambient
levels for selected volatile organic compounds;
0 Compare indoor concentrations, outdoor concentrations,
and personal exposures.
IX-2
-------
Fixed monitors will allow comparison of indoor and outdoor con-
centrations of selected pollutants and the contribution of out-
door levels to indoor concentrations. Volunteers will carry
personal monitors to permit estimation of total personal exposure
to pollutants as the person goes about his or her daily affairs.
Issue #2: Indoor Air Pollution
The TAG and the MC chose the issue of indoor (i.e., residen-
tial) air pollution for the following reasons:
o Data from other cities suggest that indoor pollution may
pose significant risk to human health.
o There is little local data on whether indoor pollution
is also a problem in the Baltimore area.
o Resolution of the problem, if there is one, is likely to
cut across a number of government agency lines and thus
be particularly appropriate for an IEMP project.
People spend the majority of their time indoors in resident-
ial, commercial, and office buildings. Thus even though indoor
air pollutants may be present at very low levels, they may never-
theless make a significant contribution to personal exposure.
Indeed, those groups of individuals most at risk from the adverse
health effects of toxic substances, such as infants, young child-
ren, the chronically ill, and the elderly, spend even greater
amounts of time indoors.
A number of factors, such as measures for increasing energy
efficiency, cleaning products, synthetic and natural building
materials, unvented heating and cooking appliances, and pesti-
cides, combine to increase the likelihood of exposure to a mix-
ture of indoor pollutants. Many of these pollutants may be
present at levels greater than for those currently controlled by
ambient air quality standards. Yet the health effects of expo-
sure to such mixtures is largely unknown.
The goals of the workplan are:
o Learn more about indoor air quality in the Baltimore
metropolitan area
o Investigate possible programs to reduce exposure to indoor
air pollutants and to recommend implementation of potential
programs to reduce air pollution.
The tasks associated with these goals are:
(1) review studies concerning indoor air pollution .to estab-
lish the importance and extent of the problem and its
relevance to the Baltimore study area;
IX-3
-------
(2) examine possible programs to reduce indoor air pollution;
and
(3) recommend programs to reduce indoor air pollution where
appropriate.
Issue #3: Underground Storage Tanks
The TAG and the MC chose leaking underground storage tanks
(USTs) as an issue to be studied in Phase II because USTs rated
high on a variety of factors associated with ground-water conta-
mination and resource impact.
Leaking underground storage tanks pose potentially signifi-
cant pollution problems in the Baltimore area. Reducing the
problem through development of requirements for underground tanks
and through identification of tanks most likely to leak has
proved to be a difficult undertaking. The characteristics of tank
failure are hard to predict; the timing and magnitude of tank
failure are highly variable even for the same type of tank under
similar circumstances. In addition, once a leak occurs, the
damage from leaks will vary even for the same type and volume of
release. Local hydrogeologic conditions, the proximity to popu-
lation, the use of ground-water, and ecological factors all play
important roles. Given limited resources, government authorities
face a need to set priorities regarding inspection and enforce-
ment activities. Those tanks located in areas of greatest hydro-
geologic vulnerability and where groundwater wells are densely
located, deserve high priority.
The goals of the workplan for underground tanks are as
follows:
0 Set priorities for inspection and enforcement. Which
leaking tanks are most likely to pose the greatest
ground-water resource damage?
0 Provide a tool which can be readily used by state and
local officials in the study area as well as elsewhere,
to help in setting priorities for managing underground
storage tanks.
The workplan identifies the following tasks necessary to
achieve these goals:
(1) develop an information management tool which can operate
on a personal computer, and which displays and integrates
information on hydrogeologic conditions, population,
ground-water use, and tank location. The tool must be
able to be used and updated on an on-going basis and
allow mapping of factors of concern.
IX-4
-------
(2) develop tools and methods to describe the extent to
which leaking underground storage tanks can potentially
result in impacts on groundwater resources. Implicit
in considering impact on the resource is some considera-
tion of health risk (based on number of people affected
on a polluted well) and economic impacts (based on number
of wells affected) With these instruments, the workgroup
will prepare a matrix displaying areas of greatest ground-
water resource impact.
(3) develop a "cookbook" which would enable other states
and localities to perform similar studies.
Issue #4: Multimedia Metals
The MC and the TAG chose the issue of "multimedia metals"
for Phase II study because metals from a wide variety of sources
and in all media were found to pose potentially significant
health and environmental risks.
The goals of the work plan are
o Characterize the environmental and health risks associated
with selected metals in the study area,
o Determine subsequently which subsets of the issue are of
greatest concern to human health and the environment, and
o Initiate analyses on the cost-effectiveness of control
strategies for issues judged to be of greatest concern.
The tasks associated with these goals are described below:
(1) coordinate with other workgroups on metals-related issues.
(2) complete work associated with the initial scoping of
issues dealing with metals. A part of this would
involve writing up the results of the priority-setting
process for metals, including a description of the
methodology. The second part would entail developing
recommendations for metals-related issues for further
study (but not as part of the IEMP).
(3) develop cost-effective techniques for lead paint removal
and dust abatement.
IX-5
-------
This latter task, because it will constitute the primary focus of
the metals work group in Phase II, demands greater elaboration.
An extensive number of studies dealing with various aspects
of exposure to lead have been carried out in Baltimore. In par-
ticular, those conducted under the aegis of the Maryland Depart-
ment of Health and Mental Hygiene and Johns Hopkins University
indicate that current techniques for abating lead paint and dust
are not very effective, pose unnecessary risks to workers and, in
some cases, actually increase the levels of lead dust in housing.
The purpose of the project will be to design and test a
technique for cost-effectively reducing lead dust levels in
households. The pilot project will be carried out by Drs.
Chisholm and Farfel of Johns Hopkins University. The approach
consists of three phases:
o Planning Phaseidentifying and reviewing abatement
techniques. The Kennedy Institute, with the assistance
of the Baltimore Building Congress and Exchange, will
convene a meeting of experts to discuss the technical
and economic feasibility of various abatement techniques.
The advisory body will include local experts in contract-
ing architecture, paint and building supplies, and envi-
ronmental health. Also represented will be organizations
involved in upgrading low-income housing, which may volun-
o Demonstration Phasedemonstrating lead abatement techni-
ques. For about six to ten homes, depending upon the
cost per house, the contractor will take household dust
samples before and after abatement to assess the effec-
tiveness of each technique. We will document the costs
in both capital and labor which are associated with each
technique.
o Evaluationevaluating cost-effectiveness of each abate-
ment technique. We will assess the costs of each abate-
ment technique against the reduction in lead in dust
which is achieved in each home. These costs will also
be evaluated against the costs incurred by the city'^s
current practices for handling homes with levels of lead
paint deemed harmful to children.
The expected product of this project will be an assessment
of the cost-effectiveness of alternative techniques for abating
lead paint in residential homes which are adaptable to the
Baltimore region. The project will also produce trained abate-
ment teams that can efficiently and effectively conduct abatement
efforts in area housing.
IX-6
-------
Issue 15: Baltimore Harbor
The TAG and the MC chose Baltimore harbor as an issue for
Phase II because of the importance that pollution of the harbor
can have on current and future uses of this vital resource.
The eco-system within Baltimore Harbor has been affected by
urban and industrial discharges over the past century and a half.
Various corrective measures, such as effluent limits, sediment
controls, and regulation of dredging spoils, are now in place.
Nevertheless, the extent to which additional remedial efforts are
in order awaits policy decisions regarding the ecological state
to which the Harbor is to be restored. These plans for the
Harbor are, in turn, dependent upon a thorough understanding of
ecosystem dynamics and the determinants of ecological health.
For example, environmental managers must understand how environ-
mental variables (such as pollutants and aquatic habitats) inter-
act to produce a potential ecological state (e.g., the ability
of fish to breed or rockfish to survive) before they can devise
effective control strategies for the pollutant.
The objectives of the work plan are to define goals for the
future use of the harbor and to identify additional research and
institutional arrangements that should occur to help environment-
al decision-makers understand how to achieve this goal. The work
group will also explore methods to assess the effects of pollu-
tants on aquatic life.
The tasks identified to achieve these objectives are as
follows:
1) prepare a statement of overall goals for the Harbor
which states the work group's consensus on the desirable
uses of the harbor. The statement will identify the
steps necessary for the harbor to progress from its
current state to these goals.
2) prepare a compendium of existing studies describing
the ecological status of the harbor and the potential
human health and ecological effects of current pollutant
levels. The studies will relate to the work group's
statement of goals which will be conducted in task #1.
This compendium will serve to identify areas for future
research. The work group intends to deliver this "s^tate
of the harbor" document to the Chesapeake Bay program,
which is planning an ongoing research project on Baltimore
Harbor.
3) coordinate environmental agencies in the identification
of management strategies for the Harbor. Several environ-
mental agencies have jurisdiction within the Baltimore
IX-7
-------
Harbor. These agencies are currently developing or have
developed studies and improvement projects to enhance the
Harbor ecosystem.
4) develop and apply a method (eco-scoring) to assess the eco-
logical effects on aquatic life of ambient concentrations
of metals and organic substances in the water column. The
results will be compared to the results of the indexing
exercise. The reasons for any inconsistencies in the
results will be explored.
Trihalomethanes in Drinking Water
The TAG, however, felt that no work plan or work group was
necessary for this issue, as the issue is already well understood.
The TAG, as a group, will compare the potential risks from tri-
halomethanes in drinking water to other potential human health
risks that may be associated with other environmental issues
under study during Phase II.
Under the requirement of the Safe Drinking Water Act to
establish National Interim Primary Drinking Water Standards, the
EPA set a maximum contaminant level (MCL) of 100 ug/liter for
total trihalomethanes (THMs) in drinking water. Trihalomethanes
are formed during chlorine disinfection of surface water supplies
used for potable drinking water. Chloroform is suspected of
being a human carcinogen. Consequently, any level of exposure
is, on the basis of EPA cancer policy, assumed to be associated
with some level of cancer riskeven those that conform to the
current trihalomethane MCL.
Levels of trihalomethanes in drinking water in the Baltimore
area currently meet this standard. Nevertheless, we estimate
that cancer risks in the Baltimore area'from drinking water con-
taining trihalomethanes to be roughly one in ten thousand (based
on current CAG estimates of carcinogenic potency for chloroform)
or three excess cancer cases. The TAG and the MC decided to
include this issue for study in Phase II because they felt that
a multimedia project like the Baltimore IEMP should consider
these "regulated" risks for comparison with other potential
chronic human health risks that may be identified in Phase II.
IX-8
« " l"nl IIP " IIIIIIIIIIHH V '111 ' W VVIIMV^MHMMmMMIW!!
-------
CHAPTER X
CONCLUSIONS
-------
X. CONCLUSIONS
The Baltimore IEMP contains three unusual features. First,
the framework in which decisions are made is different from other
lEMPs, let alone other EPA projects. In Baltimore, EPA delegat-
ed management authority to the local level. Second, the Balti-
more IEMP included examination of issues of ecological and
resource impact, in addition to health related issues. Third,
the scope of environmental problems considered in Baltimore was
wider than previous IEMP projects. We highlight four major
accomplishments of Phase I.
We have established the organizational framework at the State and
local levels for setting priorities for government action on envi-
ronmental issues.The ManagementCommitte(MC)withtheassis-
tance of the Technical Advisory Committee (TAG) Affectively iden-
tified and set priorities among a wide-ranging and diverse set of
environmental issues of concern to public officials. The Phase
II topics, which we described in Chapter IX, are the product of
this process.
The Baltimore IEMP has helped State and local governments develop
a working understanding of EPA methods for analyzing issues.
Priority-setting in the Baltimore IEMP was a hands-on process.
The TAG played an active role through its provision of expert
judgment while it used analytical tools to identify important
environmental issues and compare and rank them against evalua-
tive criteria. The success of representatives of State and local
political jurisdictions in reaching consensus on questions of
environmental priorities that unevenly affect them testifies to
the usefulness of these tools and these governments' ability to
use them. In addition, EPA held workshops for both government
officials in Maryland and the public to familiarize them with the
use of risk assessment.
The Baltimore IEMP has already helped State and local governments
identify and address a high-priority problem. In Phase I the
State and counties became aware of a potential health problem re-
sulting from the high levels of lead used in solder .and flux in
residential plumbing. They have already effected a lowering
(from 2% to 0.2%) of the standard for the use of lead in such
plumbing. This is direct evidence of a positive effect on envi-
ronmental management in Baltimore. However, most of the tangible
progress towards solving environmental problems is generally not
e'xpected until completion of the second phase of the project.
The Baltimore IEMP has provided information that will help EPA
conduct its programs. EPA has lacked analytical methods and pro-
cedures for setting priorities among issues that do not directly
relate to health. In Phase I of the Baltimore IEMP, we made pro-
gress in developing priority-setting tools for ecological issues
X-l
-------
relating to the aquatic environment and in a procedure for
achieving consensus on issues of importance to ground-water
resources.
CONCLUSIONS REGARDING THE INSTITUTIONAL ARRANGEMENTS
A key factor in the success of the process in Phase I is
EPA's delegation of management authority to State and local
governments. This sense of "local ownership" has resulted in a
greater commitment and more active involvement from State and
local officals than is typical for EPA-run projects.
The provision of a forum for discussion among governmental
agencies is also important. Indeed, we hope this organizational
framework will prove to be an enduring legacy of this phase of
the project. Its enduring nature is already attested to by the
continued enthusiastic involvement of project members after nearly
three years, the well-attended meetings of the TAG and workgroups,
and the active role that members have played in developing and
overseeing the execution of work plans.
The IEMP also provided a common language for communicating
about and understanding certain environmental problems. Risk
assessment, for example, became the "common denominator" for
discussion of a wide variety of health issues that concern many
agencies. It served to open channels of communication on diffe-
rent topics of common interest. Thus the structure of the IEMP
extended the lines of communication outside immediate profession-
al concerns and allowed for cross-fertilization of ideas.
Finally the institutional structure of the IEMP provided an
excellent forum for examining issues not clearly the responsibil-
ity of any one agency of-government. Indoor air, especially of
residences, is a case in point. By bringing together interested
government and private parties, all facets of the issue could be
objectively examined and diverse expertise from a wide variety of
disciplines brought to bear on the issue.
ANALYTIC FINDINGS
Human Health Risks
We have estimated several types of health risks. For expo-
sures that could increase the risk of cancer, we made conservative
estimates of: (1) possible incidence, or the number of excess cases
(beyond baseline occurrences) of the disease that may develop; (2)
average exposed individual (AEI) risk, or the probability that a
typical individual may develop the disease over a lifetime; (3)
maximum exposed individual (MEI) risk, or the probability that a
typical individual exposed to the highest predicted ambient con-
centration of pollutants may develop the disease over a lifetime.
For health effects other than cancer, we compared estimated
X-2
-------
exposure levels with an estimated reference dose, or no-effect
threshold, below which exposure is judged to be safe.
It is important to bear in mind the limitations and caveats
to this analysis (discussed in greater detail in Chapter V).
Among the most important are the following:
0 The Phase I analyses are based on limited data.
0 The principal purposes of the Phase I analyses are to
give a sense of the severity of a problem and to compare
problems for setting priorities for further study in
Phase II. They do not make definitive statements about
actual disease incidence in the local population.
0 Generally speaking, the assumptions used in the analyses
are conservative and tend to overestimate risks; (the
numbers presented are upper-bound estimates). However to
the extent the analyses suffer from omissions and uncer-
tainties, risks may be underestimated.
0 Finally, the results are not the product of a comprehen-
sive appraisal of all environmental risks in the Baltimore
study area.
Exposure
Of the ten organic compounds evaluated in the ambient air,
the highest measured concentrations were for benzene, xylene,
toluene, and ethyl benzene. For the metals studied, only chro-
mium appears to be in somewhat elevated concentrations in the
ambient air of Baltimore. Results from the Baltimore IEMP and
Maryland air toxics monitoring program indicated that there can
be significant variation in the concentrations of some organics
and metals in the Baltimore area. This is an important factor
to consider during Phase II when we will explore control options.
Based upon monitoring data from other cities, indoor air
pollutant levels in homes in Baltimore could be higher than ambi-.
ent air concentrations for the subset of pollutants evaluated in
the Baltimore IEMP. In some instances, indoor air concentrations
may be ten times higher.
The only class of volatile organic compounds in drinking
water to which people are significantly exposed is trihalome-
thanes, which result from disinfection of drinking water through
chlorination.
Lead can also be present in elevated concentrations in tao
water and dust from historical use of lead paint. Studies con-
ducted in other cities suggest that corrosive water can cause
X-3
II" "IMP W'"111PI»MI
-------
lead to leach out of residential plumbing that contains lead
solder or flux, leading to high concentrations in the first few
minutes of flow. Young children in older homes that have been
previously painted with lead based paint are often exposed and
ingest the dust containing lead. The result is a dramatic increase
in blood lead levels.
Cancer Risks
Of the ten organic compounds monitored in the ambient air in
Baltimore, we found that benzene clearly posed the greatest risk
(an upper-bound estimate of roughly 3 excess cancer cases per
year). The next three important organic compounds with regard to
cancer risk were perchloroethylene, chloroform, and carbon tetra-
chloride; together these three posed upper-bound risk of less
than one excess cancer case.
Of the metals examined in the ambient air, we found only
chromium to be at levels that could be of concern to human health.
Under the worst case assumption (that all the chromium is hexava-
lent and thus carcinogenic), the risk (an upper-bound estimate of
roughly 4 excess cancer cases per year) is roughly comparable to
that posed by the organic compounds studied.
In total, our analysis suggests an upper-bound estimate of
roughly 7 excess cancer cases per year from the organic compounds
and metals examined in the ambient air.
Even though trihalomethanes and other pollutants that the
IEMP was able to identify in the public drinking water were at
levels well below federal drinking water standards, these pollu-
tants may pose some cancer risk. This is because the lifetime
risk roughly 1 in 10,000 is shared by nearly the entire
population of the study area. The upper-bound estimate of 3 ex-
cess cancer cases per year is roughly equivalent to that posed
by the set of organic compounds that we examined in the ambient
air. The risk from trihalomethanes in drinking water would be
half that posed by the total ambient air riskr when metals in the
air are included and the chromium is assumed .to be hexavalent
an upper-bound estimate of 3 versus 7 annual excess cancer cases.
As previously mentioned in both the Executive Summary and
Chapter V, these numbers should be compared with the estimated
baseline incidence of cancer cases in the study area - 8227 oases
per year. [In 1983, the American Cancer Society estimated there
were 1.92 cases of cancer nationally for each cancer death. In
1984 cancer mortality in the Baltimore study area was 4285. Mul-
tiplying these factors produces the estimate of the baseline of
8227 annual cases] .
Preliminary risk calculations for indoor air using national
data suggest that indoor air pollution may be a relatively signi-
ficant potential source of risk in the Baltimore study area. The
X-4
IIP1"' "WHIIPW'III
-------
estimated risks were higher than those calculated for ambient
exposures in the other cities studied-.
Noncancer Risks
The preliminary Phase I findings suggest that most of the
Baltimore population will be exposed to chemicals at concentra-
tions high enough to cause concern over noncancer effects. Ben-
zene concentrations in the ambient air may increase the risk of
fetal developmental effects for the unborn children of pregnant
women in the study area. Ambient air concentrations of chloro-
form could also pose an elevated risk of fetal developmental
effects. Data from other cities on indoor concentrations of a
number of toxic substances, including benzene, chloroform, and
carbon tetrachloride, indicate that, also in the Baltimore area,
there may be some cause for concern over noncancer health effects
from exposure to these substances. We cannot ascertain the ex-
tent or the kinds of risks without further study.
Lead in the first few minutes of flow of tap water can con-
tribute to the Baltimore area residents" overall intake of lead;
it may put the unborn child and young children at greater risk of
adverse neurological effects and adult males at greater risk of
hypertension. Ingestion of lead in household dust is clearly a
very significant threat to the health of young children in older
homes which have been painted with lead-based paint.
Potential Threats to Ground-Water Resources
The evaluation of potential sources of pollution of ground-
water resources indicated that the most significant are: a source,
underground storage tanks; and a class of pollutants, metals.
Note that the analysis does not conclude they are problems; the
Phase I analysis focused on ranking the relative importance of
potential sources.
Underground storage tanks rated high on all factors of con-
cern. Of particular concern were the large number of sources and
their location above aquifers that are existing or future sources
of drinking water in Baltimore and Anne Arundel Counties.
Possible threats from metals include: chromium deposited as
fill in the harbor area; metals which leach from surface impound-
ments and landfills into ground water; and municipal landfills
when located in areas that are now or could be used for drinking
water.
Ecological Impact on the Harbor
The indexing exercise we conducted (using existing data
bases and results from a limited monitoring program) lead us to
conclude that mercury may be of more acute concern than the other
metals studied. Zinc, chromium, copper, and lead may also pose
problems for the Harbor.
X-5
-------
APPENDIX A
GENERAL METHODOLOGY
FOR AN
INTEGRATED ENVIRONMENTAL MANAGEMENT PROJECT
-------
APPENDIX A. GENERAL METHODOLOGY
This Appendix provides a general overview of the analytical
steps which EPA employs in Phase I of an Integrated Environmental
Management Project (IEMP) in the assessment of risks. It also
defines the concepts of risk assessment and risk management and
describes the analytical methods used for risk assessment. By
and large, the Baltimore IEMP followed these steps, with some
variations.
APPLICATION OF METHODOLOGY
This section presents a comprehensive description of EPA's
general methodology for conducting an IEMP. When we actually
apply this approach, inevitable practical limitations (such as
time and resource constraints, or limits on the state of know-
ledge) and the characteristics of the particular site force us
to tailor our efforts to our conditions. As a result, we do not
necessarily apply and develop the full framework and its analyti-
cal tools in any one IEMP. Moreover, we also believe that we
can achieve the most progress if we are flexible in applying the
approach and if we adapt to new circumstances as information
becomes available during the course of the analysis. Therefore,
the framework should not be considered to be a rigid blueprint;
its specific application can vary from one study to another,
although the general approach will remain unchanged.
There are significant limitations and uncertainties associ-
ated with the lEMP's methodology, which warrant consideration be-
fore examining the actual procedures. First, the risk estimates
are based primarily on existing knowledge about pollutant potency,
releases, and ambient conditions; these data, however, vary wide-
ly in quality and are almost always incomplete. Second, the ex-
posure estimates incorporate a series of simplifying assumptions;
although these assumptions are necessary, they they remain open
to question and may be controversial. Third, the potency esti-
mates are necessarily based on current knowledge of the toxicolo-
gical effects of various substances. Considerable controversy
exists about the degree of hazard posed by different pollutants,
and about whether some are hazardous at all. Finally, resource
and time constraints and the breadth of our focus prevent us from
analyzing individual issues in as much depth as might be possible.
We have attempted to strike a balance between the desire for
exhaustive and definitive analysis and the need for results at a
reasonable cost.
As noted at the beginning of this Appendix, the range of
potential environmental issues at any site is so large that it
A-l
-------
would be impossible, given limited resources, to conduct the in-
depth analysis that would be required to study all of the more
complex problems. Therefore, to manage an integrated study
effectively it is necessary to focus on a small set of study
topics. Although the screening process is designated as a single
step in Phase I, screening actually takes place continually
throughout the project as new information becomes available.
In its simplest form an IEMP consists of seven steps orga-
nized into two phases. An outline of the steps is provided below
and is followed by a more detailed discussion.
Phase I
1. Establishing arrangements between institutions cooper-
ating on the project
2. Defining the scope of the project:
--Setting geographic boundaries
Making an initial selection of pollutants and issues
for study
Establishing risk assessment approaches, e.g., selec-
.ting the health effects of concern
3. Collecting information on sources, pollutants, and
exposure pathways for entry into a computerized data-
base
4. Performing a screening analysis on the initial selec-
tion of pollutants and sources to determine which
of those should receive further attention in Phase II.
The screen involves two complementary approaches:
Evaluating risks to determine which pollutants,
sources, and exposure pathways are most significant
Qualitatively assessing analytical feasibility; rele-
vance to EPA, state, and local program objectives;
and potential for effective response
Phase II
5. Gathering additional data to confirm and refine the
risk assessments performed in Phase I, and to adjust
priorities accordingly
6. Analyzing and evaluating the cost-effectiveness of
alternative control options
7. Developing conclusions
A-2
-------
These steps and related issues are discussed in the following
sections.
Phase It Initial Issue Selection and Screening
Establishing Institutional Arrangements
The first step is to set up cooperative^ working relation-
ships with all responsible state, local, and federal groups with
direct responsibility for public health protection. Working re-
lationships with industry, environmental groups, and other inte-
rested parties are also highly advisable.
Defining the Scope of the Project (Scoping)
One of the first tasks is to reach an agreement with these
groups on such basic matters as the general scopei focus, and
objectives of the effort, the responsibilities of the various
parties, and the study's geographic boundaries. The key activity
of this task is to identify and select chemicals and issues for
consideration in Phase I. Decisions must be made as to which of
the many thousands of potential environmental contaminants are
likely to be present in the media of concern, and which of these
warrant the expenditure of time and resources to detect. This
is because our ambient monitoring equipment cannot adequately
identify all substances present; indeed* it is generally neces-
sary to specifically look for a pollutant in order to detect it.
There are many approaches for arriving at a list of sub-
stances which should be looked for, but they all face the same
constraints: the amount of resources available and the availa-
bility and sophistication of detection equipment. The approach
used should ideally allow a fair comparison of issues yet be
flexible enough to accommodate specific geographic constraints.
All the decisions involved in scoping are not necessarily
made at one point in time. In the process of the risk screen
(which we describe below), new issues or pollutants can be iden-
tified. The scope of the project can then be redefined to
include these new issues which are deemed to be significant
enough for inclusion in the project.
One of the more important actions made during this process
is the placing of issues which we have identified into broader
relational or reference categories.' Examples of such reference
categories, which we used in the Baltimore IEMP, are public
health .impact, ecological impact, and threats to groundwater
resources. By grouping issues into reference categories, we are
more likely to find a "common denominator" or unifying principle
to be used in evaluating the overall level of risk attributable
to each issuewithin that particular reference category. This
A-3
-------
allows like comparisons rather than having to confront the dilem-
ma of comparing different subjects for which there is no obvious
common denominator. This aspect of scoping as it applies to the
Baltimore IEMP is discussed in more detail in Chapter IV. [Some
issues can be placed into more than one reference category; toxic
contamination of fish, for example, can not only be an ecological
problem, but at the same time a public health threat.]
After issues have been grouped in reference categories,
project participants may decide to consider only one, or perhaps
just a subset, of these groupings of issues. Such was the case in
both Santa Clara and Philadelphia. At the outset of these pro-
ject participants decided to study only issues relating to human
health risks.
Developing a Database
The next step is to design a database that is appropriate to
the scope and objectives of the study. One approach is to gather
and inspect all readily available information on the issues
selected, i.e., sources, environmental releases, and exposed
populations for each medium. State, county, and city agencies,
permit writers, EPA, and local industrial facilities are the pri-
mary sources of these data, which should be collected and evalu-
ated before new data is generated.
^
After the available data have been reviewed, engineering
estimates, in many cases, can be used to fill the gaps and pro-
vide the needed information to compare issues. In almost all
cases, engineering estimates are needed to calculate intermedia
transfers. (Intermedia transfers are the relocation of pollu-
tants from one medium to another. For example, the incineration
of waste reduces -- but does not eliminate the volume of
solid material requiring landfilling. However, .the process may
also produce air-borne pollutants.) For certain common sources,
such as dry cleaners, degreasers, mobile sources (air), and non-
point run-off (water), EPA's Regulatory Integration Division in
the Office of Policy Analysis has developed algorithms to esti-
mate pollutant loadings to all relevant media that are easily
adapted to different geographic areas. In some situations, some
new monitoring may be initiated at this stage if it is deemed
necessary for setting priorities.
The Screening Process
The screening process selects and sets priorities for the
issues to be studied in Phase II. The process is emphatically
not one that can be performed in a mechanical way, but instead
relies on continuous evaluation of data, especially when compar-
ing potential risks.
A-4
-------
The screening process involves two complementary steps.
The first step quantifies risks to human health or the environ-
ment to identify the sources, pollutants, and exposure pathways
of greatest concern. This "risk screen" represents a preliminary
attempt at quantitative assessment, using available data, and
therefore may not generate results sufficiently complete to use
in setting priorities.
For that reason, we also use a second step which focuses on
issues that are not readily quantified. This approach qualita-
tively assesses an issue's analytical feasibility; its relevance
to EPA, state, and local program objectives; and its potential
for effective response (i.e., controllability).
Step 1; Risk Screen. The risk screen is a preliminary
quantitative assessment of the risks to human health or the envi-
ronment, performed to identify the sources, pollutants, and
exposure pathways of greatest concern. This assessment can be
made in a variety of ways, depending on the topic of interest.
For example, we can examine either the variety of sources that
contribute to pollution in one medium, or we can address- the
sources of pollution in several media, each of which contributes
to the same health effect.
It may take several cuts at the entire set of issues within
a reference category to develop the best common denominator. Once
this common denominator is developed, a subset of issues will
become particularly significant. For example, if the study par-
ticipants determine that carcinogenic effects should be the
"common currency" of evaluation, those issues or conditions that
might be expected to cause greater levels of cancer will become
more important than those that might cause other types of health
consequences.
Because the focus of the IEMP analysis is for policy pur-
poses and due to inevitable resource constraints, the IEMP pro-
ject manager must reserve the bulk of the study resources for
the later steps. For that reason, the risk estimates generated
for the screen are not precise estimates but rather very rough
approximations that are just suitable for setting priorities.
In fact, the information available to drive the risk screen is
not likely to be adequate even to identify the relative ranking
of issues with much certainty; instead, it serves only to group
the issues into three broad categories: (1) those that are likely
to pose high risk, (2) those with the potential for high risk
but little substantive evidence, and (3) those that are likely
to pose relatively low risk. Discretion must be used in deter-
mining when enough data have been gathered to support a success-
ful risk screen.
A-5
mi inn1 "'Pinup
-------
Step 2: Qualitative Evaluation. After the issues identi-
fied in Phase I as potential candidates for Phase II analyses are
separated into the three groups based on quantitative measures of
risk, the issues classified in the first two groups (high risks)
are further evaluated using nonrisk or secondary criteria. These
secondary criteria, applied to the first two groups of high risk
issues, provide broader perspectives for setting priorities than
those based solely on risk assessment criteria. As more informa-
tion becomes available, further reclassification of the issues
may be warranted. The secondary criteria are discussed below.
Like the rest of the priority-setting process, they should not
be regarded as being inflexible. Other groups conducting inte-
grated studies may want to modify or add to these criteria. For
example, the Baltimore IEMP added an explicit criterion of avoid-
ing duplication of existing analyses of control programs.
Analytical Feasibility
The primary elements determining analytical feasibility are
the amount of supporting data available, the availability of
analytical methods, and the level of effort required to generate
new data. These are basically program management considerations
in that they indicate how much effort would be needed for basic
data gathering before developing management alternatives (e.g.,
for control). There is an important tradeoff implicit in the
criterion of analytical feasibility: the greatest payoff in
terms of identifying and controlling toxics may be in studying
issues that have not been studied in the past or have been avoid-
ed because of the complexity or magnitude of the issue. On the
other hand, within the time and budget available for the project,
it may not be feasible to characterize these issues adequately,
and it may be necessary to curtail the analysis short of develop-
ing complete risk management strategies.
Relevance to EPA, State, and Local Program Objectives
f The elements of this criterion involve the significance of
the issue for national EPA programs and state and local interest
in the issue. One of our objectives is to use the IEMP projects
to indicate where shifts in EPA priorities and methods are appro-
priate, so the relationship of the issue to national EPA programs
is an important factor. State and local interest affects the
extent of support that local and state participants will give to
the detailed study effort, and also strongly affects the feasi-
bility of implementing control strategies, when warranted.
Potential for Effective Response
The level of existing control provides a rough measure of
the likelihood that additional controls are cost-effective, since
cost-effectiveness generally declines with increasing control.
A-6
-------
The feasibility rpf' additional controls -is also important^ if no
technological, institutional,: and political means are. available
to reduce risk, the issue ,is probably^ not ,a useful candsida,te for
study: ;~ _ ;, \ ' -\ ",. ,.....,^ '.,.., v
Risk Management Challenges in?'Interpreting Results and
Applying the Process. ' T
In the process of establishing priorities, study partici-
pants are likely to encounter several' basic problems involving
the interpretation and application of environmental risk informa-
tion. Some, specific issues are discussed below. ;
Maximum Exposed Individual Risks Versus Total Population
Risks ₯
The management or project advisory group selecting study
topics will have to confront one of the classic trade offs in
risk management--that of maximum exposed individual versus cumu-
lative population risks. Inevitably* some of the potential
study topics will involve situations where a relatively small
number of people experience risks higher than^those faced by the
aggregate population; these topics do not necessarily coincide
with those where there are widespread risks. Decisions may have
to be"made whether"to spend study resources to help a few people
by a substantial margin or to spend them toJhelp many people by a.
modest margin. This issue is one of the most difficult issues in
setting priorities.
Comparing .Across Effects "
Anotherxdiff icult issue involves the* comparison of risks
across different health effects (e.g., kidney damage ,versus can-
cer). Typically, this is /done on ting basis of j^nformed judgment.
RID is developing an approach b seating priorities that attempts
to accounti;₯for the severity of different effects. There have
been several other attempts' to scale different health effects
(for instance, EPA\s Off ice,; of Solid Waste has developed such a
scale), but this issue may be most effectively resolved by con-
sidering local concern for the severity ^p^f *dif ferent effects.
.. . ' P- y v- '* ' ... '' ""''"' " '' ''" *-"" ~" ': '''..
Comparing Human Health Effects :jtq EnvjLrpnmental : Effects
M .. ""'-'. ' '}£''.' '-;
In cases where the scope of;i the study, as in theCBaltimore
IEMP, includes human health as well as ecological issues, diffi-
cult value judgments must be made. ^ These judgmerits are not
generally/made within the same policy decision analysis because
the quantitative means -*gf! equating, -thev two^ veiry _4if:f eient.., public,.
policy objectives a^e 'lackying. It5 is (ile^arly important/ita protect
both public health and pur ecosystems,* bu/|r tKe value.,,placed oh
A-7
-------
human health* is not the same as that placed on our other environ-
mental ^goals'. Decisions cpncerning the s'etting of priorities
between "these two types of issues (because the quantitative
measures used to assess adverse impacts for each are not scienti-
fically or analytically comparable) involve the difficult and
necessarily subjective Balancing of both public policy objec-
tiyes.
Phase I Product ;
This Report contains information about theHrisks we quanti-
fied, identifies priority issues based on our risk assessments,
and describes the methods we used to assess exposure and risks.
Furthermore, it describes the process by which these products
were made possible. in this. regard, we elaborate on the institu-
tional arrangements necessary for the making of key policy deci-
sions and management judgements and the manner in which these
decisions were made.
Phase II; Risk Management Control Options
After Xfeers elect ing Phase I issues warranting further analy-
sis, we begin Phase II. Phase II consists of three major tasks:
additional data gathering to improve our exposure and , risk
assessments, a scientific review of our risk assessments, and
analysis of control options.
Additional Data Gathering
*
By definition, in. Phase I we identify the study 'topics for
further analysis using ~ existing and special engineering esti-
mates. While this is reasonable for setting priorities^ the
quality of the data becomes .even more important at the later
stages of the study. Depending on the specific issue, in Phase
II we generally engage in further data gathering in 'order to:.
(1) confirm whether an issue warrants further investment of time
and resources, (2) ensure that data used in later steps are as
accurate ^as possible* and (.3) revise Phase I priorities where
appropriate. ' ^' .
* &:'.''' - ... > " ".,,.'
* " ' \, '( : - «.
In some cases, the objective of additional data collection
may be to improve^ our understanding of the significance of parti-
cular pollutants and sources. This situation may only "call for
limited monitoring.
- '* '" *3
In other cases, which are really at the core of the Phase II
activities," the objective will be to analyze the costs of poten-
tial control alternatives, where effectiveness is measured by the
reduction in health risk^-eith'er Most Expo sect Individual (MEl)
(defined ori'p.^'A.1'5) risk or 'total population incidence. In these
A-8
-------
situations, we will need to ensure that our estimates of pollu-
tant loadings and ambient concentrations, for example, are rea-
sonable and representative of avera'ge annual conditions. These
efforts may include consulting local environmental agencies and
plant managers to confirm underlying data on ambient releases,
as well as initiating an extensive monitoring program.
As additional information is gathered in Phase II, it is
helpful to periodically re-evaluate the Phase II issues in terms
of potential risk. This can often be done by conducting a sensi-
tivity analysis using the revised data. New information may in-
dicate that the risks are not likely to be as great as once
thought, or they may be greater than originally anticipated.
Sensitivity analyses of new information provide a way to further
focus the use of project resources during Phase II.
Developing Pollution Control Options
The next step in Phase II is to assess alternative control
strategies for the subset of environmental issues identified for
this work. A very significant objective is to provide analysis
of how to reduce risks to health (total population as well as to
the maximum exposed individual) or the environment (the measure
of effectiveness) at the minimum cost. This occurs through the
lEMP's cost-effectiveness analysis, which presents, the tradeoffs
of costs and risk reductions that decision makers usually take
into consideration when formulating regulatory strategies. The
analysis considers risks that, .result from exposure to primary as
well as secondary (intermedia) releases. It is our intention
that community decision-makers use it to shape their general
strategies for providing additional environmental protection in
the area.
There are three major components to an analysis of control
options: quantitative measures of ambient concentrations at the
point of exposure; estimates of exposure and risk; and the costs
and efficiencies of feasible control options. Data collected in
Phase I and supplemented in Phase II provide the necessary infor-
mation on pollutant loadings estimates. We generally employ EPA
fate and .transport models to develop ambient concentrations at
the point of exposure. Exposure and risk calculations are then
made using standard EPA assumptions, which we discuss later
in this chapter.
We identify feasible control options and estimate their
associated costs and efficiencies by employing engineering
assessments and EPA technical documents developed to support
various regulatory activities.
The application of this process will inevitably vary from
one project to the next. In Philadelphia, we implemented the
A-9
-------
basics of this methodology. However, because Philadelphia was
our first project, we did not have experience to guide us in
assembling the steps in an efficient manner. Largely because of
lack of sophistication with the methodology, we chose not to
examine other issues which would involve ecological and non-
cancer health effects.
In the Baltimore IEMP, we have expanded the scope of the
project to include ecological impacts and have decided to examine
non-cancer health effects in Phase II. We will again be treading
new ground in developing analytic methodologies to examine eco-
logical impacts and the risk of non-cancer effects.
RISK ASSESSMENT AND RISK MANAGEMENT
The two key organizing concepts of an integrated environmen-
tal management project are risk assessment and risk management.
Risk assessment is the central task of Phase I; risk management
is the central task of Phase II. Conceptually, risk assessment
should be done independently of risk management.! In IEMP
studies, we separate assessment from management as much as
possible; however, this is not always possible so that some of
the activities in Phase I and Phase II overlap.
We emphasize that our discussion of the IEMP methodology is
presented at a conceptual level; bur actual application of the
methodology may vary from one geographic project to the next.
Furthermore, we may use different methods in response to the
needs of a particular IEMP. For example, project participants
in the Baltimore IEMP had identified issues pertaining to local
ecology and groundwater resources, in addition to health issues,
for study in Phase I. To assess the significance of these issues
and to decide which of them warranted further examination in
Phase II, we needed different analytical tools from those we used
in setting priorities among human health issues in Philadelphia
and the Santa Clara Valley. (Note that Philadelphia and Santa
Clara Valley are areas where we conducted earlier geographic
studies.) We describe these other- (nonhealth related) methods in
Chapters VI (Analysis of sources with Potential Adverse Impacts on
Ground-water) and VII (Analysis of Ecological Impact) and discuss
only our methods for analyzing risks to human health in this
Appendix.
Phase I Risk Assessment and Priority-Setting
There are always more environmental issues to study than
resources with which to address them in a manner that will lead
to their successful resolution. A key task of Phase I of an IEMP
is, therefore, this: to choose which issues demand the immediate
A-10
-------
*. i -: r . *' . ;-
attention of environmental^.--- official's-?rarid- the expenditure of
their time and resources--and which can- : be addressed at a later
date. Suchrf comparingf of issues 'and s'e.tting of priorities re-
quires thatvvall issues be reduced to -a 'common element. (Note
that an integrated approach "to all iss,ues,;-means ^we try to examine
the same issue across media. An IEMPlo.s not ,,a comprehensive look
at all the environmental issues of the study^p^ea.) Riski either
to human health or the^environment (pr ^othfe^iis the common ele-
ment we use for comparing environmental^ issues in geographic
studies. ffcRisk can be Categorized in several different ways. As
appropriate, "we can examine .risk ass,pcia*tedi;w,ith exposurjgs^ from
a specific pollutant source. We can examine risks associated
with exposure to pollutants, irrespective of source. We can
look at risk associated with exposures in a s;ingle environmental
medium, such as air, water, or food. We can examine r^isks asso-
ciated with exposure ^'pathways. In the absence" of other impor-
tant considerations,"":- the mbre significant the risk, the more
likely the issue will rank high iir priority for examination in
Phase II, the risk management phase. Important considerations
that may alter the priority ranking include, for example,
resource limitations or the fact that a particular ^IBsue is
already being handled by the Federal, state of local government.
" ' ' : ' -fti'f. '*
;'-. i- .,.,. ' ,'.v- ^ , V^.
The risks we examine are associated with exposure to toxic
chemicals in the environment and represent estimates of the poten-
tial for these exposures to cause adverse effects. Chemical's can
adversely affect human health, ands>thus pose a risk,, when people
are exposed to toxic levels in the air, water, food^ or contami-
nated soil. Chemical exposures can adversely affect the environ-
ment by disrupting the ecology of surface waters or the dynamics
of biological communitie,s. Furthermore, toxic chemicals can so
contaminate ,a natural resource that future generations can, for
all practical purposes, no longer^ use or enjoy it or must pay
greatly to do so. ^, ''-^ .
In order to examine",potential impacts to human health from
environmental exposures in a particular geographical location,
we perform a risk assessment. , Based on. the National Research
Council's^ recommendations", ^ an IEMP risk assessment contains the
following elements: ' -.. . ^. 'f
o Hazard Identification
Does the agent cause the adverse effect?
T For the chemicals '"of; interest, RID toxicologists and con-
tractors prepare individual Profile/jreports that summarize
fc the information on animal and human health effects. Each
report examines bot^ oral and, inhalation .exposure datd for
\ the following ten health effect categories: .qancer, liver,
kidney, reproductive, fetal developmental, neurobehavioral,
mutagenicity, respiratory", cardiovascular, and other. A
A-11
-------
weigh,t-of-evidence evaluation is provided for each health
effect category. For cancer, we use the EPA weight-of-
evidence evaluations that are prepared by the Agency's
Cancer Assessment Group (GAG) according to the 1986 cancer
risk assessment guidelines.3 EPA does not have weight-of-
evidence evaluation schemes for the other health effect
categories, so we use ranking schemes developed specifically
for our work. Our ranking schemes are currently being
reviewed by scientists within and outside EPA.
o Dose-Response Assessment
(What is the relationship between dose and incidence of
effect in humans?)
This information is also provided in the chemical profile
reports. With the exception of mutagenicity, a separate
evaluation is provided for each of the health effect cate-
gories. For cancer, we use the dose-response evaluations,
or "unit risk factors" that are prepared by EPA's Cancer
Assessment Group (CAG). For the remaining non-cancer
health effect categories, EPA does not have a satisfac-
tory method for quantifying dose-response relationships.
Therefore, as an alternative we use the EPA Reference
Doses (RfDs) as benchmarks to estimate exposure levels
that are "safe" and for which we assume the incidence of
effects is insignificant.*
o Exposure Assessment
(What exposuresare currently experienced or anticipated
under different conditions?)
The IEMP studies include an assessment of environmental
exposure levels for selected chemicals in the media of
interest, usually air and water. Environmental exposures
are estimated primarily by modelling. As feasible, field
monitoring is conducted and the collected data are used
to val.idate the modelling. In addition estimates of
exposed populations are made by examining census data.
0 Risk Characterization
(What is the estimated incidence of the adverse effect in
a given population?)
For the environmental media of interest, we use our expo-
sure assessment to estimate both the concentrations of
*In so doing, IEMP follows EPA practice. However, other
authorities may use other methods. For example, for
airborne contaminants. Threshold Limit Values (TLVs)
may be more suitable under certain circumstances.
A-12
-------
specific chemicals and the number of people exposed at
those concentrations. We then link this information to
our chemical-specific dose-response assessments in order
to make rough estimates of the incidence of disease that
may occur above the baseline rate in the study area.
Note that because of limitations in our ability to
quantify dose-response relationships for each of the
eight health effect categories, we must analyze cancer
differently from the non-cancer health effects. For can-
cer, we treat the CAG unit risk factor as the slope of
the dose-response curve; therefore, we can use this value
to estimate a crude incidence. For the non-cancer health
effect categories, we do not have estimates of the slopes
of the dose-response curves. Instead, we have the RfDs,
which we . use as estimates of exposure levels that are
"safe", in terms of non-cancer health risks. So for non-
cancer effects, we do not calculate the incidence of
effects. Rather, we calculate the number of people ex-
posed to levels above the RfDs in order to estimate how
many may be at some increased risk.
The risk characterization also includes a discussion of
the uncertainties in the hazard identification, dose-
response assessment, and exposure assessment. For the
hazard identification, this discussion focuses on the
weight-of-evidence evaluations. For the dose-response and
exposure assessments, the discussion focuses on uncertain-
ties in the databases and models used.
During Phase I of our first geographic project in Philadelphia
and then again in Baltimore, we also relied on priority-setting
methods other than quantitative risk assessment; e.g., the use of
expert judgement to establish relative ranking of issues.
In order to perform an IEMP risk assessment, as outlined
above, we rely heavily on approved EPA quantitative methods, CAG
unit risk values, RfDs and other appropriate standards and crite-
ria. In addition, we often use other approaches and criteria
to refine our priority ranking of issues. These criteria include
analytical feasibility, relevance to EPA, state, and local pro-
gram objectives, and the potential for effective response. These
criteria (which incorporate risk management perspectives) are
discussed in Chapter VIII of this report.
Bear in mind that the objectives of the IEMP in the use of
risk assessment procedures are to allow for relative comparisons
of potential problems and to provide a general sense of the re-
lative significance of an issue, rather than to make definitive
A-13
-------
statements about the absolute risks posed by a particular sub-
stance, source, or pathway. It must be emphasized that we conduct
risk assessments; we do not conduct epidemiological studies*
Thus, we do not, and cannot examine the disease rates in the
local population in order, to correlate health effects with current
or past environmental exposures. Moreover, the exposure assess-
ments conducted as part of an IEMP are not designed to be used in
nor are they appropriate for an epidemiological study.
" *
Although different in their objectives, design and interpre-
tation, risk assessments and epidemiological studies are actually
complementary. Risk assessment can help to identify populations
and geographic areas that appear to be at risk and therefore
might be appropriate subjects for an in-depth epidemiological
study. Epidemiological studies increase the scientific under-
standing of the relationship between exposure and health effects,
thereby strengthening the basis pf risk assessment. Epidemiolo-
gical studies may also, in some cases, be useful for confirming
specific hypotheses suggested by risk assessment.
Other qualifications regarding the use of quantitative and
other health risk assessment methods are discussed later in this
appendix.
Phase II Risk Management
In Phase II, we will focus the resources of the project on
considerations of how best to manage risk in selected areas.
Risk management is the process of evaluating and selecting
approaches to reduce the risks identified through risk assess-
ment. Risk management considers not only the level of risk
posed by a particular pollutant or source of.pollution but also
factors such as the feasibility and cost of control, public pre-
ferences, and institutional capabilities. Setting priorities
for research or government action is also a key 'aspect of risk
management. The objectives of this Phase are to investigate
control alternatives, estimate their efficiency in controlling
pollution, develop estimates of their cost, and then assess their
cost-effectiveness; in the Baltimore IEMP it also includes devel-
opment of analytical tools (i.e., a model for helping local govern-
ments establish priorities for addressing potential threats to
ground-water resources from underground storage tanks) as' well
as a research plan (the Harbor blueprint). Where necessary, we
supplement the analysis by developing detailed exposure estimates
and other specific data required to refine the quantification of
risks assessed in Phase I.
The criterion of cost-effectiveness (i.e., cost per change in
some unit of effectiveness) is generally expressed as the amount
of human health risk reduced by alternative controls in relation
A-14
-------
to the cost of implementing those .controls. However, in some
cases, such comparisons cannot be made and other economic crite-
ria or definitions of cost-effectiveness must be used (e.g., the
costs of preventing damage to a resource).
Even where mechanisms are already in place for controlling
chemical exposures, an analysis of the cost-effectiveness of
these control strategies can be useful to local communities. An
analysis of the cost-effectiveness of control options may lead
to the identification of control strategies that attain the same
level of control for less cost.
GENERATION OF RISK ESTIMATES
The following overview of methods for quantifying health
risks theoretically applies to both cancer and non-cancer health
effects. In practice, appropriate quantitative methods have
only been developed for estimating carcinogenic risks. There-
fore, unless noted otherwise, the following discussion deals
only with estimation of carcinogenic risk to an exposed popula-
tion.
In IEMP studies, we calculate risk using three measures of
exposure: risk to the most exposed individual (MEI), risk to
the average exposed individual (AEI), and the excess aggregate
population incidence. We define risk to the MEI as the increased
probability that an individual chronically exposed to the highest
concentration of one or more chemicals will have exposure-related
cancer during the course of his or her lifetime. For the MEI,
the exposure is calculated as either the highest modelled concen-
tration or the average concentration obtained for the monitoring
site with the highest daily or annual average monitored value for
one or more chemicals* We define risk to the AEI as the increas-
ed probability that an individual exposed to the area-wide aver-
age concentration of .one or more chemicals will have exposure-
related cancer during the course of his or her lifetime. Aggre-
gate population risk is the estimate of the increased incidence
of cancer, above the background rate, in an exposed population.
We use the standard EPA assumption that exposure is for 70 years.
Thus, elsewhere when we refer to "lifetime risk" or "70-year life-
time", this means 70 years of exposure. A more detailed discussion
of our assumptions and approach for calculating these risks is
provided later in this Appendix.
For a given population, the IEMP risk assessment involves
linking estimates of chemical potency, derived from animal and
human health effects data, with estimates or measurements of
area-wide contamination or exposure levels. A relatively high
degree of uncertainty unavoidably underlies this approach. In
A-15
-------
most cases, the risk estimates cannot easily be verified or dis-
proven by observation. The approach is useful, however, because
it is straightforward and allows application of the most current
scientific information to the analysis of adverse health effects
from current and projected exposures.
The two key elements in estimating health risk are: 1) the
calculation of the potency of a chemical, and 2) the determina-
tion of the level of exposure to that chemical.
1. Estimating Potency
A. Carcinogens
Potency is estimated from an analysis of relevant animal and
human (i.e., occupational, epidemiological) data concerning the
toxicity of a chemical. At EPA, data are obtained from the open
literature and from proprietary studies submitted to the Agency.
Chemical potency is determined by: 1) evaluating qualitatively
the preponderance (or weight) of evidence that a chemical causes
the adverse health effect in question; and, if so, (2) estimating
quantitatively the relationship between the dose (that is, the
amount of chemical an animal or person is to exposed over a
period of time) and the incidence of the effect within an exposed
population of a given size. The latter "dose-response relation-
ship" describes the potency of the chemical. That is, the lower
the threshold and the steeper the slope of the dose-response
curve, the more potent the chemical. For assessing carcinogenic
risks, potency estimates can be used to relate the levels of ex-
posure to the probability that individual(s) will have cancer.
The EPA Cancer Assessment Group (CAG) calculates chemical-
specific cancer potency values called "unit risk" estimates. We
use these unit risk estimates in our IEMP studies and many Pro-
gram Offices throughout EPA also use these values- for a variety
of risk assessment and risk management activities.
o Qualitative evaluation
The qualitative, or weight-of-evidence evaluation involves
reviewing the scientific literature, both human and animal,
to establish the sufficiency of evidence for a chemical's
carcinogenic potential in humans. This evaluation of the
weight-of-evidence for a causal association between the
chemical exposure and cancer is an important part of risk
assessment. It is necessary because the extent and quality
of chemical-specific databases rarely allow for an unambi-
guous determination of carcinogenic potential.
A-16
-------
Two evaluation schemes are available .for assessing cancer
weight-of-evidence. One was developed by the International
Agency for Research on Cancer (IARC) and the other was developed
by EPA. The EPA scheme is a modification of the IARC system.
All chemicals evaluated for carcinogenicity by .EPA1s GAG are now
classified according to the EPA scheme.
Briefly, the IARC scheme has the following three'categories: ^
Group I; Known Human Carcinogen. This category is used
only when there is sufficient evidence from epidemiological
studies to support a causal association between the exposure and
cancer.
Group II; Probable Human Carcinogen. This category
includes exposures for which, at one extreme, the evidence of
human carcinogenicity is almost "sufficient" and at the other
extreme, it is "inadequate". To reflect this range, the category
is divided into Group IIA and Group I IB to indicate higher and
lower degrees of evidence, respectively.
Group III; This category includes exposures that can not be
classified as to their carcinogenicity in humans.
The EPA scheme has,the following five categories: ^
Group A; Human Carcinogen. This category is used only when
there is sufficient evidence from epidemiological studies to
support a causal association between exposure to the agent and
cancer in humans.
Group B; Probable Human Carcinogen. This category includes
agents for which the evidence based on epidemiological studies
is "limited" and agents for which the evidence based on animal
studies is "sufficient". To reflect this range, the category is
divided into Group Bl and Group B2, indicating higher and lower
degrees of evidence, respectively. Group Bl has agents with
"limited" evidence from epidemiological studies. Group B2 has
agents with "sufficient" evidence from animal studies and "inade-
quate" evidence or "no data" from epidemiological studies.
Group C; Possible Human Carcinogen. This category is used
for agents with "limited" evidence of carcinogenicity in animals
and an absence of human data.
Group D; Not Classifiable as to Human Carcinogenicity. This
category is used for agents with "inadequate" human and animal
evidence of carcinogenicity or for which no data are available.
A-17
-------
Group E; Evidence of Non-Carcinoqenicity for Humans.
This category is used for agents that show no evidence of
carcinogenicity in at least two adequate animal tests in different
species or in both adequate epidemiological and animal studies.
This designation is based on the available evidence and should
not be interpreted as a definitive conclusion that the agent will
not be a carcinogen under any circumstances.
Presently, there is no similar scheme for assessing the weight-
of-evidence for agents causing non-cancer health effects.3
0 Quantitative evaluation
In general, the EPA Carcinogen Assessment Group (CAG)
estimates the cancer potency of agents that are classified
Group B and higher. Potency is expressed as the "unit
risk". This unit risk estimate for an air or water pollu-
tant is defined as "the incremental lifetime cancer risk
occurring in a hypothetical population in which all indi-
viduals are exposed continously from birth throughout
their lifetimes to a concentration of 1 ug/rn-^ of the agent
in the air they breathe, or to 1 ug/L in the water they
drink. This calculation is done to estimate in quantita-
tive terms the impact of the agent as a carcinogen. Unit
risk estimates are used for two purposes: (1) to compare
the carcinogenic potency of several agents with each other,
and (2) to give a crude indication of the population risk
that might" be associated with air or water exposure to
these agents, if the actual exposures are known."4
0 Key simplifying assumptions of the quantitative evaluation
In most cases, CAG uses a so-called linearized multistage
model to calculate chemical-specific cancer potency or
unit risk estimates. A mathematical model is needed in
order to calculate the unit risk factor because animal and
human carcinogenicity data are always associated with expo-
sure levels that are much higher than those normally found
in environmental seittinqs. Therefore, in order to esti-
mate potential cancer risk for the low exposure.levels of
concern, it is necessary to extrapolate from data associ-
ated with high exposures. This is done through the use
of appropriate models.
In explaining their choice of model, CAG notes the follow-
ing important considerations:
"The unit risk estimate represents an "extrapolation
below the dose range of experimental data. There is cur-
rently no solid scientific basis for any mathematical
A-18
-------
extrapolation model that relates exposure to cancer risk
at the extremely low concentrations,... that must be dealt
with in evaluating environmental hazards. For practical
reasons the correspondingly low levels of risk cannot be
measured directly either by animal experiments or by epi-
demiological studies. Low-dose extrapolation must, there-
fore, be based on current understanding of the mechanisms
of carcinogenesis. At the present time the dominant view
of the carcinogenic process involves the concept that most
cancer-causing agents also cause irreversible damage to
DNA. This position is based in part on the fact that a
very large proportion of agents that cause cancer are
also mutagenic. There is reason to expect that the quan-
tal response that is characteristic of mutagenesis is
associated with a linear (at low doses) nonthreshold
dose-response relationship. Indeed, there is substantial
evidence from mutagenicity studies with both ionizing
radiation and a wide variety of chemicals that this type
of dose-response model is the appropriate one to use. This
is particularly true at the lower end of the dose-response
curve; at high doses, there can be an upward curvature,
probably reflecting the effects of multistage processes
on the mutagenic response. The low-dose linear nonthres-
hold dose-response relationship is also consistent with
the relatively few epidemiologic studies of cancer respon-
ses to specific agents that contain enough information to
make the evaluation possible (e.g., radiation-induced leu-
kemia, breast and thyroid cancer, skin cancer induced by
arsenic in drinking water, liver cancer induced by afla-
toxins in the diet). Some supporting evidence also exists
from animal experiments (e.g., the initiation stage of the
two-stage carcinogenesis model in rat liver and mouse skin).
Because its scientific basis, although limited, is the best
of any of the current mathematical models, the nonthreshold model,
which is linear at low doses, has been adopted as the primary "~" ^
basis for risk extrapolation to low levels of the dose-response
relationship. The risk estimates made with such a model should
be regarded as conservative, representing a plausible upper limit
for the risk; i.e., the true risk is not likely to be higher than
the estimate, but it could be, and probably is, lower.
-'
For several reasons, the risk estimate based on animal bioas-
says is only an approximate indication of the absolute risk in
populations exposed to known c'arcinogen concentrations. First,
there are important species differences in uptake, metabolism,
and organ distribution of carcinogens, as well as species diffe-
rences in target site susceptibility, immunological response,
hormone function, dietary factors, and disease. Second, the
concept of equivalent doses for humans compared to animals on a
A-19
-------
ing/surface area basis is virtually without experimental verifica-
tion as regards carcinogenic response. Finally, human popula-
tions are variable with respect to genetic constitution and
diet, living environment, activity patterns, and other cultural
factors.
The risk estimate can give a rough indication of the rela-
tive potency of a given agent compared with other carcinogens.
Such estimates are, of course, more reliable when the comparisons
are based on studies in which the test species, strain, sex,
and routes of exposure are similar.
The mathematical formulation chosen to describe the linear
(at low dose) nonthreshold dose-response relationship is the
linearized multistage model. This model employs enough arbitrary
constants to be able to fit almost any monotonically increasing
dose-response data, and it incorporates a procedure for estima-
ting the largest possible linear slope (in the 95% confidence
limit sense) at low extrapolated doses that is consistent with
the data at all dose levels of the experiment." 4
The EPA Carcinogen Risk Assessment Guidelines indicate the
following concerning the choice of extrapolation model:
"The Agency will review each assessment as to the evidence
on carcinogenesis mechanisms and other biological or statistical
extrapolation model. ...A rationale will be included to justify
the use of the chosen model. In the absence of adequate informa-
tion to the contrary, the linearized multistage procedure leads
to a plausible upper limit to the risk that is consistent with
some proposed mechanisms of carcinogenesis. Such an estimate,
however, does not necessarily give a realistic prediction of the
risk. The true value of the risk is unknown, and may be as low
as zero."3
B. Mutagens
In the IEMP studies, we restrict our evaluation of mutagen-
icity to a qualitative assessment. We simply indicate if the
chemical is considered to be a mutagen or if there is insuffi-
cient data to make such a determination. Most of the experimen-
tal dates addressing mutagenicity involve in vitro testing of
mammalian cell cultures and non-mammalian organisms such as
bacteria. At present, we do not have methodologiess for using
these kinds of data to calculate dose-response curves amd quanti-
fy risk to humans. EPA assumes that a dose-response relation-
ship for mutagenicity, like carcinogenicity, would have no thres-
hold; therefore, any exposure level would be associated with
some degree of risk. Evidence of mutagenicity often, but not
always, suggests carcinogenic potential.
A-20
-------
C. Non-Carcinogens
Chemicals that give rise to toxic endpoints other than cancer
and gene mutations are often referred to by EPA as "systemic
toxicants" because they affect the functioning of various organ
system, such as liver, kidney, cardiovascular, and respiratory
systems. Generally, based on our understanding of physiological
mechanisms, systemic toxicants are treated as if there is an
identifiable exposure threshold (both for the individual and for
the population) below which adverse effects are not observable.
From the viewpoint of risk assessment and risk management, this
characteristic distinguishes systemic toxicity from carcinogeni-
city and mutagenicity, since the latter two are usually treated
as non-threshold processes.
Because of this concept of identifiable thresholds, systemic
effects traditionally have been evaluated by EPA through the cal-
culation of "safe" exposure levels. This is unlike EPA's ap-
proach for evaluating carcinogens, for which all exposures are
assumed to involve some measure of risk. For many years, the
concepts of "acceptable daily intake" (ADI) .and "margin of safe-
ty" have been at the heart of EPA's approach to risk assessment
for non-cancer health effects. Although there are limits to
some of these approaches, EPA is often called upon to apply these
concepts when making and explaining decisions concerning the
significance to human health of certain chemical exposures in the
environment. Thus, the threshold concept for non-cancer effects
is extremely important in the regulatory and risk management con-
text.
More recently, the Agency has tried to come to grips with
some of the technical and philosophical issues inherent in defin-
ing a "safe" exposure level for systemic toxicants. As an out-?
growth of this effort, the concept of the "reference dose" (RfD)
has been recommended to replace that of the ADI. The RfD is
calculated in the same manner as an ADI; however, it is viewed
as a .benchmark level without the value-laden connotations of
absolute safety or acceptability versus absolute unacceptability.
Thus, exposures that are less than the RfD are not likely to be
associated with non-cancer health effects and are therefore less
likely to be of regulatory concern. Conversely, as the frequency
of exposures exceeding the RfD increases and as the size of the
excess increases, the probability also increases that adverse
effects may be observed in a human population. Nonetheless, a
clear conclusion cannot be categorically drawn that all doses
below the RfD are "acceptable" and that all doses in excess of
the RfD are "unacceptable."5
To date, the Agency's calculation and use of ADIs and RfDs
has been applied only to oral exposure routes, such as drinking
A-21
-------
water and food. EPA has an established methodology for calculat-
ing oral RfDs and there are values for approximately 400 chemi-
cals. In contrast, the Agency does not have a methodology for
calculating similar benchmark levels for inhalation exposures.
EPA has not yet established a procedure for estimating such
benchmark values for chronic inhalation exposures; a methodology
is under development by a special committee in FY'87 and the
values will be called inhalation RfDs. Thus, with the exception
of the six or seven heavily studied criteria air pollutants
(ozone, carbon monoxide, sulfur oxides, etc.) there are no bench-
mark values for chronic inhalation exposures that are appropriate
for use in risk assessment. Because the Agency's RfDs are calcu-
lated for lifetime oral exposures only, and IEMP studies examine
both oral and inhalation exposure routes, it has been necessary
for us to estimate IEMP benchmark levels for inhalation exposures.
IEMP studies rely on EPA oral RfDs and on other "threshold"
values computed by our toxicologists and consultants. Our "thres-
holds" for various non-cancer effects have been calculated employ-
ing the same methodology used to estimate oral RfDs. Highlights
of the RfD methodology are as follows:
RfDs are calculated by collecting the available
animal and human data, noting the various dose
levels (in milligrams per Kg body weight per day)
at which different non-cancer health effects are
seen, and identifying the highest NOEL (No Observed .
Effect Level). The highest NOEL represents the
highest dose at which biologically or statistically
significant effects were not seen. The scientists
also try to identify the LOEL (Lowest Observed
Effect Level), which is the lowest dose at which a
biologically or statistically significant' non
cancer effect was seen.
Many different NOELs and LOELs can be identified
for each chemical, depending on the doses
selected, the spacing between the doses and the
health effects examined and reported by the
researchers. Toxicological research is very
expensive and each experiment cannot be exhaustive
in terms of the number of doses tested and the
number of effects studied. Similarily,
epidemiological research is expensive and subject
to many problems with both measurement of exposures
and identification of exposure-related effects.
This problem introduces a degree of uncertainty in
the identification of the "true" NOEL and LOEL of a
chemical and this uncertainty is carried through
the calculation of the RfD.
A-22
Hi1' I^Hl^H^H^^HM^V IP i
-------
The RfD is calculated by dividing fche highest
reliable NOEL (or lowest reliable LOEL, if
appropriate) by uncertainty factors. The selection
of appropriate uncertainty factors is based,
primarily, on the nature of the study from which
the NOEL, or LOEL, was derived. Each factor has a
value from 1-10 and represents an extra degree of
uncertainty to account for: 1) extrapolation from
the average human to sensitive members^of the human
population, 2) extrapolation from Jthe average
animal to the average human, 3) extrapolation from
subchronic exposures to chronic exposures and 4)
extrapolation from the LOEL to the NOEL.5
By definition, the RfD is based on the critical non-cancer
health effect. This is the effect first seen as exposures in-
crease above zero. Since IEMP studies often examine exposures
that are above the RfD for a particular chemical, it would be
valuable to know what kind of effect(s) might be seen as expo-
sures get higher and higher above the RfD. This type of infor-
mation is not obtainable from the single RfD that is calculated
for a chemical. To help solve this problem, we have taken the
following approach in our IEMP studies:
1) non-cancer health effects are divided into the six
broad health effect categories of liver, kidney, reproduc-
tive, neurobehavioral, fetal developmental, and other. The
category "other" includes a variety of effects such as cardi-
ovascular, respiratory, and gastrointestinal.
2) for each chemical of interest, examine both the oral
and inhalation data for humans and animals.
3) using the RfD methodology, and route-specific toxicity
data, we calculate a separate inhalation and oral "threshold"
for each of the six health effect categories. Obviously, the
more toxic a chemical, and the more it has been studied, the
more health effect categories for which we can esti-
mate "thresholds".
2.) Assessing the Level of Exposure
The second key element (see p. A-16) in estimating health risk
is determining the level of exposure to that chemical. Most of
our exposure assessments attempt to estimate ambient concentra-
tions of substances in the air and drinking water. We then make
certain assumptions about how ambient concentrations relate to
actual human exposure or dose level. For example, we take no
account of population mobility. These assumptionsthe exposure
constants regarding how much air a person breathes or water he
or she drinksare described below.
A-23
-------
We estimate ambient concentrations in two ways: direct moni-
toring or simulation modelling. There are a number of advantages
to the modelling of ambient concentrations over direct monitoring
alone:
o Modelling may provide the only way to estimate ambient
concentrations under alternative exposure scenarios.
o Modelling can take into account the geographic variabi-
lity of a large area. Because monitoring data are often
from a few specific points, they best serve as refe-
rence points for evaluating the performance of air dis-
persion models.
o Modelling is often less costly than extensive monitoring.
o Modelling enables predictions of exposures in any loca-
tion (and, in particular, the location of the most ex-
posed individual), whereas monitoring can only provide
an indication of the exposures in the vicinity of the
sampling sites.
o Modelling links concentration estimates, and hence expo-
sures, to sources. Such source information is important
as a risk management tool in that it allows us to esti-
mate the impact of various pollution control options.
o For some pollutants, there are no accepted methods for
monitoring them in ambient air.
On the other hand, constructing a model of pollutant releases
and resultant ambient concentrations involves making assumptions
about the important processes between pollutant source and human
receptor. Building such a model thus requires an understanding
of those processes, which is not necessary if one can simply
monitor ambient concentrations directly.
When carrying out an IEMP, we first employ existing data-
bases which have been reviewed for currency and quality. We
conduct monitoring as an adjunct as resources allow and use
monitoring data to verify ambient concentrations estimated from
our models. Under certain circumstances, such as situations
where we do not know sources and emissions, monitoring data are
crucial for estimating exposures. Furthermore, an advantage of
reliable, long-term monitoring data is that they provide a more
direct and often simpler means of estimating ambient conditions.
Even with direct monitoring, however, interpretation of results
can be difficult if data are limited or if lab or.sample contami-
nation is suspected.
A-24
-------
The key steps in modelling exposures are:
(1) estimating the number of sources of emissions;
(2) gathering data on emissions (in g/sec.) of target
compounds from selected sources;
(3) gathering release specifications for chemical emis-
sions, such as building dimensions, stack height,
location of release points;
(4) characterizing the processes and pathways by which a
pollutant is transported in air, soil, and water,
including the speed of transport, the extent of dilu-
tion or dispersion, and any chemical transformation
the pollutant might undergo (such as degradation to
a nontoxic, or possibly even more toxic form).
(5) locating receptor sites; and
(6) assigning populations to receptor sites.
.We try to estimate resulting ambient concentrations at
various distances from the source. Estimating long-term average
concentration levels, which are of concern for evaluating chronic
health impacts, is simpler than short-term modeling, which must
take greater account of variations in meteorological conditions.
* '
RISK CHARACTERIZATION
As noted- on pages A-12-13, risk characterization is the part
of risk assessment that brings together the exposure assessment
and the dose-response information in order to estimate the inci-
dence of an adverse effect in a given population. As further
noted, because of limitations in our ability to quantify dose-
response relationships for each of our health effect categories,
we must analyze cancer differently from the non-cancer health
effects. For cancer, we treat the GAG unit risk factor as the
slope of the dose-response curve; therefore, we can use this
value with the results of our".-exposure assessment to estimate a
crude incidence. For the non-cancer health effect categories,
we do not have estimates of the slopes of the dose-response
curves. Instead, we have RfDs and our own "thresholds" for each
of the effect categories. Therefore, for non-cancer, instead of
calculating the incidence of effects, we calculate the number of
people exposed to levels of chemicals that are above the RfDs and
other "thresholds" in order to estimate how many may be at some
increased health risk. For each chemical of interest, we examine
each of the six non-cancer health effect categories separately.
A-25
-------
If exposures to a substance are below the threshold for a
particular non-cancer health effect category, then we expect
little increased risk of effect from that substance. If expo-
sures exceed the estimated threshold for a particular health
effect category, we then indicate the possibility of increased
risk of that effect and try to estimate the size of the popula-
tion exposed to such concentrations. Exposures exceeding an
estimated threshold are generally more appropriate subjects for
further investigation.
Note that the use of GAG potency estimates to evaluate can-
cer risk and the use of RfDs to identify exposures of potential
concern for non-cancer health effects are relatively straightfor-
ward practices and have been used throughout EPA for many years.
Risk Characterization as Applied to an IEMP
In IEM studies, we calculate cancer risk using three measures
of exposures: risk to the most exposed individual (MEI), risk
to the average exposed individual (AEI), and the excess aggregate
population incidence. For all calculations, we use the following
standard assumptions: 1) the average lifetime is 70 years, 2)
the average adult breathes 20 cubic meters of air per day, and
3) the average adult drinks 2 liters of water per day.
We express individual risk as either MEI risk or AEI risk.
We define risk to the MEI as the increased probability that an
individual exposed to the highest concentration of one or more
chemicals will have exposure-related cancer during the course of
his or her lifetime. For the MEI, the exposure is calculated as
either the highest modelled concentration or the average concen-
tration obtained for the monitoring site with the highest daily
or annual average monitored value for one or more chemicals. We
define risk to the AEI as the increased probability that an indi-
vidual exposed to the areawide average concentration of one or
more chemicals will contract cancer during the course of his or
her lifetime. Aggregate population risk is the estimate of the
increased incidence of cancer, above the background rate, in an
exposed population. ,
To estimate the risk to the most exposed individual, we
typically need to know how far an individual lives from the maxi-
mum pollutant concentration near a source. To estimate average
individual risk, we estimate an average pollutant concentration
to determine exposure. To estimate population risk, we must iden-
tify the number of people exposed to a given pollutant concentra-
tion. In our exposure estimates, we are assuming that people are
exposed to outdoor ambient air concentrations 24 hours a day for
a lifetime. In reality, people spend most of their time indoors,
either at home or at work. While our assumption overstates
A-26
-------
actual exposure to outdoor air, the' bias introduced by this pro-
cedure may not be too great, since recent EPA (TEAM) studies
show that many outdoor air contaminants are also found indoors
at equal or greater concentrations. This represents current
standard practice at EPA.
Under these assumptions and exposure scenarios, lifetime
cancer risk to the exposed individual is simply the product of
exposure and potency:
R = Ex P (1)
individual risk exposure potency factor
As discussed above, exposure is the product of the ambient
concentration of the pollutant in the medium of.concern (air or
drinking water) and exposure constants (i.e., the standard assump-
tions of body weight, breathing rate and water consumption):
E = Y x Z (2)
exposure ambient concentration exposure constants
in medium of concern
and
R Y x Z x P (3)
individual ambient exposure potency
risk cone, in medium constants factor
of concern
SAMPLE CALCULATION
The following is a simple example to illustrate how we would
calculate the lifetime risk to the MEI, as well as the lifetime
risk to a population, associated with an ambient air exposure of
ten micrograms per cubic meter (ug/m3) of benzene. We have
divided the geographic area of concern into sections which we
will call "grids."
In our example, we have four grid sections. For each grid, we
have the following average ambient air concentrations and popula-
tion data:
Annual Average
Benzene
Concentration Exposed
Grid f (ug/m3) Population
1 10 20,000
2 8 30,000
3 8 25,000
4 5 25,000
A-27
-------
The MEI risk of cancer for an individual exposed to 10 ug/m3
of benzene in the air (the highest benzene exposure) is calculat-
ed by using equation (3). The ambient concentration (Y) is 10
ug/m3, which is equivalent to 0.01 milligrams per cubic meter or
0.01 mg/m3. The .potency factor (P) for inhaled benzene, devel-
oped by EPA's Carcinogenic Assessment Group (GAG), is 0.029
(mg/kg/day)-1. This value means that an individual inhaling one
milligram (mg) of benzene per kilogram body weight per day for a
lifetime has an estimated increased probability of developing
cancer of about three in one hundred (upper-bound estimate).
Using equation (3) and the maximum ambient concentration
value of 0.01 mg/m3, we calculate the risk to the MEI as follows:
R = Y x Z . x P (4)
MEI Avg. annual exposure potency
Risk ambient cone. constants factor
where, the exposure constants (Z) are adult body weight (70 kg.)
and adult breathing rate (20 m3/day)
therefore,
R = 0.01 mg/m3 x (20 m3/day x 1/70 kg) x 0.029 (mg/kg/day)-1
= 8.3 x 10-5
The lifetime upper-bound estimate of risk to the MEI (R) in this
example is 8.3 x 10-5 or roughly eight chances in 100,000 of
developing cancer over a lifetime given constant exposure, to 0.01
mg/m3 per day.
The risk to the average exposed individual (AEI) is calcu-
lated the same way, using an average exposure determined from
either monitoring data or air dispersion models.
The final step in our example is to estimate the increased
incidence of cancer in the total population in our four grid
geographic area. To do this, we multiply the risk to the AEI in
each grid by the number of people in that grid. We then sum
these estimates of incidence across all grids to calculate the
lifetime aggregate incidence of cancer for our four .grids.
Generally, EPA presents incidence as the expected number of
excess cancer cases per year. Dividing the upper-bound lifetime
estimate by 70, we arrive at the upper-bound estimate of the
annual average number of excess cancer cases in the population of
concern.
We estimate individual and population risks from ingestion
of drinking water in exactly the same way we estimate the risks
from inhalation. The only difference is that the potency factors
A-28
-------
have different values and we use the water consumption exposure
constant of 2 liters/ day. Because the potency of a chemical may
differ for different routes of exposure, we must use the potency
factor that is appropriate for ingestion. This potency factor is
provided to us by the Carcinogen Assessment Group. Concentrations
of selected chemicals in the drinking water are measured or esti-
mated and then used with the oral potency factors in equation
(3) to estimate risk. We calculate both maximum individual risk
and aggregate excess incidence in a population in the same way
indicated in our four grid example.
Risk Characterization as Applied to the Baltimore IEMP
The Baltimore Phase I activities emphasized cancer as the
primary health effect of concern but some attention was given to
non-cancer effects of chemicals that were addressed as carcino-
gens. In Phase II, we will more broadly consider noncarcinogenic
effects to the extent that available data will allow. We empha-
sized cancer in Phase I because the public has expressed frequent
concern about the possible link between exposure to environmental
pollution and the incidence of cancer Also, the quantitative risk
assessment methodology for cancer has more scientific credibility
within EPA than do similar methods for assessing non-cancer
effects.
INTERPRETING RISK ASSESSMENT RESULTS
The estimates of individual health risk and aggregate inc-
idence from exposure to toxics should not be interpreted as pre-
cise or absolute estimates of future health effects. The simpli-
fying assumptions and uncertainties in both the toxicology and
exposure components are simply too great to justify a high level
of confidence in the precision of the results.
The potency and the threshold estimates used in this study
are consistently conservative in the direction of overestimating
risk; they may overestimate the likely effects of chemical expo-
sure but are unlikely to underestimate them. Such an estimate,
however, does not necessarily give a realistic prediction of the
risk. The true value of the risk is unknown, and may be as low
as zero." This leads to assessments that have biases for health
protection when the results are used. By contrast, our exposure
estimates are not as clearly conservative; some assumptions are
conservative while others are our best guess of an actual value.
Overall, we have tried to be somewhat conservative in our expo-
sure assessments, as is appropriate in a priority-setting exer-
cise. It is important to read the detailed chapters carefully to
understand what confidence to place in a particular risk estimate.
A-29
-------
On the other hand, we may understate risks to the extent
that we do not estimate exposure and risks from all sources and
pollutants that may have toxic effects. We have tried to identi-
fy the sources and pollutants of greatest concern and address
those that data and methods allow us to cover. We acknowledge
that we have not addressed all sources and pollutants.
Because of the uncertainties involved, the results should
not be interpreted too literally. For example, one should not
conclude that a source projected to cause three cases of cancer
per year is clearly worse than a .source projected to cause two
cases; given the overall uncertainty of the analysis, these two
results are indistinguishable. On the other hand, for example,
it is reasonable to conclude that a source projected to result in
one case per year represents less risk to the population than a
source projected to result in ten cases per year. Despite the
uncertainties, our risk estimates are useful for roughly assess-
ing the potential magnitude of the overall risks from particular
pollutants, sources, and pathways; comparing issues with one
another; and setting priorities among environmental issues and
concerns.
A-30
-------
REFERENCES FOR APPENDIX A
1. National Research Council (1983). Risk Assessment in the
Federal government: Managing the Process. Committee on the
Institutional Means for Assessment of Risks to Public Health,
Commission on Life Sciences, national Academy of Sciences.
National Academy Press, Washington, D.C.
2. International Agency for Research on Cancer (1982). IARC
MONOGRAPHS on the EVALUATION OF THE CARCINOGENIC RISK OF CHEMI-
CALS TO HUMANS. World Health Organization, Volumes 1 to 29,
Supplement 4, pp. 13, 14.
3. U.S. Environmental Protection Agency (1986). Guidelines for
Carcingenic Risk Assessment. Federal Register, Vol. 51, No. 185,
33993 - 34003.
4. U.S. Environmental Protection Agency (1985). Health Assess-
ment Document for 1,2-Dichloroethane (Ethylene Dichloride).
Office of Health and Environmental Assessment; Washington, D.C.
pp. 9-235 to 9-241.
5. U.S. Environmental Protection Agency (1986). Reference Doses
(RfD): Description and Use in Health Risk Assessments. Draft
prepared by the RfD Workgroup for the Risk Assessment Forum,
May 1986.
A-31
-------
APPENDIX B
LIST OF ISSUE PAPERS
SAMPLE ISSUE PAPER: SEPTIC TANKS
-------
List Of Phase I Issue Papers
Developed For
Technical Advisory Committee
Benzene
Combustion/Incinerator
Chromium in/around Baltimore Harbor(4/24/85)
Ground-water Resource Degradation Ranking for Chromium in
the Area around Baltimore Harbor(4/25/85)
Landfills
Multi-media Metals
-Addendum to Multi-media Metals
Pesticides
Publicly Owned Sewage Treatment Facilities
Septic Tanks
Surface Impoundments in the Baltimore Study Area
Water Discharges to Baltimore Harbor
Baltimore Health Data
Note Regarding Health Scores
-------
BALTIMORE IEMP ISSUE PAPER. SEPTIC TANKS
SUMMARY
This issue paper explores the impacts of ground water
contamination by septic tank systems in the IEMP Baltimore study
area. About 12% of the households in Baltimore county and 32% of
the households in Anne Arundel county dispose of domestic wastes
in septic tanks; most of these households use private wells to
supply drinking water. Septic tanks are not used in Baltimore
City.
In Baltimore County both the coastal plain and piedmont
regions have experienced localized groundwater contamination from
septic tanks. The greatest risk appears to be from the improper
disposal of chlorinated solvents, particularly in areas of the
piedmont with fractured ground-water flow. In a few isolated
cases ground-water concentrations of trichloroethylene from sep-
tic tanks have reached concentrations as high as 47 ug/L. Life-
time exposure to this concentration would be associated with
predicted upper bound cancer risks exceeding 1 x 10~5; however,
contaminated wells in this area are no longer used as sources of
drinking water.
In Anne Arundel County the far eastern regions along
Chesapeake Bay are at greatest risk from septic tank use. In
the Mayo Peninsula area nitrate concentrations more than.two
times the drinking water standard of 10 mg/L have been found. It
is difficult to quantify the risks from these exposures, but it
is clear that infants ingesting this water are at increased risk
of methemeglobinemia. This area is characterized by a shallow
water table, high septic tank density, and low soil permeabili-
ties . Contamination of drinking water wells by septic tanks
does not appear to be a major problem in other areas of the
county.
In some parts of the U.S., septic tank leachate fouls
streams, but this is not recognized as a problem in the Baltimore
study area.
Although bacterial and viral contamination from septi9 tanks
is probably one of the most widespread pollution problems in the
area, this analysis does not include microbial pathogens.
SOURCE CHARACTERISTICS AND POLLUTANTS
Septic tank systems consist of a tank, connected to a local
dwelling or set .of dwellings, which receives the raw wastes for
treatment by sedimentation, flotation and anaerobic biological
degradation; and a system of open-jointed or perforated pipes
B-l
-------
which take the effluent from the tank, and distribute it through
the soil for treatment via oxidation and microbial degradation or
adsorption. As such, the system is designed to release contami-
nants to the subsurface as a generally efficient means of efflu-
ent treatment. However, in situations where the treatment and/or
assimilative capacity of the soil is exceeded these contaminants
can pollute ground water.
Contaminants released from septic tank systems can be
classified into two categories. The first are contaminants which
are natural components of biological and household wastes and
includes the following:
o nutrients and other dissolved materials, particularly
nitrates, and
o microbial pathogens (bacteria and viruses), which are not
being considered in this analysis.
The second category of contaminants are those introduced to the
system through septic tank cleaning and the disposal of consumer
products or industrial wastes. Unlike contaminants in the first
group, these pollutants are typically chlorinated organic sol-
vents used for household cleaning or industrial degreasing pro-
cesses, and are not always present in the household effluent.
The most common of these solvents are probably trichloroethylene
and methylene chloride; we limit the project discussion of sol-
vents to these two.
Contaminants from both of the above categories have been
identified as ground-water pollutants in the Baltimore study area.
TOXICITY
Hazard Identification
As described earlier, nitrates and chlorinated solvents are
the pollutants covered by this paper. Adverse effects of each
are discussed below, emphasizing human health effects.
Nitrate is a highly soluble and mobile compound with a half-
life in ground water of around one and one-half years. It is
commonly found in high concentrations in septic effluent. At the
low concentrations generally found in ground water the most seri-
ous health effect associated with nitrates is methemoglobinemia,
a reversible disorder in which blood hemoglobin is oxidized to
methemoglogin which is unable to transport oxygen to body tissues.
Ground-water contaminated by septic tanks rarely or never has ni-
trate, concentrations high enough to cause this effect in healthy
adults, but infants are susceptible at much lower levels. Nitrate
B-2
-------
concentrations for less than 20 tng/L have caused the condition in
infants, although cases are more often seen in association with
concentrations greater than 50 mg/L.
Methylene chloride has been classified as a potential human
carcinogen byEPA1sCarcinogen Assessment Group (GAG) on the
basis of laboratory animal studies. In addition, IEMP toxicolo-
gists have reviewed studies performed by EPA and others which
indicate that methylene chloride exposures are associated with
teratogenic effects, fat vacuolization in the liver, and tempor-
ary increases in the levels of carboxyhemoglogin in the blood.
Trichloroethylene is also classified as a human carcinogen
by GAG. It may also cause fat vacuolization and necrosis in the
liver.
i
Ecological effects from septic tank leachate are primarily
associated with high loading rates of nutrients and dissolved
organic carbon. Dilution of septic tank leachate is sufficiently
high that direct toxic effects are probably negligible; the major
effect on aquatic systems is increased potential for eutrophica-
tion.
Dose-Response Assessment
Exhibit A contains information on the dose-response
functions for the pollutants covered in this issue paper. The
potency estimates for carcinogens are upper bound estimates de-
rived by EPA's Carcinogen Assessment Group (GAG). Thresholds
and potency estimates for other effects from methylene chloride
exposures were derived by IEMP and represent median values. The
no observed effects level (NOEL) reported for nitrate was identi-
fied by the National Research Council.
On the basis of the above hazard assessment discussion the
primary human health hazards from septic tank contamination of
ground-water in the study area are:
o cancer from ingestion of trichloroethylene, and/or
methylene chloride, and
o acute methemoglobinemia in infants from nitrates.
B-3
-------
Exihibit A
Dose-Response Functions for
Septic Tank Pollutants
Contaminant
Nitrate
Methylene
chloride
TCE (Trichloro-
ethylene)
Effect
Methemeglobinemia
Cancer
Teratognicity
Liver (fatty vacuol-
ization)
Cancer
Threshold
(tng/kg day
NOEL = 10 tng/1
0
0.186
0.186
0
Potency^ /
6.3x10-4
1 .65x10-3
2.28x10-3
1 .9x10-2
Source
NRC
GAG
IEMD*
IEMD*
CAG
1 . Potency unit for Methylene chloride and TGE is (mg/kg day)"
peer reviewed potency is not available for nitrate, thus the
National Research Council NOEL will be used to qualitatively
assess nitrate risks.
EXPOSURE
In this section we examine the potential for exposure to the
previously discussed contaminants in terms of the sources, path-
ways, ambient contaminant levels, and exposed populations.
Sources
Septic tanks are the most common alternative to municipal
sewers as a means of disposing domestic wastes. In the Baltimore
study area they are used in both Baltimore and Anne Arundel coun-
ties. Septic tanks are not used in Baltimore City, so the ba-
lance of this section will focus only on their use in the two
counties of interest. The map in Exhibit B indicates those areas
in these conties which are most vulnerable to septic tank conta-
mination.
Baltimore County
According to 1980 census data approximately 11.7 percent of
the households in Baltimore County use septic tank systems. The
majority of these are located in the northern part of the county.
B-4
-------
Exhibit B
Map of Baltimore and
Anne Arundel Counties
Phoenix-Nike-Missile
Site
Greenspring-N
Missile Site
Geologic Fall
«I
+/ Areas in which septic 5£
tanks are the most
common method for
sewage disposal
Mayo
Peninsul
-------
EXHIBIT C
Factors Affecting Contaminant Fate and Transport
Factor
Role in Fate and Transport of
Contaminants from Septic Tanks
Unsaturated Zone
Depth
Affects transport-time of constituents;, from
tank field ,to drinking water source-; thus
affecting amount of hydrolysis or denitrifi-
cation in unsaturated zone. >
Soil Permeability
Low permeabilities result in reduced trans-
port time, high permeabilities cause; pooling
of effluent at soil surface, both high and
low decrease efficiency of effluent treatment.
Fraction of Orga-
nic Carbon and
Cation Exchange
Capacity
Important determinants of contaminant mobil-
ity; chlorinated organics adsorbe on soils
high in organic carbon, increasing transport
.time. Neither factor has an appreciable
effect on nitrate mobility. ,
Subsurface
Lithology
**
Non-productive aquifers and aquitards can
protect underlying sources of ground-water
by significantly slowing or preventing conta-
minant transport, -i'lmportant factor in several
regions of Bait, study area. '
Ground-water
Velocity and
Direction
High velocity promotes contaminant dilution
and dispersion, decreases time available for
contaminant decay. Low velocity decreases
dilution but increases amount .of contaminant
decay. Upgradient wells are unlikely to be
containinantd , downgradient well are most
likely to be contaminated. ,. ,./ -
Septic Tank
Density
High and medium densities result in large in-
puts' of contaminants to the subsurface, in-
crease potential for well contamination.
Distance From Sep-
tic System to
Drinking Water
Well
Greater distances increase time for contami-
nant decay and area for effluent treatment
and contaminant assimilation.
-------
EXHIBIT - D
Typical Hydrogeologic Conditions in Regions
Using Septic Tanks in the Baltimore Study Area
Characteristic
County
Baltimore
Depth to Ground-
Water (feet)
Permeability
Fraction of
Organic Carbon
Cation Exchange
Capacity (meq)
Expected Ground-
Water Velocity
(ft/yr)
Lithology
0-10 (coastal plain)
15 - > 100 (piedmont)
very low (coastal
plain)
low to high (piedmont
NA
NA
NA
Shallow unconsolidat-
ed sands, gravel, and
silt (Teritary & Qua-
tenary Sediments),
underlain by Upper
Cretaceous.clays and
coarse sands and Pata-
psco sands, gravel and
clay containing high
concentrations of dis-
solved minerals
Coastal Plain - Unconso-
iidated sands, gravel
and clay (Pleistocene
and Recent Age), under-
lain by Anne Arundel
clay and unconsolidated
Patuxent formation.
Piedmont - Crystalline
rock formations includ-
ing well weathered Coc-~
keysville marble; frac-
tured Wissahickon schist
and phyllite; highly
jointed Peters Creek ,;
quartzite; and some low
permeability gabbro,
serpentine, granite, and
pyroxenite.
NA - .Information not available. ;
J_/ Anne Arundel County information characteristic of Mayo penin-
sula area, which is considered to be the most vunerable are in
the county.
2/ Estimate based on aquifer type. Obtained in personal communi-
cation with Geraghty and Miller, Inc., Annapolis, Md, January
23, 1975.
Anne Arundel^ /
0-2 (wet months)
2-9 (dry months)
Very low to low
NA
NA
1 - 1027
-------
EXHIBIT E
Summary of Ground-Water Pathway Characteristics in
Baltimore and Anne Arundel Counties
Baltimore County
1 . Coastal Plain (Southeast region of county along Chesapeake
-local ground-water obtained from unconsolidated sands, gra-
vels, and clays with very shallow, fluctuating, water table
(0 - 10 feet)
-low soil peraeabilities cause ponding and surface contamina-
tion, reduce ability of soil to adequately treat effluent
-high density septic tank population, most tanks installed
prior to current county regs . requiring that drinking water
wells be placed no closer than 150 feet downgradient or 100
feet upgradient from any septic system, thus there is close
proximity of wells and tanks.
2. Piedmont (Area north and west of geologic fall line, over 80
percent of county)
-ground-water flow primarily in fractured bedrock and solution
cavities, some flow in highly weathered bedrock. These crys-
talline formations are not generally productive sources of
ground -water.
-widely varying soil permeabilities
-ground-water depths generally greater than- 30 feet, some
isolated locations of shallower ground-watr used as drinking
water source ,
-in low permeability areas tanks have been placed up to 12
feet below ground ito avoid surface contamination. This has
led to a shallow buffer zone between tanks and ground-water
and resulted in local ground-water contamination in isolated
cases .
Anne Arundel County
-located entirely in coastal plain, subsurface made up of
unconsolidated sediments overlaying clay and coarse sands
-eastern region of county most vulnerable, ground-water in
this area obtained from very shallow Aquia sand sediments
(water table 0-10 feet)
-------
EXHIBIT E (con11)
-soils j;im vulnerable areas have .very low permeability
-septic tanks in area oftn modified to avoid surface contami-
nation by piping effluent directly into local ground-water.
This has resulted in vseripus ground-water contamination
problems in general area. -*"-""
'*.."
-slow ground-water flows in eastern regions do not allow for
much contaminant dilution
i,. " -^
-most eastern communities were developed before county ordi-
nances requiring the placement of shallow wells, at least 100
feet from a septic tank, deep wells 75 feet from tank. Thus
wells and tanks are in close proximity in these areas.
-balance,ftof county not as vulnerable *due to extensive sewering,
better soil permeabilities, less "tank density, and more
common ii&e of deeper Patapsco formation as source of drinking
water. r ?
-------
EXHIBIT F
Identified Constituents and Concentration Levels in
Drinking Water Wells Contaminated by Septic Tank Use
(all concentrations mg/1)
Contaminant
Nitrate (as N)
Ave . cone
Cone, range
Methylene
Chloride
TCE
Cone, range
MCL
10
Baltimore County
Not identified as a
contaminant
Not identified as a
contaminant
.006 - .047t
Anne Arundel County
14.0*
10.0 - 23.0
Not identified
as a contaminant
Not identified
as a contaminant
* Nitrate levels for Anne Arundel county are representative of
Mayo peninsula area only. Source of data: Anne Arundel
County Health Dept. Mayo Peninsula Water and Sewerage Survey,
1978.
t TCE concentrations pertain to wells near the Phoenix-Nike and
Greenspring-Nike missel bases. Elsewhere, no TCE has been
detected in limited sampling of ground water.
-------
Some communities located on the Chesa'peake Bay to the southeast
also use septic tanks.
Septic tank density in the county is quite variable. The
southeast region was originally populated by summer homes and
cottages which have now become year-round homes. Density in this
area is relatively high because it was developed prior to the
development of current sewage disposal ordinances. Lot sizes in
this area can be as small as 1/4 acre, each with their own septic
tank. In the northern sections development is less dense, al-
though some pockets of high density can be found. Much of this
region has been developed under county ordinances requiring a
minimum lot size of four acres for septic tank use.
Several of the past serious contamination incidents in
Baltimore County have been the result of using large septic tanks
as disposal systems at military installations. These systems
were properly designed to dispose of human wastes from a large
number of people. However, they were also used to dispose of
many other types of waste including chlorinated solvents which
eventually contaminated ground-water. Although few other inci-
dents of this type have been discovered, the chemical contamina-
tion of ground-water from septic tank use by commercial units
such as shopping centers and gasoline stations could be a wide-
spread problem. */_
Anne Arundel County
1980 census data indicate that approximately 32.2 percent of
the households in Anne Arundel county use septic systems. Al-
though septic tanks are used to some degree in all regions of the
county, the large majority of the areas most vulnerable to ground-
water contamination because of poor hydrogeology and/or high popu^-
lation density are now served by municipal sewer systems. How-
ever, in some highly vulnerable areas located south of Annapolis
along the western shore of the Bay, septic tanks are still in use
in high densities, and ground-water contamination has resulted.
The most highly studied region in this vulnerable area is
the Mayo Peninsula. Septic tank density in this area is extreme-
ly high, with about 230 tanks/sq. mile. This very high density
is not typical of the majority of the county. In the Mayo peni-
nsula area drinking water is obtained from privately owned
shallow wells which are particularly vulnerable to contanimina-
tion. Water obtained from deeper wells in these coastal areas
generally contains high concentrations of iron and sulphur,
I/ Personal communication, Mr. Thomas Ernst, Supervisor, Water
and Sewerage Programs, Bait. Co. Dept. of ^Health, with Paul
Lewandowski, Sobotka & Co., Inc. .
B-5 .
-------
making it an unsuitable option as a drinking water source. Al-
though the May Peninsula is an extreme case, other areas located
in the coastal region may also be particularly vulnerable. Un-
like Baltimore County, Anne Arundel County has not experienced
chemical contamination due to commercial scale septic tank use,
and it is not expected to be a major problem.^/
Pathways
Because septic tanks are -designed to release contaminants
into the subsurface environment, any septic system can potential-
ly contaminate ground-water. The use of this contaminated water
is a source of drinking water is the most likely pathway for
exposure to pollutants resulting from the use of septic tanks,
and thus ingestion via drinking water is the exposure pathway of
interest. Although septic tank use can result in surface water
contamination, we do not regard this as likely to be a signifi-
cant exposure pathway because: (1) local water suppliers monitor
for nitrate and maintain concentrations under the no observable
effects level (NOEL) to comply with Safe Drinking Water Act
requirements, (2) solvents entering surface water via septic tank
leachate are initially present in dilute concentrations, and
quickly volatilize, and (3) treatment of surface water will
remove some of the contaminants if they are present. Therefore,
we do not consider this pathway to be an important source of
exposure.
There are several factors which play important roles in
determining the extent to which a drinking water well will be
contaminated by effluent from septic tank systems. These include
contaminant release rates, soil and hydrogeologic conditions,
location factors, and system design. Although releases will vary
depending on the number of individuals using the system and the
composition of wastes disposed, previous studies have found that
a single tank serving an average 4-person household will release
approximately 280 m^/yr of effluent into the unsaturated zone.
This results in a typical system releasing over 5 kg/yr of nitr-
ate. Chlorinated solvents are not as easily typified; however,
it is clear that total releases of the above constituents can be
substantial in areas of thigh, or even medium, septic tank use.
Soil and hydrogeologic conditions largely determine the,fate
and transport of the released pollutants. Contamination incidents
in both counties indicate that problems are of ten very site-
specific, and that predicting fate and transport on the basis of
local hydrogeologic conditions is rarely a simple matter. How-
ever, there are a number of important characteristics for which
2/ Personal communication, Singh Dhillon, Anne Arudel Co. Dept.
of Health, with Paul Lewandowski, Sobotka & Co., Inc.
B-6
-------
general statements concerning their Impact can be made. These
include:
o unsaturated zone depth,
o soil characteristics, particularly permeability, fraction
of organic carbon and the cation exchange capacity of the
soil,
o lithology of the sub-surface, including the presence of
overlying aquifers and aquitards, and
o local ground-water flow rates and direction.
In addition, locational factors such as septic tank densities,
and distances from septic systems to drinking water wells affect
the level of well contamination.
Exhibit C summarizes the role of each of the above factors
in contaminant fate and transport, and Exhibit D contains infor-
mation concerning typical hydrogeologic conditions in the regions
of Baltimore and Anne Arundel counties where septic systems are
commonly used.
In these exhibits we briefly present the most important
factors pertaining to ground-water contamination as the pathway
for human exposures. In Exhibit E we summarize the potential for
exposure in Baltimore and Anne Arundel counties via this pathway.
Ambient Concentrations
The movement of contaminants based on the factors discussed
in the previous section can potentially result in contamination
of drinking water wells at concentrations above drinking water
standards. These concentrations can result in human health risks
that could be considered significant.
Exposures to nitrate and methylene choride from drinking
ground-water contaminated by septic tank effluent have recently
been modeled in a study conducted by EPA's Office of Policy
Analysis that compared health risks from six different sources of
ground-water contamination. Preliminary results indicated that
in high and medium septic tank density areas methylene chloride
concentrations in wells can reach levels resulting in upper bound
cancer risks on the order of 10"^. This contaminant moves rapid-
ly through the unsaturated zone, thus both shallow and deeper
wells can be contaminatd. Because of this rapid movement and its
moderate carcinogenic potency, methylene chloride was found to be
a more important contributor to risk than nitrate. This was par-
ticularly true in areas of rapid ground-water flow, and would
likely also be the case in situations where wells were located
short distances from the septic tank field. As mentioned earlier,
B-7
-------
methylene chloride is considered a carcinogen by EPA and it was
modeled as such.
Although nitrate exposures can have serioud effects, nitrate
concentration3 in septic effluent are generally quite low (around
wO mg/L), resulting in/relatively small loading rates to* the sub-
surface, particularly an relation to the low toxicological potency
of nitrate. In addition, nitrate degradation in groun'd-water is
significant, with a half-life of 1.5 years. The combination of
these factors results -an insignificant exposures to nitrate from
septic effluent in most cases. However, high- revels cSf nitrate
associated with septic tnk use have been found in Ann Arundel
County. ,
.- ^Below we present data on concentrations 6f contaminants
found in drinking water wells which can potentially be linked to
septic tank use in each county. These data are*wsummarized in
Exhibit F, which also contains current drinking water standards
for comparison purposes. We also discuss some significant past
contamination incidents to illustrate the potential level* of
exposure in the study area. f! ;5H>
Baltimore County ~ - » ; ?
Baltimore County has experienced several major ground-water
contamination incidents directly linked to septic tank useage
which have resulted in drinking water wells contaminated with
trichloroethylene. The most important and well-documented of
these occured at the Phoenix-Nike missile location outside of
Jacksonville^ At this site contamination of ground-water appa-
rently resulted from the disposal of TCE in the septic tank used
to dispose of the base's sewage. Reasons for this disposal are
not clear; however wells in- the area are considered unfit for
drinking water use because 6f TCE levels as high as 47 ug/L at
site boundaries and 8.6' ug/L in private wells 180 meters north
of the site, which is the predominant direction of ground-water
flow. Although the site was closed in 1966, this contamination
was not discbvered until the late 1970' s. Eight to tfn private
wells have been contaminated. In a similar incident, -the septic
system servicing .the now closed Greenspring-Nike missiie base,
locatejl near the Chestnut Ridge subdivision 10 milWs west of
Cockeysville, has also been implicated as a source ofi/^TCE co'nta-
mination. In this case, contaminant levels in drinking water
wells on the base were as high as 15>
-------
Nitrate contamination from septic tanks has not been identi-
fied as a major problem in the county, including the vulnerable
coastal plain area. In the northern .portions of the county ni-
trate concentrations higher than^the drinking water standard have
been found in ground-water. However, this is most likely associ-
ated withv naturally occurring soils.f,and agricultural uses of ni-
t-y-af-o *i ^'"' '- -'*'
ULciUC. /- . .^ . , .,
> . , . ..,3j ; \&. . $ '--. .
Anne, Arundel: County *^.:x .x .-.,-,..' .
.*- ' «.. ' ':' ' *''' -'; *;*£"* - >
Past surveys, of ground-water^users in the area indicate that
ground-water"''pollution has been a significant problem. A large
survey conducted, approximately 15 years ago found that 66 spercent
of the drinking water wells infethe county exceeded the Maximum
Contaminant Levelis ,f(MCLs), established .under the Safe Drinking
Water Act, for at.-least one pollutant.^/ The importance of septic
tank use as a s.ojUyceje of .this contamination was not explored; hpw-
ever^~|i*fe is likely- to be less »of a contributing factor*now Ithan
it was* then, because of the continued sewering of many vulnerable
areas'.i«*£:""> £ *.-::;?' '.%*.,*.£ '
"^'"' ''. . ' 'A . . *?*£ ' .:;.''' , .
',ft'f - '*'< . . -
ijA 1978 survey of the Mayo peninsula area found that over 35
percent of the households in the* area had experienced problems,
with their septic tank operations.^/ A drinking water sampling
study undertaken concurrently found~that about ten percent of the
drinking water wells in the peninsula were contaminated with fecal
coliform bacteria (i.e., bacteria were present) and five percent
of the .wells-were contaminated with nitrate in concentrations in
excess ofv the TO mg/L MCL. ^As Exhibit F illustrates, nitrate con-
centrations ranged from 10 to 23 mg/L, with an average of around
14 mg/L. Because Mayo county is a particularly vulnerable area,
these frequencies and contaminant levels are likely to be the
highest in
-------
Anne Arundel County
Although 32 percent of the households in Anne Arundel county
use septic tanks, and 26.7 percent obtain their drinking water
from private wells, the majority of these individuals are unlike-
ly to be exposed to contamination caused by septic tank use.
Those most likely to be exposed are individuals living in the
eastern portions of the county south of Annapolis, particularly
people living in the Mayo peninsula area. The 1978 study of the
peninsula identified 1925 households in this area; with a total
population of 5,922 according to the 1980 census. This repre-
sents approximately 1.5 percent of the total population of the
county.
The above exposure assessment indicates that contamination
of ground-water due to the use of septic systems can and has
occured in the Baltimore study area. Human exposures to?septic
tank pollutants via the ingestion of contaminated ground-water is
possible at significant concentration levels in some specific
areas, and most problems will be very localized. The next sec-
tion discusses the human health risks associated with these poten-
tial exposures.
RISK CHARACTERIZATION
Human health risks from the use of septic tanks in the
Baltimore study area are primarily due to exposures to hazardous
concentrations of nitrate, trichloroethylene (TCE), and possibly
methylene chloride. Because the use of septic tanks in Baltimore
and Anne Arundel counties is local in nature, estimated risks
pertain only to those areas where septic tanks are used. Also,
because there are no data available on chemical pollutants in
many privately-owned wells, it is difficult to determine whether
there are substantial chronic human health risks from septic tank
releases over much of the area.
Below we specifically discuss the risks from each of these
pollutants. We have divided the chemical pollutants into two
categories, those whose major effect is cancer (trichloroethylen
and methylene chloride), ' and those considered to be non-carcino-
genic (nitrates) because of important differences in dose-
response assumptions used in our risk assessments.
Non-Carcinogens; Nitrate
Because we currently do not have a dose-response function
for nitrate which has been peer-reviewed, we can not estimate
individual risks or population risks for this pollutant. However,
in Anne Arundel County concentration levels of nitrate as high as
23.0 mg/L (as nitrogen) have been found. These levels are more
B-10
-------
than two times greater than the no ^observed effect level for
nitrate of 10 mg/L nitrate-nitrogen. Because of this, infants
exposed to drinking water in the Mayo peninsula area are at some
risk of methemoglobinemia. According to Singh Dhillon of the
Anne Arundel Dept. of Health, the Mayo Peninsula community has
been warned several times that children should not use well water
for drinking, so there may be little exposure. Adults are prob-
ably not at risk.
In Baltimore County no nitrate contamination has been relat-
ed to septic tanks, therefore added risk of methemoglobinemia is
likely to be very low in this area.
Carcinogens; TCE and Methylene Chloride
t. j
Trichloroethylene is the only carcinogen known to be present
due to septic tank releases in the study area.Risks from trich-
loroethylene exposures can be calculated using CAG potency values
derived based on modified multistage model for carcinogenicity.
This model assumes a linear dose-response curve at low doses and
no biologic threshold. The potency estimates are the upper bound
95% confidence limits, so the resulting risk estimates are conser-
vative (i.e., tend to overestimate risk).
There is currently no exposure at the sites where TCE conta-
mination is known to exist. However, if we assume that TCE con-
tamination exists undetected at other sites, and is present at
the same concentrations as those detected around the Nike missile
sites, we can estimate ri~sks at other potential contamination
sites by using the following assumptions:
o average adult consumption of drinking water is 2 liters/
day,
o average adult body weight is 70 kg, and
o contaminant concentrations remain constant over an indi-
viduals lifetime, which is assumed to be 70 years.
Using these assumptions, individuals upper bound cancer risks at
the TCE concentrations listed in Exhibit D range from 3.2 x 10~»6
to 2.6 x 10~5.
In Anne Arundel County no contamination of drinking water
wells by chlorinated solvents via septic tanks has been identi-
fied.
In a study of human health risks posed by ..releases from
different sources of ground water contamination, EPA's prelimi-
nary results indicated upper bound cancer risks... from exposure to
methylene chloride range from about 10~4 to 10~*0 under different
B-ll
-------
hydrogeological and release scenarios. If people are exposed to
methylene chloride from septic tank releases, this range probably
brackets the risks in the Baltimore study area.
To summarize, contamination of ground-water due to septic
tank use is not a widespread problem in the Baltimore study area.
However, a history of contamination incidents shows that the
study area does contain some vulnerable areas in which concentra-
tions of nitrates and trichloroethylene can reach levels suffici-
ent to pose human health risks. Future studies of this issue
would need to focus on accurately identifying these areas, deter-
mining the types and volumes of releases, and detecting concen-
trations of pollutants in drinking water wells. This would be
particularly useful in determining the importance septic tanks as
a source of chlorinated solvent contamination of ground-water.
Although few incidents of solvent contamination have been report-
ed in the study area, several local environmental scientists
believed such contamination to be more widespread than previously
thought, particularly in Baltimore County.
Because the exposure pathway of interest is the drinking of
contaminated ground water, any future studies should also include
risk mitigation and prevention strategies as the process of char-
acterizing the possible extent of septic tank-related problems.
Drinking water contamination problems due to septic tanks have
traditonally been solved by the extension of municipal water and
sewer services into vulnerable areas. However, this is often a
very costly option which encourages increased development and
consequently can bring new human health risk factors into these
areas. Because the population at risk from septic tank pollution
appears to be fairly low in the study area, additional focus for
future work should probably be on the development of relatively
inexpensive and efficient means of managing contamination problems
at the source level (e.g., limits on use of solvents for cleaning
tanks) as well as the continued development of local regulations
pertaining to septic tank siting.
COMMENTARY
This issue is directly related to two other issues in the
Baltimore study:
o risks from sewage treatment (the primary alternative to
septic tanks for treating domestic wastewater), including
sludge disposal, and
o risks from residential and agricultural application of
fertilizers and pesticides (these sources also are in the
immediate proximity of residential wells, and can release
B-12
-------
nitrates and chlorinated organics).
Although septic tanks are recognized as potential sources of
pollution in several areas, and have been studied for several pol-
lutants in seveal areas, there is a moderate-to-high level of un-
certainty in most of the risk assessment aspects of this issue.
In terms of hazard assessment, a dose-response function for
nitrate is not readily available, and there is considerable uncer-
tainty in the dose-response functions and weight of evidence for
carcinogenicity for TCE and methylene chloride. And in terms of
exposure assessment, routine monitoring of private wells does not
occur for even "conventional" pollutants like nitrate, much less
trace organics, so it is very hard to estimate exposure levels.
For these reasons, and because of the time lag that occurs bet-
ween release of pollutants to ground water and exposure to pollu-
tants in drinking water wells, it is relatively difficult to
characterize the risks from septic tank releases.
B-13
-------
APPENDIX C
EPILOG
-------
APPENDIX C. EPILOG
In this "concluding unscientific postscript" we describe
some of the ways in which we think the Baltimore IEMP has already
contributed to improved environmental management for the study
area. These improvements include tangible accomplishments with
an immediate consequence for improved public health (e.g., the
change which reduced the standard for the lead used in solder and
flux in residential plumbing). They also include less tangible
events which may prove to be more important in the long run for
environmental management; our use of an explicitly risk-driven
process to help set priorities among the environmental problems
is an example. This is a new, and we hope permanent, way of doing
business. Here, then, are our concluding thoughts on Phase I.
The Baltimore IEMP Established a Prototype for Locally-Managed
Setting of Priorities
o In a first for IEMP projects EPA delegated authority to
the local-level to manage the Baltimore IEMP with the
result of greater activity and commitment on the part of
local officials. This includes project direction, polic-
ies budget allocation, priority setting, and technical
work EPA functions in a staff capacity.
o This is the first time different governmental agencies
(state, local. Federal) in the study area have had a forum
in which to discuss a wide range of environmental issues
and to set priorities across them using a common language
(risk) and tools. This has provided an opportunity for
faster action in identifying and acting on potential
environmental problems (e.g. the lead solder issue dis-
cussed below).
o The IEMP provided an opportunity to learn about and use
various analytic tools for defining environmental prob-
lems and setting priorities. Project staff from EPA:
assembled various data bases at the request of the
Technical Advisory Committee to help identify a set of
32 potential pollution problems.
- helped develop methodologies to establish criteria to
rank these 32 problems and narrow this set for further
(Phase II) study.
- conducted a day-long seminar for the general public to
explain the methodologies and results of Phase I.
o We hope that this new way of doing business will become
a permanent feature of environmental management in
Baltimore.
C-l
-------
- We cannot be certain that the institutional arrange-
ments fostered will become a permanent part of the
Baltimore area's environmental infrastructure. We can,
however, take encouragement from the involvement of
representatives from the key jurisdictions and their
various environmental organizations and approximately
60 workgroup members, as indicative of a potential
group of apostles. Their investment also makes us
optimistic that the local and state governments will
implement our Phase II findings or reflect them in
future environmental management decisions.
The IEMP provided direct advice and guidance to Maryland's air
toxic program.
o The IEMP provided a conceptual basis (risk assessment)
for the proposed state program and guidance on how it
works in practice.
- EPA staff assisted the State in development of their
emissions inventory and provided practical assistance
in how risk assessment could be used to determine
acceptable emissions/exposure levels.
- EPA staff conducted preliminary monitoring over a six
month period and did the initial risk assessment work.
o The IEMP focused the State's attention on unusual sources
of toxic emissions and additional problems.
- We identified POTWs not only as a source of toxic
emissions but also as a classic case of intermedia
transfer of pollutants.
o The IEMP provided direct technical assistance to the
State.
- A member of the IEMP staff was a member of the advisory
panel for the State's program.
- EPA helped the state organize a one-day seminar for
citizens and industry. It dealt with the role of risk
assessment in the proposed program.
o The IEMP paved the way for ORD's Total Exposure Assessment
Methodology (TEAM) project in Baltimore.
- We provided the introduction to local officials and
are influencing the study design so its results will
be available for use in local policy decisions regard-
ing outdoor versus indoor air pollution.
C-2
-------
The IEMP has influenced a change in local codes on the use of
lead solder in residential plumbing.
o The IEMP convened a workgroup to examine the problem of
metals across environmental media. This fresh examina-
tion of human exposure to lead versus other metals showed
that this "well understood and controlled" pollutant
remains a major threat to human health.
- A simple analysis of hospital records (which showed
150 hospitalizations per year of children in Baltimore
City for lead chelation therapy) indicated that lead
remains one of the primary environmentally-caused
threats to human health. Further, our analysis of
potential risks from lead in drinking water in compari-
- son with other risks showed that lead in drinking water
can cause significant risks. This occurs primarily in
areas with corrosive water and in buildings with newly-
soldered plumbing.
'-r
The local plumbing board has moved voluntarily to re-
duce from 2% to 0.2% the amount of lead in solder and
flux used in Maryland's plumbing.
The I£MP is developing a methodology to help set priorities for
action among the wide variety of leaking underground storage tank
situations.
o The work may be applicable elsewherefor instance, in
Maryland's Ground-water Strategyin helping set priori-
ties for a planned state-wide monitoring program.
o This approach is also of major interest to EPA's Office
of Underground Storage Tanks for possible application in
other states and localities.
o EPA's Office of Research and Development is also inte-
rested in determining if our approach could be used to
set priorities for determining the location and frequency
of their program to monitor the effects of acid precipita-
tion on ground-water.
The IEMP identified and convened a group of city* county and state
officials who regulate activities affecting Baltimore Harbor.
o This coalition is one of the first successful efforts to
bring together representative of key agencies and to
work cooperatively on the Harbor.
The IEMP identified indoor air pollution as an issue for study.
o Partly as a result of our efforts, Maryland will seek EPA
assistance (from the Office of Radiation) to conduct a
, state-wide survey for radon during winter
* U.S. GOVERNMENT PRINTING OFFICE: 1987-716-002/60604 C"3
------- |